University of Southern California Dissertations and Theses

Consumer choice under budget constraint: Why consumers overspend

(USC Thesis Other)

Consumer choice under budget constraint: Why consumers overspend

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content CONSUMER CHOICE UNDER BUDGET CONSTRAINT:
WHY CONSUMERS OVERSPEND
Copyright 2002
by
Yun-Oh Whang
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BUSINESS ADMINISTRATION)
August 2002
Yun-Oh Whang
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 3094382
UMI
UMI Microform 3094382
Copyright 2003 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES. CALIFORNIA 90007
This dissertation, written by
Y b W -.Q h /. ..................................
under the direction of his. Dissertation
Committee, and approved by all its members,
has been presented to and accepted by The
Graduate School, in partial fulfillm ent of re
quirements for the degree of
DOCTOR OF PHILOSOPHY
of Graduate Studies
Date ....August.. 6*..£ Q .Q 2 .............
DISSERTATION COMMITTEE
Chairperson
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T a b l e o f C o n t e n t s
List of Tables...................................................................................................................... v
List of Figures................................................................................................................... vi
Abstract............................................................................................................................ vii
I. I n t r o d u c t io n 1
II. C o n c e p t u a l F r a m e w o r k ......................................................................................................5
1. Problem Definition........................................................................................ 5
Assumptions ........................................................................................................................................................................................................................5
Scope o f Research...................................................................................................................................................................................................7
2. Budget Formation Process...........................................................................10
3. Budget, Reference Price, Reservation Price, and Aspiration Price..........13
4. Theoretical Background...............................................................................15
Regulatory Focus ...............................................................................................................................................................................................16
Regulatory Reference...................................................................................................................................................................................18
Self-regulation and Prospect Theory.......................................................................................................................21
5. A Conceptual Framework for Decision Processes Under Budget
Constraint...................................................................................................... 25
Benefit Gain From Buying..............................................................................................................................................................27
Benefit Nongain From Not Buying................................................................................................................................29
Fund Loss From Buying .....................................................................................................................................................................31
Fund Nonloss From Not Buying..........................................................................................................................................33
An Integrated Conceptual Framework................................................................................................................34
Graphical Representation o f Overspending Behavior.........................................................36
III. E m p ir ic a l T e s t s o f T h e P r o p o s e d C o n c e p t u a l M o d e l .............................40
1. Pilot Study 1: Budget Formation Process.................................................. 40
2. Pilot Study 2: Overspending Behavior.......................................................42
3. Study 1: Consumer Choice Under Budget Constraint...............................48
Independent Variables and Manipulations..................................................................................................49
Task.................................................................................................................................................................................................................................................51
Dependent Variables...................................................................................................................................................................................53
Covariates..........................................................................................................................................................................................................................53
Hypotheses........................................................................................................................................................................................................................54
4. Study 2: Interaction Between Self-regulatory State and Absolute
Budget Level..................................................................................................64
Experimental Design...................................................................................................................................................................................65
Hypotheses........................................................................................................................................................................................................................67
ii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5. Study 3: Interaction Between Self-regulatory State and Core
Attribute Type................................................................................................68
Experimental Design 69
Hypotheses.............................................................................................. 70
IV. R e s u l t s ...............................................................................................................72
1. Study 1: Consumer Choice Under Budget Constraint............................... 72
Manipulation Checks_............................................................................. 72
Overspending Behavior..........................................................................................................................................................................77
Process Measures 80
Budget Revision vs. Violation................................................................ 87
Time 89
Emotional Responses. .................................................................................................................................................................................91
2. Study 2: Interaction Between Self-regulatory State and Absolute
Budget Level..................................................................................................94
3. Study 3: Interaction Between Self-regulatory State and Core
Attribute Type................................................................................................98
V. G e n e r a l D i s c u s s i o n ............................................................................................................. 104
1. Exclusivity vs. Addition Hypotheses........................................................104
2. Emotional Responses.................................................................................108
3. Interaction with Situational Variables.......................................................110
4. Validation of Proposed Conceptual Model Using Path Analysis 112
V I. T h e o r e t ic a l C o n t r ib u t io n s a n d M a n a g e r ia l I m p l ic a t io n s 118
1. Application of Self-regulation Theory to Consumer Context 118
2. Groundwork for Generalized Constraint Management Model................120
3. Expansion of Proposed Model to Underspending Behavior 121
4. Managerial Implications............................................................................ 121
VII. C o n c l u d i n g R e m a r k s , L i m i t a t i o n s , a n d F u t u r e S t u d i e s 124
R e f e r e n c e s 127
iii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
L is t o f T a b l e s
Table 1. Pilot Study 1: Response Summary.................................................................. 42
Table 2. Excerpts from Pilot Study 2: Why Overspend?..............................................44
Table 3. Pilot Study 2: Summary of Overspent Product Categories Reported........... 47
Table 4. Chronic Frequency of Activating A Self-regulatory State
(Pilot Study 2)................................................................................................... 56
Table 5. Contrast Analysis of Overspending Probabilities in Study 1 79
Table 6. Contrast Analysis of Reservation Price Importance Ratings in Study 1...... 82
Table 7. Contrast Analysis of Fund Availability Importance Ratings in Study 1...... 84
Table 8. Contrast Analysis of Anticipated Regret Importance Ratings in Study 1.... 86
Table 9. Contrast Analysis of Perceived Saving Importance Ratings in Study 1....... 86
Table 10. Factor Scores for Emotional Responses 92
Table 11. Parameter Estimates of Path Analysis Using Data from Study 1............114
iv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
L is t o f F ig u r e s
Figure 1. Human Self-regulation.................................................................................... 20
Figure 2. Consumer Self-regulation in Choice Context................................................27
Figure 3. A Conceptual Model for Consumer Choice Under Budget Constraint 35
Figure 4. Graphical Representation of Consumer Choice Under Budget
Constraint......................................................................................................... 37
Figure 5. Summary of Pilot Study 2: Chronic Self-regulatory Focus and
References........................................................................................................ 45
Figure 6. Overspending Probabilities in Study 1........................................................... 78
Figure 7. Importance Ratings of Reservation Price in Study 1.....................................81
Figure 8. Importance Ratings of Fund Availability in Study 1......................... 83
Figure 9. Importance Ratings of Anticipated Regret in Study 1...................................84
Figure 10. Importance Ratings of Perceived Saving in Study 1...................................85
Figure 11. New Budget After Choice............................................................................. 88
Figure 12. Emotional Responses Depending on Choice in Study 1........................... 95
Figure 13. Overspending Probabilities in Study 2 96
Figure 14. Overspending Probabilities in Study 3.........................................................99
Figure 15. Fund Focus Manipulation Check for Study 3............................................ 100
Figure 16. Budget Importance Rating During The Evaluation Stage in Study 3 102
Figure 17. Revised Reservation Price After Choice in Study 3..................................103
Figure 18. Path Analysis Results for Proposed Conceptual Model in Study 1......... 115
v
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A b s t r a c t
A conceptual model for consumer choice process under budget constraint is
proposed and empirically tested in a series of laboratory experiments. The model
builds upon self-regulation theory (Higgins 1997), modifying the notions of
regulatory focus and regulatory reference to fit a consumer context. Benefit (e.g., the
features to be acquired) and fund (e.g., the money to be spent) are defined as
regulatory foci, and overspending (e.g., spending over one’s budget) and not-
overspending (e.g., staying within the budget) as regulatory references. The
combination of these regulatory foci and references (i.e., self-regulatory state) should
activate the following motivations: 1) a benefit focus and an overspending reference
motivate consumers to maximize the benefit gain from overspending, 2) a benefit
focus and a not-overspending reference motivate consumers to minimize the benefit
nongain from not overspending, 3) a fund focus and an overspending reference
motivate consumers to minimize the fund loss from overspending, and 4) a fund
focus and a not-overspending reference motivate consumers to maximize the fund
nonloss from not overspending. The model further suggests that each of these
motivations makes a particular decision variable salient: 1) reservation price, 2)
anticipated regret, 3) fund availability, and 4) perceived saving, respectively.
The empirical tests provided support for the proposed relationships among
the self-regulatory states, motivations, and salient decision variables using
(M)ANOVA and path analysis. It is also found in an experiment with digital
vi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
cameras that the motivation to minimize the benefit nongain from not overspending
(benefit focus and not-overspending reference) is most effective in encouraging
consumers to overspend their budget. Another set of experiments indicated that
situational variables such as the absolute level of budget and the nature of the core
attribute of the product (e.g., functional vs. hedonic) influenced the type of
motivation activated instead of interacting with the decision making process.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I. I n t r o d u c t io n
When making a purchase decision, a consumer has to consider two aspects of
the exchange: 1) the features and benefits to be acquired through the purchase and 2)
the price to be paid for the purchase. It would not be a difficult decision if
consumers had unlimited resources. However, with 85 percent of the population in
the US earning less than $100,000 a year (Schor 2000), not every purchase is a
simple decision based only on desirable features and benefits. Instead, consumers
are challenged to make choices that maximize what they get considering how much
they have to give up.
The fact that an average consumer has limited resources brings an important
construct—budget—into consumer purchase behavior. Before making a purchase
decision, consumers often set a certain price level that they do not intend to go over
for the purchase; this is called ‘budget’ in laymen’s terms. A rational consumer
would set the budget at a level that maintains the balance between satisfying his
needs and safeguarding his overall finance. However, economic data show that
consumers often do not behave rationally in managing their finances. An average
American household has negative savings with an average unpaid credit card balance
of around $7,000 per household, and bankruptcy rates continue to grow from
200,000 a year in 1980 to 1.4 million in 1998 (Schor 2000). In short, consumers
tend to spend more than they can afford, which is defined as ‘overspending’ behavior
in this study.
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The evidence of overspending behavior is not limited to macroeconomic
statistics. A budget is often set for a single purchase to help a consumer keep track
of expenses and manage overall finance. However, it is not uncommon that the
actual spending goes over this budget. Suppose John wants to buy a house. He hires
an agent, and the first question the agent asks is what price level he has in mind.
John thinks about his various needs (e.g., square footage, number of bedrooms, etc.)
and how much he can afford based on his income stream and other necessary
expenses. He decides $200,000 would be the price he is willing to spend, and this
figure can be considered as the budget for John’s house purchase. However, John
soon finds out there are homes that are more desirable but more expensive than his
budget. At the same time, the agent tries to persuade John to go for a more
expensive home because it means more commission for the agent. It would be
interesting to see how many people in John’s shoes stay within the budget and buy a
home that costs no more than $200,000. When John spends more than $200,000 for
his home purchase, it is, in this study’s term, overspending.
This scenario leads to the questions of this study. Why do people overspend?
Why can’t people always stay within the budget and avoid accumulating debt? What
are the psychological processes people go through when they make a purchase
decision under a budget constraint? What are the key factors that govern the
overspending process? Unfortunately, neither marketing nor psychology has yet
provided a systematic understanding of these issues. Practitioners (e.g., real estate
agents and retail sales representatives) have apparently been using various sales
2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
tactics motivated by high commission or profit margin to encourage buyers to
overspend (e.g., Whittier 1994). However, it is not clear which tactic to use when
and how effective those tactics really are. Therefore, providing a theoretical
framework for consumer choice under budget constraint, which is the main objective
of this study, would be of great contribution to both marketing theory and practice.
This study attempts to theorize the consumer choice process under budget
constraint where overspending is one of the possible outcomes. Understanding the
choice process under budget constraint would allow researchers to shed light on why
and how people decide to overspend or stay within the budget. Chapter II lays out
the assumptions and scope of the present research along with the definitions of the
terms. Given the assumptions, a budget formation process is proposed, and the
recent development of self-regulation theory (Higgins 1997) in social psychology
serves as the theoretical backbone for developing a conceptual model behind choice
under budget constraint. The proposed model conceptualizes the various choice
processes under budget constraint and identifies the key decision variable for each
process. The external validity of the proposed model is tested through two pilot
studies (Chapter III). Pilot study 1 examines the budget formation process and its
role in consumer purchase, and Pilot study 2 confirms through content analysis of
essay responses that consumers indeed activate one or more of the proposed choice
processes. Based on the findings in the pilot studies, three studies are designed to
test the proposed conceptual model and the role of two situational variables (Chapter
III). First, a set of hypotheses formulated from the proposed conceptual model is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
empirically tested in Study 1. The respondents are manipulated to activate one of the
decision processes conceptualized in the proposed model and are asked to make a
purchase decision under a budget constraint. The dependent variables include
overspending likelihood and cognitive and affective responses. Two additional
studies are designed to empirically test the moderating effects of two situational
variables on overspending behavior. The situational variables were the nature of the
core attribute (functional vs. hedonic) and the absolute budget level (high vs. low)
for Study 2 and 3, respectively. Integrating the proposed theoretical model tested in
Study 1 and the past research on these two situational variables leads to a set of
predictions on overspending behavior, which are empirically tested in two
experiments. The results from the studies described above provide strong support for
the proposed conceptual model for consumer choice processes and their affective,
cognitive, and behavioral outcomes (Chapter IV). Summarizing the results, a
general discussion is presented in Chapter V focusing on the findings that are not
consistent with the predictions. The present research’s theoretical contributions and
managerial implications (Chapter VI) are summarized later and are followed by the
concluding remarks and future research (Chapter VII).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
II. C o n c e p t u a l F r a m e w o r k
1. Problem Definition
Assumptions The present study starts with two basic assumptions: 1)
consumers have limited resources and 2) consumers set a budget for an individual
purchase incidence. The notion of limited resources has to be an assumption because
the availability of resources is a perception rather than a fact. A consumer could
keep spending regardless of the income until bankruptcy as if he had unlimited
resources, and another consumer with enough income not to worry about his
expenses would be very sensitive about his spending. For the former, there is no
need for a conceptual model to explain the spending behavior because they tend to
choose the most desirable alternative without considering the cost. Budget rarely
plays an important role in their purchase processes. Since there is no budget to
compare the actual spending against, overspending defined as spending over the
budget becomes meaningless. It is more interesting to examine consumer choice
behavior given that consumers set budgets in recognition of limited resources.
Therefore, it is assumed in this research that consumers perceive themselves to have
limited resources, and that they put effort into balancing their checkbooks. Pilot
Study 1 will test the external validity of this assumption.
The second assumption is that consumers do set a budget for an individual
purchase. Since budget is a construct that has been understudied, we still do not
have a clear understanding on the budget formation process. Remote evidence in
5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
economics and consumer behavior literature indicates that consumers set budgets at
the expense category level (for a review, see Zelizer 1993). The early works by
Bakke (1940) and Rainwater, Coleman, and Handel (1959) introduced the notion of
separate ‘envelopes,’ ‘china pitchers in the cupboard,’ and ‘tin cans’ where people
often physically placed funds for various expense categories after setting the budget
for each. More recently, Heath and Soil (1996) argued that consumers indeed have
mental budget categories to which they allocate their available funds, and when they
overspend in one budget category, consumers transfer funds among the categories to
balance the checkbook. Despite the ample research on the category level budget
allocation behavior, the evidence for setting a budget for a single purchase incidence
is hard to find. The only relevant study by Brockner and Rubin (1985) showed in an
experiment regarding personal investment decision making that the majority of
subjects spontaneously set an upper limit (i.e., budget) up front on their expenditure
for each individual investment decision.
The assumption of budgeting at the individual purchase level seems to be a
logical extension of the evidences found at the category level. However, many
questions about this assumption remain unanswered. When do consumers set a
budget for a single purchase? What are the products/services for which consumers
set a budget for a single purchase? These questions are beyond the goal of the
present study, so for a practical reason, the present research assumes that consumers
set budgets at the individual purchase level. Pilot Study 1 will attempt to verify the
external validity of this assumption, and more discussion on this issue will be
6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
presented in Chapter VII (Concluding Remarks and Future Studies) because the
questions around this assumption are worth revisiting later as a separate study.
Scope o f Research Given the assumptions that consumers have limited
resources and set a budget for a single purchase, the present research limits its focus
to a consumer choice context with the following characteristics. First, the present
research focuses on rational purchase behavior where consumers make deliberate
choice toward value maximization under perfect market information. It would be
more than possible to explain overspending behavior with irrationality (e.g., “I just
like this product”), bias (e.g., “I like this product because it is from Japan”), or
imperfect marketing information (e.g., “I didn’t know this product had such a great
design”). However, overspending behavior is not limited to the purchase occasions
where irrationality, bias, or imperfect market information exists. The present study
attempts to formulate a theoretical model that explains overspending behavior even
without these factors, and it would provide the most conservative test of the
proposed model.
This limited scope of research also differentiates the present study from the
research on unplanned buying (e.g., Iyer 1989) and impulse buying (e.g., Rook 1987;
Eagle and Blackwell 1982). These behavioral patterns can be considered as an
extreme cases of overspending because consumers do spend over the budget, which
is set to zero before engaging in the purchase process. However, this study is more
interested in overspending behavior in the context of planned and deliberate purchase
process where overspending behavior often occurs. For example, buying a home as
7
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
presented in the previous chapter is generally a planned purchase with an extensive
search process and requires a lengthy deliberation, and yet it is not uncommon to see
people overspend. I believe the overspending behavior in a planned and deliberate
purchase occasion needs a different conceptual framework from unplanned or
impulse buying, and therefore, the present study proposes a theoretical model that is
based on a context that does not fall into the category of unplanned or impulse
purchases. It is important to clarify the scope of research as defined above because
the findings provide insights for the case where consumers have all the resources
(e.g., time and information) to make a rational choice but still fail to do so.
In short, the present study limits its scope to a rational and deliberate
purchase under perfect market information given that consumers have limited
resources and set a budget for a single purchase. In this context, what the present
study attempts to investigate is the choice behavior under a budget constraint with
two or more alternatives, all of which are on the same value maximization line where
the price-performance ratio is constant. Since all the alternatives are on the same
value maximization line, consumers are expected to be indifferent among the
alternatives that equally maximize the value. For example, suppose John’s house
shopping leaves him with two finalists: A) a $200,000 home with 3 bedrooms, 2
bathrooms, 2,000 sqft and B) a $220,000 home with 3 bedrooms, 2 bathrooms, 2,200
sqft, and a pool. Without considering the price, it is an easy decision to buy a bigger
home with a pool (Home B). However, with the price attribute, John faces a
challenge to decide if he wants to give up the extra $20,000 to buy the more
8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
attractive Home B. If John perceives the extra benefits from Home B are exactly
worth the extra $20,000, both homes are considered to be on the same value
maximization line. In other words, both homes provide exactly what the money is
worth (“bang for the buck”). John’s decision is predicted to be a coin flip (i.e., 50-50
chance) because he does not have a preference based on the attributes of the homes
including the price. However, the introduction of budget constraint changes this
picture. When John does not have a preference between the homes, a budget
constraint becomes the determining factor and leads to a decision to buy Home A
and stay within the budget. The budget constraint could still keep John from buying
the more expensive Home B even if he perceives the extra space and the pool to be
worth more than the extra $20,000 and prefers Home B. However, to be more
conservative, the present study focuses on the context where John does not have a
preference and is supposed to stay within the budget. By doing so, the proposed
theoretical model would be able to help us understand the explicit role of budget in
purchase decisions without the confounding with preference. In other words, it is
predicted that a rational choice results in no overspending in the context described
above, and the present study examines why we observe overspending and attempts to
explain why.
The clearly defined scope of research is very important for developing a
parsimonious theoretical framework. By limiting the scope of research to a context
where there is no inherent preference, the conceptualization can focus on the
decision criterion of interest—budget. This will enable us to identify the key
9
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
variables that influence the effect of budget on consumer choice and further extend
the proposed framework to the context where preferences do exist. In addition, the
operationalizaton of the relevant constructs becomes cleaner, and, as a result, it
improves the precision of the empirical tests of the proposed model.
2. Budget Formation Process
Considering the fact that budget formation is an antecedent to overspending
behavior, it is necessary to define what budget is and conceptualize how it is set. As
discussed in the previous section, there is evidence that consumers do set budgets at
the expense category and single purchase levels. However, it is unfortunate that past
literature fails to provide a conceptual framework on how the budget figures are set.
What do consumers consider when allocating a certain amount of budget for a
purchase? This section provides a simple conceptual model for budget formation
process and uses it as a steppingstone to develop a process model for overspending
behavior.
I suggest that there are two sources of information consumers use to set a
budget. First, consumers retrieve and/or search for the information about the market.
Market information includes product specifications, price distribution, customer
evaluations, and so on. Based on the compiled market information, a consumer
figures out how much he is willing to pay for a product with the desirable set of
features and benefits. This willingness-to-pay can be labeled as the consumer’s
‘reservation price’ based on its definitions in economics and marketing (e.g., Emery
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1970; Monroe 1973; Weiner 1988). Economists define reservation price as the price
at which a consumer does not resist the purchase (e.g., Scherer 1980), and it is
defined in marketing as the maximum price a consumer is willing to pay (e.g., Rao
and Gautchi 1982). Both definitions lead to an important implication that consumers
would eliminate alternatives from the choice set when those alternatives exceed the
reservation price (Monger and Feinberg 1997), which is similar to the role of budget
in choice behavior. However, reservation price, as defined above, is not an identical
construct to budget. Since reservation price is narrowly defined to be formed based
only on market information in this study, it is, in principle, possible to set a
reservation price for a product/service that is not affordable due to lack of funds.
More specifically, budget is defined in the present study as the product of
willingness to pay and the amount of available financial resources (‘wallet
information’). It is possible that consumers form a reservation price through an on
going information search but do not have a budget for purchase due to limited fund
availability. Therefore, budget is influenced by but not identical to reservation price.
The wallet information includes income stream, disposable cash in hand, savings,
credit lines, etc., and it is defined as ‘fund availability’ in this study. Fund
availability at a household level is often called ‘buying power’ or ‘spending power’
in economics. While reservation price represents the willingness to pay for a product
or service based on the market information, fund availability determines the ability to
pay based on the wallet information.
11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Both reservation price and fund availability have a positive influence on the
budget figure. The higher the reservation price and fund availability, the higher the
formed budget would be. However, reservation price and fund availability are
independent of each other because they do not share the same set of antecedents.
The antecedents for reservation price include product and market information, while
those for fund availability are related to the consumer’s finance, which has nothing
to do with the product or market. For example, consumers sometimes form a
reservation price for a product they cannot afford (e.g., “I am willing to pay the
price, but I don’t have that much money”), or they can afford it but are not willing to
pay the price (e.g., “I do have enough money, but I am not willing to pay that much
for it”). Fund availability serves as a ceiling for the budget in the former case, and so
does the reservation price in the latter. In these extreme cases, the budget is equal to
either the fund availability or the reservation price, whichever is lower.
Suppose John initially had a reservation price of $200,000 for a 3-bedroom
home and was able to afford $250,000. The budget is set somewhere between these
two figures, and it would be set at $200,000 (lower of the two) in an extreme case
because John is not willing to pay more than $200,000 even though he can afford
more. On the other hand, suppose John initially had a reservation price of $250,000
and was able to afford only $200,000. The budget is the same at $200,000 in an
extreme case because John cannot afford more even though he is willing to pay
more. The same budget of $200,000 is set by two different processes in these cases.
In reality, a consumer would try to find an optimal budget figure after making
12
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
adjustments to the willingness-to-pay (reservation price) and/or the ability-to-pay
(fund availability) if there is any difference.
3. Budget, Reference Price, Reservation Price, and Aspiration Price
The budget formation process discussed above leads to the definition of
budget as conceptualized in the present study. With reservation price and fund
availability as its antecedents, budget is set by a consumer at an optimal amount he is
willing and able to spend for a purchase. One could contend budget as equivalent to
the concept of reference price, reservation price, or aspiration price. Keeping the
definition of budget presented above, there is a clear distinction between these
concepts and budget. First, there are a number of operationalizations of reference
price reported in marketing literature (for review, see Briesch, Krishnamurthi,
Mazumdar, and Raj 1997). However there is only one study that incorporated non-
market information in its operationalization by considering deal proneness of the
household (Kalwani, Yim, Rinne, and Sujita 1990). The other studies used strictly
market information (e.g., past prices, current prices, price trend, promotion
frequency, and market share). In short, reference price is often defined without
considering the financial situation (fund availability) of the consumer, while the
present study posits that budget is influenced by fund availability as well.
As mentioned above, reservation price is also defined as a construct that is
formed based on market information only. This definition of reservation price is in
line with past literature that theorizes reference price as a determinant of (Monger
13
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and Feinberg 1997) or equivalent to reservation price (Emery 1970; Monroe 1973;
Weiner 1988), which suggests that reservation price is also independent to fund
availability.
The same argument applies to aspiration price. Klein and Oglethorpe (1987)
defined aspiration price as one type of reference price— ‘the price one would like to
pay.’ The literature on negotiation provides a more specific definition that
hypothesizes aspiration price as ‘the buyer’s belief about the seller’s reservation
price (lowest acceptable price)’ (Kreistensen and Garling 1997; White and Neale
1994). In these previous studies, aspiration price is again defined as an outcome of
predominantly market information (e.g., seller information). Therefore, the present
study postulates that budget is a different construct from reference price, reservation
price, and aspiration price.
In short, the major distinction is the fact that budget is strongly influenced not
only by market information but also by wallet information (fund availability).
Suppose a consumer perceives $50,000 as a fair price for a Porsche (reference price),
is willing to pay up to $52,000 (reservation price), but would like to pay $45,000 if
possible (aspiration price). The consumer would set a budget somewhere between
$45,000 and $52,000 depending on the fund availability if a Porsche is affordable, or
does not have a budget at all (i.e., no intention to make a purchase at the moment)
because the consumer cannot afford one. This example illustrates that reference,
reservation, and aspiration prices can be formed regardless of fund availability.
14
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
It is presented above a simple conceptualization of budget formation process,
and it is critical to the conceptual model development for overspending behavior.
Budget formation process must be more complex than described above with many
other individual, contextual, and situational factors affecting the outcome. Since the
present study’s focal problem is overspending behavior and the budget can be
controlled fairly easily in an experimental setting, the present study does not attempt
to test the proposed budget formation process. A pilot study is conducted to gain
support for the proposed budget formation process, and a further investigation on
budget formation is left for future studies.
4. Theoretical Background
Given that a budget is formed based on reservation price and fund
availability, a conceptual model for purchase decision processes under budget
constraint is presented in this section. Suppose a consumer has to decide between
two alternatives. One fits the budget, and another goes above. Assuming that they
both offer exactly what the money is worth (i.e., equal value maximization), a
rational consumer would choose the alternative of which the price is less than or
equal to the budget. However, as discussed in the previous chapter, consumers do
not always make a rational choice. This anomaly is the focus of the present study,
and identifying the motivational forces that influence the choice process is crucial to
developing a conceptual model.
15
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
There are two independent motivations at work toward the opposite direction.
One encourages the consumer to choose the alternative with desirable features and
benefits that provides pleasure. Another discourages the consumer to spend over the
budget because it causes pain by taking away the resources that were to be spent on
something else. In other words, making a choice under a budget constraint can be
theorized with ‘hedonic principle’ that governs the motivation to approach pleasure
and avoid pain. The role of hedonic principle in human behavior has long been
documented in marketing and psychology. Conceptualized as an approach-
avoidance conflict, hedonic principle has served as the motivational assumption for
the theories in cognitive and organizational psychology (e.g., Dutton and Jackson
1987; Edwards 1955; Kahneman and Tversky 1979), expectancy-value models (e.g.,
Feather 1982; Sheppard, Hartwick, and Warshaw 1988; Wilkie and Pessimier 1983),
consistency theories in social psychology (e.g., Festinger 1957; Fleider 1958), and
achievement motivation (Atkinson 1964).
Regulatory Focus More recently, self-regulation theory in psychology
further extends hedonic principle and provides a more systematic conceptualization
of human motivation by introducing the constructs called regulatory focus and
regulatory reference (Fliggins 1997). Hedonic principle posits that human
motivation approaches desired end-states and avoids undesired end-states, but self
regulation theory further extends the hedonic principle by specifying different ways
of approaching (avoiding) different types of desired (undesired) end-states. Self
regulation theory suggests that different self-regulatory foci are operational under
16
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
different types of desired or undesired end-states. Following the categorization by
Higgins (1997, 1999), there are: 1) nurturance related (e.g., aspiration and
accomplishment) and 2) security related (e.g., responsibilities and safety) desired
end-states. Nurturance related end-states and security related end-states are defined
as promotion focus and prevention focus, respectively. The motivation is shaped
around gaining or not gaining strong ‘ideals’ of attaining advancement in promotion
focus, and around losing or not losing strong ‘oughts’ of insuring safety in
prevention focus. In other words, promotion focus has a motivational consequence
of approaching positive outcome and prevention focus has a consequence of
avoiding negative outcome.
In the context of the present study, consumers with promotion focus would
put more emphasis on acquiring more features and benefits in a purchase, which is
an accomplishment to be gained. On the other hand, consumers with prevention
focus would concentrate on not giving up too much money on the purchase, which is
financial safety to be lost. For example, John, who is shopping for a home, would be
determined to own a dream home and feel the accomplishment from the purchase.
Buying a less desirable home is not necessarily a negative outcome, but buying a
dream home is definitely a positive outcome. Here John focuses on a nurturance
related end-state (approaching the positive outcome of gaining a dream home),
which is a promotion focus. Alternatively, John would be determined to maintain his
finance by finding a home that is not going to cost him an arm and a leg. Not
spending too much is not necessarily a positive outcome, but going over the budget
17
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
is definitely a negative outcome. He is in a prevention focus where a security related
end-state (avoiding the negative outcome of overspending) is what is to be
maintained in this case. In short, promotion focus deals with acquiring the benefits
while prevention focus involves maintaining the finance. Therefore, a promotion
focus is defined as a ‘benefit’ focus and prevention focus as a ‘fund’ focus in the
present study’s specific context of consumer choice.
Regulatory Reference Another construct in self-regulation theory that
offers important insights to consumer decision making is regulatory reference
(Higgins 1997). As defined above, regulatory focus involves the valence of the
outcome (positive vs. negative) where people approach the presence of a positive
outcome (promotion focus) and avoid the presence of a negative outcome
(prevention focus). Conceptualizing different types of positive and negative
outcome provides us with another set of motivational consequences. More
specifically, the positive outcome in promotion focus can be either the presence of a
desired end-state or the absence of an undesired end-state (Markus and Nurius 1986).
The desired end-state is perceived as gain, and the undesired end-state as nongain in
promotion focus. As a result, people approach both the presence of gain and the
absence of nongain. Similarly, the negative outcome in prevention focus can be
either the presence of an undesired end-state or the absence of a desired end-state.
The former is perceived as loss and the latter nonloss. As a result, people avoid both
the presence of loss and the absence of nonloss (Idson, Liberman, and Higgins
2000).
18
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
For example, John may not consider not having an extra room for an office at
home as a negative outcome. However, having an office surely is a positive
outcome, and he would be motivated to approach either having one (presence of a
desired end-state) or making sure he does not buy a house without one (absence of an
undesired end-state). Presence of a desired end-state and absence of an undesired
end-state are both positive outcomes, but the behavioral outcome would be different
depending on the motivational direction where different regulatory references are in
place. Similarly, spending too much money on a house is a negative outcome while
spending just the budgeted amount is rather a status quo to be maintained than a
positive outcome. John would definitely avoid either spending too much (presence
of undesired end-state) or failing to stay within the budget (absence of desired end-
state). In other words, desired and undesired end-states, which are called regulatory
references, can activate both promotion (approach) and prevention (avoidance) foci
(Carver and Scheier 1981, 1990).
The introduction of regulatory reference is a significant theoretical
contribution to understanding human motivation and behavior. For example, self
regulation with desired end-state reference has long been reported to involve
behavioral production while undesired end-state reference involves behavioral
inhibition or suppression (e.g., Estes 1944; Gray 1982; Lewin 1935; Skinner 1953;
Thorndike 1935). Another example is the asymmetry between the closed nature of
desirable end-states and the openness of undesirable states. People have a certain
19
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
destination when approaching a desirable end-state but can infinitely move away
from an undesirable state (Carver and Scheier 1981, 1990).
Integrating regulatory focus and regulatory reference provides us with a 2x2
matrix for self-regulation with promotion vs. prevention focus and desired vs.
undesired end-state reference as its two independent dimensions (see Figure 1).
Figure 1. Human Self-regulation
Regulatory Reference
Desired end-state Undesired end-state
Regulatory
Promotion focus
Approach
the presence
of desired end-state
Approach
the absence
of undesired end-state
Focus
Avoid Avoid
Prevention focus the absence the presence
of desired end-state of undesired end-state
Applying this matrix to a consumer purchase leads to the following
definitions of regulatory references in the context of the present study. Desired end-
state reference includes approaching the presence of a desired end-state (promotion
focus) by acquiring desirable features and benefits from a purchase or avoiding the
absence of a desired end-state (prevention focus) by not spending more than the
budget. Here the consumer asks himself a question that starts with, “if I buy the
alternative with more features and benefits, ...” or “if I buy the expensive alternative
and spend over the budget, ...” where both questions feature the alternative that costs
20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
more than the budget as the reference. On the other hand, undesired end-state
reference represents approaching the absence of an undesired end-state (promotion
focus) by not losing the desirable features and benefits from a purchase or avoiding
the presence of an undesired end-state (prevention focus) by spending more than the
budget. Here the consumer asks himself a question that starts with, “if I don’t buy
the alternative with more features and benefits, ...” or “if I don’t buy the expensive
alternative and stay within the budget, ...” where both questions posit the alternative
that fits into the budget as the reference. Therefore, the present study calls desired
end-state reference ‘buying’ reference, and undesired end-state reference ‘not-
buying’ reference.
Self-regulation and Prospect Theory It is important to clarify how the
regulatory focus used in the proposed conceptual framework differs from the
prospect theory (Kahneman and Tversky 1979). The conventional manipulation of
framing in prospect theory literature leads people to attend to either the positive or
the negative consequences of a decision while regulatory focus manipulation
provides both positive and negative consequences of an occurrence (Brockner and
Higgins 2001). For example, a framing manipulation in the context of consumer
choice under budget constraint would state, ‘you acquire a desirable product if you
buy this expensive alternative’ vs. ‘you spend over the budget if you buy this
expensive alternative’ while a regulatory focus manipulation would state both.
In other words, regulatory focus does not manipulate the valence of the
outcome. Instead it manipulates the nature of the information by varying the nature
21
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
of the goal to achieve. For example, acquiring a desirable product and not spending
too much for the purchase are both positive outcomes in consumer choice under
budget constraint. Regulatory focus makes a distinction between these two different
types of positive outcome by identifying the goal as promotion and prevention,
respectively. It is important to note that prospect theory is not concerned about the
difference between these two outcomes because they are both positive consequences.
In short, regulatory focus examines the effect of the nature (i.e., promotion vs.
prevention) rather than the valence (positive vs. negative) of the information
presented.
Borrowing the notions of regulatory focus and regulatory reference from self
regulation theory in psychology, a 2x2 matrix for the motivations behind consumer
purchase decision can be formulated. As explained in the previous section, one
dimension represents the regulatory focus with benefit (promotion) and fund
(prevent) foci, and another represents the regulatory reference with buying (desired
end-state) and not-buying (undesired end-state) references. The reason for using a
different terminology from self-regulation literature in psychology is that it better
represents the intuitive connotation attached to the construct in a consumer purchase
setting. More specifically, consumers could be in prevention focus when attributes
such as termite control, security system, and fire alarm for a home are considered. In
other words, fund is not the only decision variable of a purchase transaction that
triggers prevention focus by the buyer but one of many decision variables.
22
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Therefore, using the term prevention focus in the context of the present study could
be misleading.
The same argument applies to regulatory reference. There are multiple
desired and undesired end-states that could serve as a reference other than buying
and not-buying, such as upgrading/downgrading the product category (e.g., from 3-
bedroom house to 4- or 2-bedroom house) and deferring the choice, which could
shift or even reverse the desirability level of the end-state. For example, suppose
John starts his house hunting with 3 bedrooms in mind. If only low-end 3-bedroom
houses fit the budget, staying within the budget is an undesirable end-state in benefit
focus. However, downgrading the choice set to 2-bedroom houses suddenly puts
most of the high-end 2-bedroom houses under the original budget and makes them
desirable for benefit focus in the new choice set. To eliminate this kind of ambiguity
and control the construct validity, the regulatory reference is defined strictly to
staying within the budget vs. overspending in the given choice set without allowing
any changes in the choice set. Therefore, the levels of regulatory reference in the
present study are labeled with ‘buying’ and ‘not-buying.’
These strict definitions of regulatory focus and regulatory reference
constructs as discussed above do not hurt the external validity of the proposed
conceptual model. First, the objective of the model development is to understand the
differential effect on the motivation and behavioral outcome by benefit and fund
foci, and the model conceptualizes benefit and fund foci as a specific case of
promotion and prevention foci, respectively. In other words, benefit focus can
23
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
always be considered as promotion focus, but promotions focus does not always
represent benefit focus, which means benefit focus is a subset of promotion focus.
Similarly, it is assured that fund focus is always a specific treatment of prevention
focus by defining it as a subset of prevention focus. It allows a theoretical model
building based on the past findings regarding promotion and prevention foci given
that benefit and fund foci exclusively lead to promotion and prevention foci,
respectively. Regulatory reference is also narrowly defined in a specific context of
consumer purchase without the ability to change the choice set or defer the choice in
the present study. The definition limits ‘buying’ and ‘not-buying’ as a subset of
desired and undesired end-state references, respectively, which again warrants the
adoption of the past findings regarding regulatory reference in the present study’s
conceptual model building.
5. Conceptual Framework for Decision Processes Under Budget Constraint
The application of regulatory focus and reference to the present research
context of consumer choice under budget constraint has been discussed above. In
short, a consumer in a conflict between buying a more attractive alternative by
overspending and staying within the budget by buying a less attractive alternative is
expected to activate one of the regulatory foci and references to initiate the process
of resolving the conflict. The activated regulatory focus and reference is called ‘self-
regulatory state’ in the present study. The self-regulatory state (focus and reference)
shapes the decision process by laying out the main question the consumer would ask.
24
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
For example, a consumer who focuses on the ‘benefit’ and takes ‘buying’ as
the reference asks a question, ‘what benefit do I get if I buy the expensive alternative
and spend more than my budget?’ Similarly, a consumer with ‘benefit’ focus and
‘not-buying’ reference tries to answer the question, ‘what benefit do I lose if I do not
buy the expensive alternative and stay within the budget?’ If the consumer is drawn
to (‘buying’ reference) or cannot let go of (‘not-buying’ reference) the extra benefits
from overspending, overspending is more likely to take place. If not, the consumer
stays within the budget. The extra benefits are defined in the present study as gains
when acquired and nongains when not acquired.
Under the ‘fund’ focus, the questions become, ‘how can I pay for the extra
over the budget if I buy the expensive alternative?’ (‘buying’ reference) and, ‘how
much do I save if I do not buy the expensive alternative and not spend the extra?’
(‘not-buying’ reference). If the consumer can come up with a way to pay for the
extra (‘buying’ reference) or perceive the potential saving to be trivial (‘not-buying’
reference), overspending is likely to occur. If not, the consumer stays within the
budget. The extra expense over the budget is defined in the present study as loss
when spent and nonloss when saved.
These key questions suggest four distinctive decision variables that influence
the decision processes consumers experience through which consumers elect to
overspend or not. Consumers with benefit focus and buying (not-buying) reference
weigh the ‘benefit gain (nongain) from buying (not buying)’ more than any other
decision variables, and consumers with fund focus and buying (not-buying) reference
25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
weigh the ‘fund loss (nonloss) from buying (not buying)’ more. Therefore, the
proposed conceptual model labels the four motivations behind overspending
behavior after these key variables: 1) maximizing benefit gain from buying, 2)
minimizing benefit nongain from not buying, 3) minimizing fund loss from buying,
and 4) maximizing fund nonloss from not buying (see Figure 2). As mentioned
earlier, ‘buying’ in these labels means ‘buying an alternative with more desirable
features and benefits and spending over the budget,’ and ‘expensive’ means the price
is over the budget.
The next step for theoretical model building is to identify the subsequent
decision making processes of these self-regulatory states. A process model is
suggested by linking the four types of motivation drawn from self-regulatory states
to the marketing literature. The model identifies a distinctive marketing variable for
each self-regulatory state and hypothesizes its influence on the consumer choice to
overspend or not.
Benefit Gain From Buying Self-regulation theory suggests that people in
promotion focus and desired end-state reference—the state parallel to benefit gain
from buying in consumer choice context—are motivated to approach the presence of
a desired end-state. The desired end-state in consumer purchase would be the
acquisition of an alternative with extra features and benefits, and the approach
motivation toward the presence of it leads to overspending behavior. If a consumer
does not consider the extra features and benefits attractive, the consumer is unlikely
26
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
to be motivated to maximize the benefit gain from buying and is likely to stay within
the budget.
Figure 2. Consumer Self-regulation in Choice Context
Regulatory Reference
Regulatory
Focus
Benefit focus
Fund focus
Buying Not Buying
Maximize Minimize
Benefit Gain Benefit Nongain
From Buying From Not Buying
Minimize
Maximize
Fund Loss
Fund Nonloss
From Buying
From Not Buying
It is shown above that benefit gain from buying condition posits the perceived
worth of the features and benefits as the main decision variable. Considering that the
features and benefits of the alternatives in the choice set are the core of market
information, and market information is the key determinant of the reservation price
as discussed earlier, the model posits that reservation price plays a critical role in
benefit gain from buying condition. More specifically, it is postulated in the
proposed conceptual model that consumers in benefit gain from buying condition
(benefit focus and buying reference) assess the features and benefits of the
alternatives and revise the reservation price accordingly. Since reservation price is
identified as one of the determinants for budget, revising the reservation price has an
impact on the budget. A raised reservation price is likely to boost the budget. Since
27
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the budget itself escalates after raising the reservation price, choosing a more
expensive alternative, in principle, cannot be considered as overspending because it
is not over the revised budget. However, the present study categorizes this case as
overspending because the consumer still spends more than the original budget set
before engaging in the choice task. It makes practical sense because the original
budget is often the only measure that is observable by marketers because budget
revision is processed on the fly during the decision making process.
Suppose John, who is in the market for a house with a $200,000 budget,
narrows down his choice set to two houses: 1) a $200,000 home with 3 bedrooms, 2
bathrooms, and 2,000 sqft, which is right on the budget and 2) a $220,000 home with
the same specification with 300 sqft extra space and a pool, which runs over the
budget. When motivated to maximize the benefit gain from buying, he becomes
more willing to pay extra for more space and a pool than he initially was, which
translates to a higher reservation price. When setting the initial budget of $200,000,
suppose he was strongly influenced by the reservation price of $200,000 along with
the fund availability level of $250,000. The reservation price served as the ceiling
that capped the budget at $200,000. As the reservation price grows, the budget gains
the room to grow as well. On the other hand, raising the reservation price does not
affect the budget significantly if the original budget is capped by the fund availability
(e.g., reservation price of $250,000 and fund availability of $200,000). The fund
availability still remains as the key determinant of the budget, and the impact of
inflated reservation price is discounted.
28
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In short, the proposed conceptual model posits that consumers with benefit
focus and buying reference make a choice between overspending or not by revising
the reservation price. The revised reservation price subsequently influences the
budget and ultimately determines the choice outcome. An overspending decision is
made when the benefit gain from buying is maximized by inflating the reservation
price.
Benefit Nongain Front Not Buying Self-regulation theory suggests that
consumers with promotion focus and undesired end-state reference are motivated to
approach the absence of an undesired end-state. Undesired end-state in a consumer
choice context would be not buying the desirable alternative because of the budget,
and the absence of it is the positive outcome consumers approach. The decision
becomes whether or not to buy an inexpensive alternative and give up the desirable
features and benefits of the expensive one. Giving up the desirable features and
benefits is conceptualized in the present study as nongain. It is different from loss
because not having the desirable features and benefits is not necessarily a loss.
Instead, it is a gain that is given up and not attained. This self-regulatory state,
therefore, is called benefit nongain from not buying, and consumers who are
motivated to minimize it are likely to overspend. If consumers consider the nongain
insignificant, overspending is unlikely to occur.
The attempt to understand this self-regulatory state leads to a question of
what consumers try to minimize. Nongain involves the features and benefits that
consumers could enjoy, and not being able to enjoy them would trigger regret in
29
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
consumers’ minds. Regret is defined in psychology as the sense of sorrow or
disappointment over something done or not done (Landman 1987). People are also
found capable of anticipating the potential regret associated with their action and
making a decision with the anticipation of regret (Bell 1982). In other words, what is
minimized by the motivation to minimize nongain is anticipated regret, and the
motivation to minimize anticipated regret often leads to overspending. Fear appeal is
one of the examples of emphasizing the negative outcome and the regret linked to it
(e.g., Keller and Block 1996; Tanner, Hunt, and Eppright 1991). In an experiment
with two alternatives to choose from, Simonson (1992) also reported that consumers
are more likely to choose a better-known, more expensive brand than a less-known,
less expensive brand if they anticipate regret.
Based on these findings, the proposed conceptual model assigns anticipated
regret as the key decision variable for the consumers in benefit nongain from not
buying condition. For example, John in his choice process would anticipate regret of
not being able to enjoy the extra space and pool by not choosing the expensive
alternative. The final choice depends on the magnitude of anticipated regret. The
bigger the anticipated regret, the more likely John is to overspend. Note that
anticipated regret in benefit nongain from not buying condition does not influence
budget while the key variable of benefit gain from buying condition—reservation
price—influences the decision through the change in budget. More discussion on
this issue is presented later in this chapter.
3 0
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fund Loss From Buying In prevention focus and desired end-state
reference, it is suggested by self-regulation theory that people are motivated to avoid
the absence of a desired end-state. Prevention focus and desired end-state reference
translate to fund focus and buying reference in the context of consumer purchase. In
other words, consumers focus on the money they have to give up for the purchase
and posit spending more than the budget by buying the expensive alternative as the
reference point. Consumers would want to minimize the fund loss from buying the
expensive alternative. The more a consumer spends, the more loss of fund the
consumer has to suffer.
This condition has direct relation to fund availability, the construct defined in
budget formation process earlier. With ample fund availability, fund loss from
buying would not be perceived as significant so that overspending is likely to occur.
With lack of fund, the loss would keep the consumer from overspending. Therefore,
the model posits that fund availability is the main decision variable for the
consumers in the fund loss from buying condition. Since buying is the reference, the
fund loss is specifically defined as the difference between the budget and the price of
the expensive alternative. To minimize or even eliminate the loss, consumers would
raise the fund availability, which in turn raises the budget. Not buying the expensive
alternative also eliminates the loss, but it is the case where not-buying serves as the
reference, which will be discussed later. In short, under fund focus and buying
reference, consumers make a choice based on the perceived loss of fund and are
31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
likely to overspend when the perceived loss is minimized by increasing the fund
availability.
For example, John would perceive the $20,000 difference between an
attractive home ($220,000) and the home within his budget ($200,000) as a fund loss
if he decides to buy the attractive home. Motivated to minimize the loss, John would
come up with a way to earn or secure an extra $20,000 he can spend on the home
purchase such as working extra hours and postponing a car purchase. By doing so,
John’s fund availability and, subsequently, his budget goes up. The increased budget
of $220,000 eliminates the difference between the budget and the price to pay, and
John gets rid of the perceived fund loss. If John cannot come up with a way to
eliminate the perceived loss, overspending is unlikely to take place. In addition, if
the budget is set with reservation price not fund availability as its cap, the increase of
fund availability would have a discounted effect on the budget. For example,
suppose the budget of $200,000 is set with the reservation price at $200,000 and
fund availability at $250,000. Where the reservation price serves as the cap, raising
the fund availability to $220,000 still results in a budget of $200,000 because the
budget cannot go over the reservation price. The perceived loss is not reduced in this
case, and overspending is unlikely to occur.
Note that the role of fund availability in an overspending decision is similar
to that of reservation price in benefit gain from buying condition. Both have an
impact on the budget, and consumers are likely to choose the expensive alternative
because it falls within the new budget. As mentioned earlier, the present study still
32
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
defines this behavior as overspending by keeping the original budget as the
benchmark for overspending. We will discuss more about the change of budget and
its theoretical implications later in this chapter.
Fund Nonloss Front Not Buying People in prevention focus and
undesired end-state reference are motivated to avoid the presence of an undesired
end-state, according to self-regulation theory. Applying this notion to consumer
choice context, consumers are in fund focus and not-buying reference. The main
decision variable is fund nonloss, which is defined as the difference between the
prices of the expensive alternative that goes over the budget and the inexpensive one
that fits into the budget. By not buying the expensive alternative, a consumer saves
the price difference, which is conceptualized in this study as nonloss. Compared to
the fund loss defined above as the loss from overspending, fund nonloss is equivalent
to the amount saved by not overspending.
Therefore, what consumers are motivated to avoid is not being able to save
money on a purchase, where the potential saving is equal to the perceived price
difference. Perceived price difference among the alternatives is reported to have an
impact on consumer choice (e.g., Russo and Schoemaker 1989; Thaler 1980; Monroe
and Petroshius 1981), and the present study suggests that perceived saving should be
the key decision variable consumers consider in fund nonloss from not buying
condition. If the price difference (nonloss) is perceived as significant, overspending
is unlikely to occur and vice versa.
33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
For example, the perceived saving of $20,000 between the two alternatives
John has to choose from has direct impact on the decision when John is motivated to
minimize the fund nonloss from not buying. If the $20,000 saving is perceived as a
significant amount, John would not overspend. On the other hand, John is likely to
overspend when the nonloss is perceived as trivial (minimized).
An Integrated Conceptual Framework Four variables (reservation price,
anticipated regret, fund availability, and perceived saving) are identified as the key
decision variables for consumer choice under budget constraint. Figure 3
summarizes the proposed integrated conceptual model that explains four different
processes behind overspending with these four decision variables. By connecting a
decision variable to a self-regulatory state, a set of testable hypotheses can be
developed based on the past research findings on the key variables and in self
regulation theory. The effects of the key decision variables on various cognitive,
affective, and behavioral responses are hypothesized and tested in the next chapter
based on the proposed process model.
Graphical Representation o f Overspending Behavior Before moving on
to the proposed studies to empirically test the conceptual model presented above, a
graphical representation of overspending behavior is presented to further understand
the choice process under budget constraint. Figure 4 shows a diagram with the price
and benefit as its axes where four alternatives are shown on the same value
maximization line. Product C is preferred to A because C offers more benefits at the
same price. Product B is preferred to D because B has a lower price with the same
34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
amount of benefits. However, the decision becomes a challenge when choosing
from products A and B. They are on the same value maximization line, which means
both provide exactly the same bang for the buck. A is cheaper but offers fewer
benefits, and B has better benefits at a higher price. Any product in the shaded
triangular region on the diagram is preferred to both products A and B, and any
product in the white region is considered inferior. Since products A and B are on the
same value maximization line, consumers are not expected to prefer one to another.
Figure 3. A Conceptual Model for Consumer Choice Under Budget Constraint
Reservation
Price
Budget
Anticipated
Regret
Maximiz ng Minimizing
Benefit Gain Benefit Nongain
from Buy ing from Not Buying
I
Evaluation
1
Minimizi
Fund Lo
from Buy
Fund
Availability
< ig Maximizing
ss f^find Nonloss
ng fro)n Not Buying
Perceived
Saving
Choice for
Purchase
Budget Revision Processes
Budget Violation Processes
35
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4. Graphical Representation of Consumer Choice Under Budget Constraint
Benefit
B2 -
Equal
Value Maximization
Lin^
Benefit Gain y *
from Buying
Fund Loss
from Buying
B1
- > Fund Nonloss
from Not Buying
Benefit Nongain
from Not Buying
P1 P2
Price
One of the major theoretical contributions of the proposed conceptual model
is the introduction of budget revision and violation. As discussed earlier, consumers
in benefit gain from buying and fund loss from buying conditions are expected to
revise the budget during the choice process. It is defined as ‘budget revision’ in the
present study. On the other hand, consumers in benefit nongain from not buying and
fund nonloss from not buying conditions are expected to keep the original budget. If
they decide to overspend, it is defined as ‘budget violation’ in the present study.
Even though the same outcome is observed as the behavioral outcome in both budget
36
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
revision and violation cases, the processes are different in their nature, and as a
result, the implications are different. For example, budget revision and violation
would have different implications on repurchase behavior. Once a budget level is
raised during a purchase process, the next purchase would have a higher budget than
the original, while budget violation would not have as much impact on the next
budget formation.
Consumers who are motivated to maximize benefit gain from buying focus
on the benefits of product B, which costs more than the budget. They then decide if
the extra benefits of product B are worth paying more. The benefit perception
decides the choice outcome, and consumers overspend when the perceived amount
of benefits offered by product B moves upward into the preferred region.
Consumers who are motivated to minimize benefit nongain from not buying focus on
the benefits of product A. They decide if losing the extra benefits of product B is
important enough to make product A’s benefit offering inferior. Again, the benefit
perception decides the choice outcome, and consumers overspend when the
perceived amount of benefits offered by product A moves downward into the inferior
region. While benefit focus shapes the overspending decision by moving the benefit
perception vertically on the diagram, fund focus creates a horizontal shift of the price
perception. By coming up with ways of paying extra for the expensive product B,
minimizing fund loss from buying moves the product B’s price perception leftward
into the preferred region. On the other hand, consumers maximize the fund nonloss
from not buying by considering the price difference (P2 - PI) as trivial.
37
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Overspending takes place when the price perception of product A moves to the right,
an inferior region.
Figure 4 illustrates the reason for applying the regulatory focus and reference
to consumer choice under budget constraint. Regulatory focus determines what
direction (up/down or left/right) a consumer’s perception shifts, and regulatory
reference determines which alternative’s perception moves. Putting these two
dimensions on a plane, the diagram clearly shows four distinct processes behind
overspending.
38
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
III. E m p ir ic a l T e s t s o f T h e P r o p o s e d C o n c e p t u a l M o d e l
Two pilot studies and three main studies are designed to empirically test the
proposed conceptual model for consumer choice under budget constraint. Pilot
Study 1 investigates the budget formation process and provides the brief verification
of the assumed budget formation process. Pilot Study 2 surveys the motivations
behind overspending to confirm that consumers do go through one of the proposed
processes when choosing to overspend. Pilot Study 2 also serves as a pretest for
choosing the product categories for the stimuli of the main studies. Main Study 1
tests the relationships among the self-regulatory states, key decision variables, and
the outcomes. The outcomes include cognitive, affective, and behavioral responses.
Main Studies 2 and 3 examine the moderating effect of situational variables on
overspending behavior. The results offer the boundary conditions within which the
proposed model works.
1. Pilot Study 1: Budget Formation Process
The first pilot study asked 140 business major undergraduate students at a
west coast university to write a short essay on how they plan to purchase a stereo
boombox for their dorm rooms. The scenario was presented that their boombox is
broken, so they have to buy a new one to replace it. Two independent female judges
coded the responses, and the agreement rate was 93%. The author resolved the
disagreements after consulting with the judges.
39
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
132 responses (94.3%) mentioned that they were going to set a budget before
searching for information or going to a store. To decide the budget amount, 45
responses (32.1%) stated that extra features were important, 43 responses (30.7%)
mentioned that they were going to look into the functions, 39 responses (27.9%)
considered advertising as a viable source of information for their budget formation,
and 31 responses (22.1%) wrote that they would compare the prices of the
alternatives. Combining these three major responses and others that are related to
market information (e.g., ask friends, current boombox price, sale, quality, brand
name, warranty, and market average price) yielded 102 responses (72.9%) reporting
that they would search for market information (see Table 1 for the summary of
responses). The result implies that the majority of consumers do collect market
information to set a reservation price for a purchase. At the same time, 59 responses
(42.1%) mentioned affordability would play an important role in planning the
purchase, which shows that fund availability, as assumed earlier, would be a part of
the budget formation process. In addition, there were 43 responses (30.7%) that
mentioned both market information and affordability. Considering the fact that a
boombox (less than $50 in general) does not cost a substantial portion of a college
student’s monthly budget, the interpretation of the responses supports the definition
of budget used in the conceptual model proposed earlier; the subjects often consider
both market information and fund availability in their budget formation process.
4 0
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 1. Pilot Study 1: Response Summary
Response Items Frequency*
Affordability 42%
Extra features of the alternatives 32%
Functions of the alternatives 31%
Information from advertising 28%
Compare prices of alternatives 22%
Quality of the alternatives 21%
Price and features of the current boombox 19%
Value of the alternatives 18%
Ask friends 10%
Average market price 10%
My need for a boombox 10%
Other products than boombox 10%
Warranty coverage 8%
Personal experience 7%
On sale 6%
Brand name 4%
New technology 4%
How important boombox is 3%
My usage of boombox 3%
Top brand 3%
Design of the alternatives 1%
Wait if not affordable 1%
* Percentage out of 140 total responses.
2. Pilot Study 2; Overspending Behavior
The second pilot study asked 108 business major undergraduates at a West
Coast university to write an essay about two occasions they overspent and why they
did. This pilot study was designed to examine the external validity of the proposed
motivations discussed in the previous chapter. Since the motivations were
manipulated and their relationships with decision variables and behavioral outcomes
were tested in the main study later, Pilot Study 2 attempted to verify that these
motivations were indeed in effect when consumers made a choice under budget
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
constraint. This serves an important purpose of confirming the activation of the
motivations specified in the proposed model in real life choice contexts, and the
confirmation provides the external validity of the manipulations in the main study.
The responses were coded by independent judges, one male and one female, into
four categories of motivation (maximize benefit gain from buying, minimize benefit
nongain from not buying, minimize fund loss from buying, and maximize fund
nonloss from not buying) as proposed in the conceptual process model. The judges
agreed on the classification 92% of the time, and the author resolved the
disagreements after consulting with the judges.
The results showed that 138 out of 216 reported overspending experiences
(63.9%) stated one of the four self-regulatory motivations as the reason behind
overspending. The rest (78 responses) failed to explain why they overspent or did
not fit to any of the motivation types. For example, some responses mentioned a
clear preference or better price-performance ratio as the reason behind overspending,
and some others attributed the overspending experience to lack of information.
These cases are outside the scope of the present study, so they are not included in the
analysis. Excluding the nonresponse error, the result grants external validity of the
present study’s conceptual model suggesting that the proposed motivations are
indeed at work when consumers make purchase decisions under budget constraint.
Typical examples for each self-regulatory state are shown in Table 2.
4 2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 2. Excerpts From Pilot Study 2: Why Overspend?
Regulatory
Focus
Regulatory
Reference
Motivation Quote Example
Benefit
Buying
Maximize
benefit gain
from buying
“It had better features and also
looked newer and nice.”
Not Buying
Minimize
benefit nongain
from not buying
“The less expensive product
seems like its performance will
not suit my needs.”
Fund
Buying
Minimize
fund loss
from buying
“I couldn’t afford it, so I had to
ask mom and dad to help.”
Not Buying
Maximize
fund nonloss
from not buying
“$30 more did not seem to
much.”
Another important contribution of this pilot study is the assessment of the
chronic self-regulatory state people automatically activate when facing a choice
under budget constraint. It is reported in psychology as an individual personality
trait stating that, ‘although individuals are likely to develop the capacity for using
both strategies, only one strategy will become their habitual mode of responding to
the world’ (Rohan and Zanna 1998, p280) regarding promotion versus prevention
focus. Due to this chronic tendency of individuals to activate a certain regulatory
focus, studies in social psychology often measure regulatory focus instead of
manipulating it (e.g., Brendl, Higgins, and Lemm 1995; Higgins, Shah, and
Friedman 1997). Figure 5 shows the probability of activating each chronic
regu latory fo c u s and referen ce w ith ou t any m anip u lation for the 138 resp o n ses that
were categorized into a self-regulatory state in the pilot study.
43
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 5. Summary of Pilot Study 2: Chronic Self-regulatory Focus and References
70%
£ 60%
c
o
50%
Buying \
Reference
O o
L-— ■ "
40%
30%
e
© <
3
CT
20%
Not Buying '
Reference 10%
0%
Benefit Fund
Regulatory Focus
The probabilities in Figure 5 show that benefit focus with buying reference is
the most common self-regulatory state people activate without any outside influences
(60.1%), and it is followed by benefit focus with not-buying reference (21.0%), fund
focus with not-buying reference (10.9%), and fund focus with buying reference
(8.0%). It is important to identify the probability of a particular chronic self-
regulatory state because of its potential interaction with the manipulated state in an
experimental setting. Although it is reported that momentary situational
manipulations in laboratory experiments have similar effects to that of chronic self-
regulatory states developed by socialization (Higgins 1998), it has to be noted that
the effect of a self-regulatory state on cognitive, affective, and behavioral outcomes
is a product of both chronic and manipulated self-regulatory states. Therefore, the
results from the empirical tests in the present study should be analyzed considering
the distribution of the chronic tendency found in the pilot study.
44
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Pilot Study 2 also serves as a pretest for the product categories to be used in
the instruments for the empirical testing. Table 3 lists the product categories the
respondents reported as the ones they overspent on. 197 responses are compiled
from 108 respondents, each respondent reporting one or two product categories he or
she overspent on. Thirty product categories are identified from the responses, and
the frequency of each category is reported in the table. Considering the fact that the
respondents are college students, it is not surprising that stereo system, car, and
computer are ranked as the three most overspent categories.
The list is valuable because the highly overspent categories would not be
ideal candidates for the instruments in the main study. It is possible that the
manipulations in the experiments would not be able to show their effects on the
overspend decision due to a ceiling effect, which the overspending probability has
already reached without any manipulations. In other words, an ideal product
category should have enough room for the manipulations to influence the outcome so
that the effect can be observed. At the same time, a product category with very little
overspending frequency should be avoided as well. It is a case of a floor effect,
which reflects the unwillingness to overspend on the category. In short, product
categories somewhere in the middle would be the best for detecting the effects of
manipulations. The selection of the product categories for the main studies is further
discussed in the following sections.
45
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 3. Pilot Study 2: Summary of Overspent Product Categories Reported
Product Category Frequency %
Stereo 36 18.3%
Car 29 14.7%
Computer 28 14.2%
Clothing 18 9.1%
TV 15 7.6%
Bike 14 7.1%
Phone/Celluar/Pager 9 4.6%
Shoe 7 3.6%
Camera 5 2.5%
Gift 5 2.5%
VCR 4 2.0%
Vacation 3 1.5%
CD 3 1.5%
Entertainment 3 1.5%
Watch 2 1.0%
Apt 2 1.0%
Golie Pads 1 0.5%
Surf Board 1 0.5%
Salon 1 0.5%
Treadmill 1 0.5%
Stove 1 0.5%
Dog 1 0.5%
Flower 1 0.5%
Concert Tickets 1 0.5%
Make-up 1 0.5%
Dss 1 0.5%
Guitar Amp 1 0.5%
Fax 1 0.5%
Chair 1 0.5%
Carpet 1 0.5%
Total 197 100.0%
3. Study 1: Consumer Choice Under Budget Constraint
Study 1 attempted to confirm the cognitive and affective processes behind
overspending conceptualized in the model presented in Figure 3. Pilot Study 2 has
already shown that consumers do spend over the budget in their purchases being
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
motivated by one of the self-regulatory states proposed in the present study. While
Pilot Study 2 was based on essay responses, Study 1 examined the proposed model
in a laboratory experiment with tight control of the possible confounds. The main
objective of Study 1 was to confirm the proposed model and identify the differences
among the four self-regulatory states in terms of their effects on the behavioral
outcome. Validating the proposed model included the identification of the main
decision variable in each self-regulatory state and confirming its role in overspending
behavior.
As mentioned earlier, the scope of the present research is limited to the
overspending behavior, not the budget formation process. Therefore, the budget was
given to the subjects in Study 1 and they were asked to make a purchase decision
with the budget in mind. A binary choice was presented to the subjects to make the
task as simple as possible. The simplicity of the task limited the amount of
information for the subjects to process and precluded possible order effect. The
simplicity also helped maintain high accessibility of the product information in the
subjects’ memory when making the choice.
A forced choice task was given to the subjects to observe a clean effect. Of
course, choice deferral is a viable option for consumers when they face a conflict
between acquiring a better product and staying within the budget. However, giving
the subjects an option to defer the choice provides no observable behavioral
outcome, which blocks the opportunity to understand the process behind the choice
decision. The same choice deferral can come from multiple processes, which would
47
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
be difficult to identify. Even though we can observe the delayed choice later, the
delayed choice would also be influenced by various factors outside the scope of the
present study. Therefore, the subjects performed a forced choice task, which
provided a clean and efficient empirical test of the proposed model without the
possible confounds.
Independent Variables and Manipulations A 2x2 factorial design with a
control group is operationalized in Study 1. Two dimensions for the factorial design
are regulatory focus and regulatory reference. Regulatory focus has two levels:
benefit and fund foci. Regulatory reference has two levels: buying and not-buying
references. The subjects in the benefit and fund foci were asked to think about ‘the
features and benefits’ and ‘the prices’ of the alternatives, respectively. The subjects
in the buying and not-buying references were asked to think about buying the
expensive (overspending) and the less expensive (staying within the budget)
alternative, respectively. The instructions for the objective of the choice task, which
served as the key manipulation, read, “you want to buy one that comes with great
features” (benefit gain from buying), “you don’t want to buy one that lacks great
features” {benefit nongain from not buying), “you want to buy one that is not going
to hurt your finance” (fund loss from buying), and “you don’t want to buy one that is
going to hurt your finance” (fund nonloss from not buying).
The manipulations were imposed in two stages of the task: one before
evaluating the alternatives and another before making a choice after the evaluation.
The reason for having two manipulations was to reinforce the self-regulatory state
48
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
the subjects were manipulated to activate and to avoid the confounding between the
chronic and manipulated self-regulatory states. As discussed earlier, people have a
chronic tendency to activate a certain self-regulatory state, which is a personality
trait. If the chronic tendency is at work in the experiment, the observed outcomes
cannot be attributed to the manipulations. In other words, the chronic and
manipulated self-regulatory states fight for dominance in the decision making
process, and the stronger the manipulation, the less the effect of chronic tendency on
the outcomes is. A pretest where the subjects were asked to think aloud while
performing the task confirmed that the manipulations were strong enough to override
the chronic tendency.
A control condition was added to serve as the baseline to measure the
effectiveness of the manipulations on overspending behavior. No manipulations
were given to the subjects, and the chronic self-regulatory states of the subjects were
measured instead of being manipulated. Of course, overspending could still be
observed in the control group, but the probability of overspending was expected to
be lower due to the lack of manipulation that reinforces the self-regulatory state and
motivation.
Task Each subject was randomly assigned to one of the conditions
described above and asked to complete a choice task. A computer based shopping
task was designed with digital camera. Camera is a product category that was not
mentioned frequently in Pilot Study 2 (2.5%) but was ranked 9th on the most
overspent category list. After a series of pretests using personal interviews, digital
49
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
cameras were found to be high in involvement and low in product category
familiarity among the subject pool—college students. High involvement reduces the
response error because experimental condition manipulation tends to work better
under high involvement, and low familiarity reduces the noise from the variance of
product category familiarity. Therefore, digital cameras were used as the stimulus to
achieve high involvement and better control over the product category familiarity.
The subjects were told that they had come up with a choice set of two
alternatives as the finalists after evaluating the available alternatives in the market.
Two real digital camera models with Olympus brand were used as the finalists. A
series of pretests confirmed that both models were perceived to be on the same value
maximization line (i.e., equal price-performance ratio), which minimized the
possibility of preferring one to another based on the features, benefits, and price. In
other words, the presented models were perceived to offer exactly what the money
was worth.
The task started with the general instructions for the task, and subsequently
the subjects were presented with a scenario they were asked to imagine themselves
in. The scenario included the objective of the task and the budget they had to keep in
mind. The budget of $300 was given with a short description of how it was formed
based on the assumed budget formation process presented in the previous chapter.
Then the subjects were presented with a summary of a Consumer Report article that
contained the first manipulation of self-regulatory state (focus and reference). After
processing the article content, two models with extensive attribute information were
50
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
presented as the finalists. One was priced at $295 and the other at $395. The
subjects were asked to evaluate them and think about which one to buy. The next
screen showed a paragraph with a title, ‘now let’s think about this.. as a second
manipulation reiterating the first manipulation (Consumer Report article) in a more
specific context. The subjects were constantly reminded of the budget during the
manipulations and evaluation stages. Making the budget salient throughout the
whole task was expected to reduce the probability of overspending, which warrants a
conservative test of the effect by the manipulation on overspending. In other words,
if overspending occurred even under salient budget constraint, it could be concluded
that the manipulation did override the impact of budget constraint and induce
overspending. The screen where the subjects were asked to make a choice came
next, and it was followed by an open-ended question asking the subjects to explain
the choice process. After the computer shopping simulation was done, a paper and
pencil questionnaire with cognitive and affective dependent measures was given to
the subjects and filled out without allowing the subjects to go back to the computer
shopping simulation.
Dependent Variables Various behavioral, cognitive, and affective
variables are observed or measured as dependent variables. The choice made by the
subjects between the alternatives is the key behavioral outcome. Choosing between
one that goes over the budget and the other that does not has a binomial distribution,
which can be translated into overspending probability. The amount of time spent to
make a choice is another behavioral measure that is recorded. There are also a
51
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
number of process measures that are critical to test the proposed conceptual model
for choice under budget constraint. These include the key decision variables such as
reservation price, anticipated regret, fund availability, and perceived saving. In
addition to the process measures, cognitive outcomes such as the revised reservation
price, fund availability, and budget after the choice are recorded along with the
perceived difficulty of the task. Different types of emotions are also measured after
the choice to test the effect of the independent variables on affective responses. The
hypotheses related to these dependent variables are formulated later.
Covariates Two individual factors are measured to sort out the effects of
personality traits in the analysis: buying impulsiveness and need for cognition.
Buying impulsiveness has a direct impact on overspending behavior. It is defined as
‘a consumer’s tendency to buy spontaneously, unreflectively, immediately, and
kinetically’ (Rook and Fisher 1995, p306), and therefore, the subjects with high
buying impulsiveness are more likely to overspend in the present study. Since the
present study is interested in the effect of manipulated self-regulatory states on
overspending, not of the inherent impulse buying tendency, a subset of the buying
impulsiveness scale composed by Rook and Fisher (1995) is included in the
questionnaire and input to the analysis as a covariate to separate its effect on
overspending.
Need for cognition (NFC) is also found to have an impact on consumer
choice when they have to resolve conflicting information about attitude objects
(Thompson, Zanna, and Griffin 1995). The choice context in the present study falls
52
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
into this category where consumers have to resolve a conflict between acquiring a
desirable alternative and sticking to the budget. It is expected that the individuals
with high NFC are likely to spend more time to resolve the conflict and enjoy the
cognitive investment. Since NFC is a personality trait independent of the
manipulated self-regulatory states, it is measured and thrown into the analysis as a
covariate. A subset of the NFC scale composed by Cacioppo, Petty, and Kao (1984)
is adopted in the questionnaire.
Hypotheses With the experimental design described above, the present
study formulates a set of hypotheses to test the conceptual model for choice under
budget constraint presented in the previous chapter. The proposed conceptual model
posits that consumers activate four motivational states: 1) maximizing benefit gain
from buying, 2) minimizing benefit nongain from not buying, 3) minimizing fund
loss from buying, and 4) maximizing fund nonloss from not buying. The first set of
hypotheses targets the behavioral outcome of the choice task given to the subjects.
Four experimental conditions with different motivational states and one control
condition are compared in terms of their impact on overspending probability, and the
effectiveness of the manipulations is tested. Observing higher probabilities for
overspending in the four experimental conditions than in the control group confirms
that the motivational states that are identified in the proposed conceptual model
indeed lead to overspending behavior. The formal hypotheses are stated below.
53
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Hja: Consumers who are presented with a self-regulatory state o f benefit
focus and buying reference are more likely to overspend than the ones
without the presentation.
H ib'. Consumers who are presented with a self-regulatory state o f benefit
focus and not-buying reference are more likely to overspend than the
ones without the presentation.
Hic: Consumers who are presented with a self-regulatory state offund focus
and buying reference are more likely to overspend than the ones
without the presentation.
H iy Consumers who are presented with a self-regulatory state offund focus
and not-buying reference are more likely to overspend than the ones
without the presentation.
The second set of hypotheses is formulated around the differential effect of
motivational state on overspending probability. Is there any difference among the
motivational states in encouraging overspending? To make the predictions, it is
necessary to revisit the results of Pilot Study 2 where the chronic self-regulatory
states are recorded (see Table 4 for the summary of results). Benefit focus showed
higher activation frequency than fund focus (81.2% vs. 18.8%) as did buying
reference than not-buying reference in benefit focus condition (60.1% vs. 21.0%).
The results lead to a conclusion that consumers are more likely to activate the
motivation to maximize the benefit gain from buying than in the other motivational
54
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
states when facing a decision to overspend or not. This chronic tendency is a
personality trait, which gets activated regardless of manipulation.
Table 4. Chronic Frequency of Activating A Self-regulatory State
(Pilot Study 2)
Regulatory Reference
Regulatory Focus
Buying Not Buying Total
Benefit 83 (60.1%) 29 (21.0%) 112(81.2%)
Fund 11 (8.0%) 15 (10.9%) 26(18.8%)
Total 94 (68.1%) 44 (31.9%) 138 (100%)
When a consumer is exposed to a marketing message that manipulates the
consumer to temporarily activate another motivational state, there are two possible
interactions that can take place. First, consumers would still rely on the chronic
activation and make a choice. The motivational state is considered exclusive in the
self-regulation theory literature in social psychology (e.g., Fliggins 1998; Crowe and
Fliggins 1997). In other words, only one regulatory focus and reference can be a
dominant motivational force that regulates the behavior. This belief lets social
psychology literature use a median split of the measured self-regulatory states and
use them as independent variables in the analysis. This ‘exclusivity’ argument leads
to a hypothesis predicting that the activation of a motivational state congruent to the
55
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
chronic tendency (i.e., maximizing the benefit gain from buying as found in Pilot
Study 2) should result in the strongest impact on the likelihood of overspending as
suggested by Higgins (1997). For example, a consumer with a chronic tendency of
maximizing the benefit gain from buying would find it confusing when presented
with a marketing message encouraging the consumer to minimize the benefit nonloss
from not buying. The manipulation would force the consumer to rely either on his
chronic tendency or on the imposed manipulation. The overspending likelihood
should be somewhere between that of the two self-regulatory states, and it cannot
exceed the case of benefit gain from buying, which is found to be the highest in
inducing overspending. The formal hypothesis is stated below.
H2a ■ ' Consumers who are presented with a self-regulatory state o f benefit
focus and buying reference are more likely to overspend than the ones
in the other regulatory foci and references.
Another possibility is that more than one motivational state can be activated
simultaneously and interact with each other. The motivational states are considered
additive in this case where activating more than one regulatory focus and reference
are involved in making a choice. The additive nature of different regulatory foci and
references are assumed to be complementary, which leads to a higher chance of
overspending in consumer choice. For example, a consumer who has a chronic
tendency of activating benefit gain from buying and is exposed to a marketing
56
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
message designed around benefit nongain from not buying, the consumer now has
two tools to attack the conflict in choice. In other words, wanting to acquire the
desirable benefits and perceiving the price difference as trivial at the same time
should be more effective in inducing overspending than by either one of the
motivations alone. In other words, given that benefit gain from buying is a chronic
self-regulatory state, manipulating consumers to activate the other states would
increase the chance for overspending. A set of competing hypotheses to H2 a is
formulated below, and the empirical test will show if the self-regulatory states are
exclusive or additive in influencing consumer choice.
H2 b: Consumers who are presented with a self-regulatory state offund focus
are more likely to overspend than the ones in benefit focus.
H2c: Consumers who are presented with a self-regulatory state o f not-buying
reference are more likely to overspend than the ones in buying
reference.
Hu: Consumers who are presented with a self-regulatory state offund focus
and not-buying reference are more likely to overspend than the ones in
the other regulatory foci and references.
If H2 a is supported, it supports the argument that self-regulatory states are
exclusive (i.e., one self-regulatory state dominates the process), and if H2 b - H u are
supported, self-regulatory states are considered additive (i.e., multiple self-regulatory
57
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
states can be activated in a choice task simultaneously). More specifically, H2b
hypothesizes that regulatory focus is exclusive. Suppose consumers can put
themselves in benefit and fund foci simultaneously, manipulating them into fund
focus would strengthen the motivation to overspend. Therefore, the overspending
probability in the fund focus condition is expected to be higher than in benefit focus.
H2 c hypothesizes the additive nature of regulatory reference. Since buying reference
is chronically dominant, manipulation of not-buying reference would increase the
probability of overspending if regulatory reference is additive. Finally, H2 d
hypothesizes that both regulatory focus and reference are additive in nature. If this is
the case, consumers would show the highest overspending probability when they are
manipulated into the state that is not chronically dominant in both dimensions.
This set of hypotheses provides an important insight into self-regulation in
consumer choice. Theoretically, the nature of self-regulation in a complex task such
as consumer purchase needs to be understood before it is applied to marketing. It
also has strong managerial implication to marketers. If self-regulatory states are
exclusive, marketers would have little incentive to manipulate self-regulatory states
in their marketing communication programs because the manipulation would be
effective only when it is congruent with the chronic state. If not, it can be an
important tool to effectively persuade consumers to choose a particular alternative.
The next set of hypotheses test the process measures for the proposed
conceptual model for consumer choice under budget constraint. As explained in the
previous chapter, four decision variables are suggested to play a major role in the
58
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
four self-regulatory states: reservation price for benefit gain from buying, anticipated
regret in benefit nongain from not buying, fund availability in fund loss from buying,
and perceived saving in fund nonloss from not buying. The importance ratings for
these decision variables are compared among the experimental conditions to confirm
the proposed model. The formal hypotheses are stated below.
H3a: Consumers with benefit focus and buying reference rate reservation
price more important for making a choice than the other decision
variables.
H ib'. Consumers with benefit focus and not-buying reference rate anticipated
regret more important for making a choice than the other decision
variables.
Hsc: Consumers with fund focus and buying reference rate fund availability
more important for making a choice than the other decision variables.
Hip. Consumers with fundfocus and not-buying reference rate perceived
saving more important for making a choice than the other decision
variables.
The previous chapter discussed the difference between budget revision and
violation processes in the proposed conceptual model for consumer choice under
budget constraint. Even though the observed results are the same, it is suggested that
the process behind them is different in terms of how the budget is maintained. The
59
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
proposed conceptual model posits that consumers with buying reference would
revise either reservation price (benefit focus) or fund availability (fund focus), and
subsequently revise the budget. This budget revision process should have an impact
on the budget formation process of the next purchase occasion because the current
purchase shifts the original budget. Once the reservation price, fund availability, and
the budget itself are changed, it is believed that it is difficult for consumers to go
back to the level where the budget was before the revision. On the other hand, the
proposed conceptual model suggests that consumers with a not-buying reference do
not revise the budget. Instead, the decision variables—anticipated regret (benefit
focus) and perceived saving (fund focus)—directly influence the choice outcome.
This is called budget violation because the original budget is not affected during the
choice process and stays the same after the choice. The formal hypotheses for
testing budget revision and violation are stated below.
H4a: Consumers with buying reference are more likely to raise the budget
after the choice than the ones with not-buying reference.
Iff,: Consumers with not-buying reference are more likely to maintain the
budget after the choice than the ones with buying reference.
Another behavioral outcome of interest in the model is the amount of time
consumers spend to arrive at the final choice. The past research on self-regulation
theory in social psychology reports that people in promotion focus are likely to spend
60
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
more time to accomplish the goal of a task (e.g., Roney, Higgins, and Shah 1995).
This translates to a prediction in the present study’s context where consumers in
benefit focus are expected to spend more time on making a choice. Fortunately, the
present study collected an extensive set of data on the time spent on each stage of the
task because a computer shopping simulation is used as the instrument. There are six
time measures recorded for each subject: 1) time spent to read the instructions, 2)
time spent to read the first manipulation (Consumer Report article), 3) time spent to
evaluate the alternatives, 4) time spent to read the second manipulation, 5) time spent
to make a choice, and 6) the total time taken to complete the task. Among these
measures, the present study is interested in the time spent to evaluate and make a
choice, and it would be a great opportunity to identify on what stage of the purchase
process the self-regulatory state has the most differential impact. Inferring from the
findings in social psychology, the following hypotheses will be tested.
Hja: Consumers with benefit focus are likely to spend more time to evaluate
the alternatives than the ones with fund focus.
H sb'- Consumers with benefit focus are likely to spend more time to make a
final choice than the ones with fund focus.
The last set of hypotheses includes the affective responses by the consumers
after making a choice. It has been suggested in self-regulation literature that the
successful (failing) approach toward a positive outcome activated dejection-related
61
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
emotions such as being happy and satisfied (sad and dissatisfied), while the
successful (failing) avoidance of a negative outcome activates agitation-related
emotions such as being calm and relieved (tense and nervous) (Carver and White
1994; Higgins, Shah, and Friedman 1997; Roney, Higgins, and Shah 1995). It would
be interesting to test if these findings can be applied to a consumer choice context.
Six scales were included in the questionnaire to measure the affective responses:
three dejection-related and three agitation-related. A composite measure will be
used to test the hypotheses after a confirmatory factor analysis. The formal
hypotheses are listed below.
H (,a: Consumers with benefit focus are more likely to experience dejection-
related emotions than the ones with fund focus.
II< ; / , ■ ' Consumers with fundfocus are more likely to experience agitation-
related emotions than the ones with benefit focus.
In conclusion, Study 1 attempts to support the proposed conceptual model for
consumer choice under budget constraint by testing the hypotheses that are derived
from the proposed model. The hypotheses deal with the cognitive, affective, and
behavioral responses along with the processes behind overspending, as does the
proposed model. The results of Study 1 will be presented in the next chapter.
62
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
4. Study 2: Interaction Between Self-regulatory State and Absolute Budget
Level
Building on the proposed conceptual model tested in Study 1, Study 2
examines if self-regulatory states have a different impact on consumer choice
depending on the absolute budget level. The term ‘absolute budget level’ is defined
as whether consumers perceive the budget as a substantial portion of their overall
expenses or not. For example, a home purchase is an extremely high budget task
with a considerable amount of expense, while a detergent purchase is a low budget
task involving a small amount of expense. Do these purchases differ in terms of
consumer choice behavior? Do they lead to different overspending decisions?
These questions are investigated in an on-line experimental setting.
Experimental Design In addition to the same set of independent variables
(regulatory focus and reference), absolute budget level is added as another dimension
of the 2x2x2 factorial design of the experiment. Regulatory focus has two levels—
benefit and fund focus. Regulatory reference has two—buying and not-buying
reference. Absolute budget level has two—high and low. The manipulations are
imposed the same way as in Study 1 for regulatory focus and reference using a
Consumer Report article. Two product subcategories are used to manipulate the
absolute budget level, one with high budget and prices and another with low budget
and prices.
The subjects were asked to participate in an online shopping task for a
computer monitor. A computer monitor was chosen because of its high frequency of
63
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
overspending reported in Pilot Study 2 and because it had two subcategories for
which the price range is considerably different. A computer is ranked as the third
most frequently overspent category in Pilot Study 2 (14.2%). In other words, a
computer is a product category the subject pool overspends on very often. By using
a component (monitor) instead of the whole system as the stimulus, the ceiling effect
on the overspending likelihood can be avoided. A series of pretests using interviews
confirmed that a computer monitor is a category subjects would not overspend on.
Another merit of using a computer monitor as the stimulus is that the
stimulus can easily manipulate two absolute budget levels without significantly
changing the content of the stimulus. Study 2 used a 17-inch regular cathode ray
tube (CRT) monitor for the low absolute budget level condition and 15-inch liquid
crystal display (LCD) flat panel monitor for the high condition. Both subcategories
shared the same attributes such as screen size, resolution, refresh rate, compatibility,
and color balance system, so the information presented to the subjects could be
controlled across conditions. NEC was used as the brand in both conditions because
it was found to be one of the most respected monitor manufacturers of both CRT and
LCD monitors from a pretest. The prices of the CRT monitor alternatives were $150
and $200 with a $150 budget, and the prices of the LCD monitor alternatives were
$599 and $799 with a $600 budget.
The amount the subjects had to overspend to choose the expensive one in
both conditions was controlled to be 33% of the baseline model ($50 over $150 and
$200 over $600 for low and high budget level conditions, respectively). The overall
64
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
structure of the task is exactly the same as in Study 1, the only difference being the
different product category (please see Appendix C for the sample screenshots for the
online experiment). While the subjects switched over to a paper and pencil
questionnaire in Study 1, the responses were collected online for Study 2. After
making a purchase decision, the subjects were asked to write an open-ended response
that explained their choice, and the online questionnaire followed.
The same set of cognitive, affective, and behavioral measures were recorded
and analyzed as dependent variables of the study with the same set of covariates
(buying impulsiveness and need for cognition).
Hypotheses The conceptual model proposed in the present research posits
that consumers make a decision with a motivation to acquire desirable features and
benefits with the least amount of expense when a budget constraint exists. Acquiring
desirable features and benefits can be categorized as a positive motivation to
approach a positive outcome, and spending as little as possible as a negative
motivation to avoid a negative outcome. The conflict between these two opposite
motivational forces is found to create indifference or ambivalence (Cacioppo and
Berntson 1994; Cacioppo, Gardner, and Berntson 1997). More specifically, people
feel indifferent when there are weak positive and negative motivational forces at
work and ambivalent and unstable under strong positive and negative motivational
forces.
The present study suggests that the absolute budget level determines the
strength of the motivational forces and leads to indifference (low absolute budget
65
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
level) or ambivalence (high absolute budget level). When a consumer feels
indifferent among alternatives, the role of the self-regulatory state in the choice
process would be minimized. In other words, the choice is unlikely to be an
elaborate process, and it is doubtful that the differential effects of self-regulatory foci
and references would be significant. On the other hand, when a consumer
experiences ambivalence, the role of the self-regulatory state is expected to be
critical in the choice process. The choice task becomes elaborate, and the outcome is
likely to be influenced by the self-regulatory state of the consumer.
Therefore, it is predicted that the overspending probabilities for self-
regulatory states are significantly different from each other only in high absolute
budget level condition and not in low absolute budget level condition. The present
study formally hypothesizes the effects of absolute budget level on choice behavior
as following.
H7a: Consumers are likely to show significant differences in overspending
probability among different self-regulatory states when the absolute
budget level is high.
Hjb'. Consumers are not likely to show significant differences in overspending
probability among different self-regulatory states when the absolute
budget level is low.
66
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
5. Study 3: Interaction Between Self-regulatory State and Core Attribute
Type
Study 3 investigates the effect of the nature of the core attribute, the main
decision variable for the choice, on consumer choice under budget constraint. The
nature of the core attribute is classified into two categories in the present study:
functional and hedonic. For example, a consumer who is in the market for a new
home would put the square footage of the house or the design as the most important
decision variable. The square footage is a typical functional attribute of the product
category of interest, and the design is a hedonic attribute. Past research found that
consumers show different purchase behavior depending on the nature of the core
attribute. Study 3 examines its influence on choice behavior under budget constraint.
Experimental Design In addition to the same set of independent variables
(regulatory focus and reference), the nature of the core attribute was added as
another dimension of the 2x2x2 factorial design of the experiment. Regulatory focus
has two levels—benefit and fund focus. Regulatory reference has two—buying and
not-buying reference. Core attribute nature has two—functional and hedonic. The
manipulations were imposed the same way as in Study 1 for regulatory focus and
reference using a Consumer Report article. Television was used in the experiment as
a target product category, which ranked 5th on the most frequently overspent product
category list with 7.6% in pilot study 2. Two models of Panasonic televisions were
presented for subjects to choose from in the experiment. A budget of $300 was
given, and the alternatives were priced at $299 and $399.
67
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
The manipulation of the core attribute nature was performed by emphasizing
either functional attributes or hedonic attributes in the stimulus as the most important
decision criteria. For example, the functional attribute condition presented the
subjects with the features such as new tube technology, picture-in-picture, and
audio/video inputs/outputs for connecting to other components of a home theater
system. The hedonic attribute condition highlighted the viewing pleasure for the
eyes and extremely realistic sound to the ears with the same features presented in the
functional conditions.
The overall structure of the task was exactly the same as in Study 1, the only
difference being the different nature of the core attribute. The responses were
collected online as in Study 2. After making a purchase decision, the subjects were
asked to write an open-ended response that explains their choice, and the online
questionnaire followed.
The same set of cognitive, affective, and behavioral measures were recorded
and analyzed as dependent variables as in Studies 1 and 2 with the same set of
covariates (buying impulsiveness and need for cognition).
Hypotheses Past research found that consumers are likely to maximize the
features and benefits when they are more emotion-laden, meaning that consumers are
likely to overspend when the nature of the core attribute is hedonic. For example,
naive subjects are found to consider losses more salient than gains on hedonic
attribute under emotion-laden choice (Luce, Payne, and Bettman 1999). More
specifically, some hedonic attributes are said to be ‘sacred’ (Tetlock, Peterson, and
68
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Lerner 1996) or ‘protected’ (Baron 1986; Baron and Spranca 1997), so they make it
difficult for consumers to give them up. Emotion-laden choices with hedonic core
attributes are also found to be more difficult to give up in exchange for monetary
savings (Luce, Payne and Bettman 1999). Based on these findings, overspending
is more likely to occur when the core attribute is hedonic in nature. More
specifically, the losses are more heavily weighted when the core attribute is hedonic,
and therefore, the loss/nonloss conditions (fund focus) are expected to show higher
probabilities of overspending than gain/nongain conditions (benefit focus) when the
core attribute is hedonic. In addition, the perceived saving in payment nonloss for
the not-buying condition becomes relatively unimportant with the hedonic core
attribute as suggested by Luce et al. (1999). This leads to a high overspending
probability for the condition. The formal hypotheses are stated below.
Hsa: Consumers are more likely to overspend when the core attribute o f the
product is more hedonic than functional.
H8 b: Consumers with fund focus are more likely to overspend on a product
with a hedonic core attribute than the one with benefit focus.
H 8 c ■ ' Consumers with fund focus and not-buying reference are more likely to
overspend on a product with a hedonic core attribute than the one with
other regulatory foci and references.
The next chapter presents the results of Studies 1, 2, and 3 described above.
69
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
I V . R e s u l t s
The results from the three studies designed to empirically test the proposed
conceptual model are summarized in this chapter. Study 1 focuses on testing the
effects of self-regulatory states on overspending behavior and the processes behind
it. Study 2 examines the role of absolute budget level in overspending behavior and
its interaction with self-regulatory states. Study 3 tests the effect of the nature of a
product’s core attribute on overspending behavior and its interaction with self-
regulatory states.
1. Study 1: Consumer Choice Under Budget Constraint
190 undergraduate students taking marketing courses from a West Coast
university and a Southeast university participated in the experiment. Four cells in
the 2 (benefit vs. fund focus) x 2 (buying vs. not-buying reference) experimental
design collected 38 to 41 responses each, and 31 subjects were assigned to a control
group without receiving any manipulation.
Manipulation Checks The $300 budget given to the subjects for a digital
camera purchase was tested to determine if it was indeed a major influence on the
subjects in the choice task. The questionnaire included two questions regarding the
role of budget in the choice task. The first question asked the subjects if the budget
of $300, ‘was a major concern when I started to evaluate the alternatives,’ and the
overall average score was 5.25 in 1 (not at all) to 7 (very much) scale including the
70
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
control group. The ANOVA without the control group showed a marginally
significant difference between buying and not-buying references for this scale
indicating that the subjects in buying reference were more likely to consider budget a
major concern than the ones in not-buying reference (Maying = 5.48 vs. Mn o t-b u y in g =
4.94, F(l, 153) = 3.32, p < 0.10, s2 = 0.021). As predicted, buying impulsiveness as
a covariate in the analysis turned out to be marginally related to the influence of
budget (F(l, 153) = 2.82, p < 0.10, s2 = 0.018). Need for cognition was also
marginally significant (F(l, 153) = 3.65, p < 0.10, s2 = 0.023).
The second question asked the subjects if the budget of $300, ‘was a major
concern when I made my purchase decision,’ and the overall average score including
the control group was 4.83. The ANOVA without the control group showed
significant differences in both regulatory focus and reference dimensions. The
subjects in fund focus was more influenced by the budget than in benefit focus
(M benefit = 4.20 vs. Mfu n d = 5.22, F(l, 153) = 11.88, p < 0.005, s2 = 0.072), and the
ones in buying reference were more influenced by the budget than in not-buying
reference (Mbu y in g = 5.08 vs. Mn o t-b u y in g = 4.32, F(l,153) = 6.11, p < 0.05, s2 = 0.038).
Regarding the covariates, need for cognition was not found significant, but buying
impulsiveness was found significant in affecting the importance of budget in the
choice process (F(l, 153) = 5.41, p < 0.05, s2 = 0.034).
The combined measure of the above two measures for the importance of the
budget showed Cronbach’s alpha of 0.80 and the overall mean was 5.04 including
the control group. Both regulatory focus and reference had a main effect when
71
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
analyzed without the control group. The subjects in fund focus were more sensitive
to the budget than in benefit focus (Mbe n e fit = 4.64 vs. Mfu n d = 5.28, F(l, 153) = 6.44,
p < 0.05, s = 0.040), and the ones in buying reference were more sensitive to the
budget than in not-buying reference (Maying = 5.28 vs. Mn o t-b u y m g = 4.63, F(l, 153) =
5.76, p < 0.05, s = 0.036). The effect of buying impulsiveness on the importance of
budget was also significant (F(l, 153) = 5.02, p < 0.05, £2 = 0.032), and need for
cognition was found only marginally significant (F(l, 153) = 3.25, p < 0.10, £2 =
0.021).
These findings indicate that consumers are more sensitive to the budget when
the budget gets in the way and causes a conflict. In other words, consumers perceive
the task as a conflict when they are in fund focus and buying reference. As stated
earlier, fund focus activates the motivation to either minimize the fund loss from
buying or maximize the fund nonloss from not buying. Overspending works against
these motivations and causes a conflict in choice process. Similarly, budget
constraint enhances the conflict for the consumers with buying reference because
they consider spending over the budget.
In sum, the budget given to the subjects indeed influenced the decision in
both evaluation and choice stages of the task. The reported magnitude of the
influence varied among self-regulatory states in the choice stage, and a personality
trait—buying impulsiveness—was found to have an impact on the perception of
budget in the task. This confirms the need for including the covariates in later
analyses.
72
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
The second part of the manipulation check deals with the manipulations
themselves. It was critical to check if the manipulations for the self-regulatory state
worked in the experiment because of thier conflict with the chronic states of the
subjects. To test the proposed conceptual model, the manipulated self-regulatory
state should override the chronic state so that the results could be interpreted as the
effects by the manipulations.
Responding to the question that asked if, ‘I was mainly thinking about the
features and benefits of the alternatives,’ the subjects showed a significant difference
between benefit and fund focus conditions (M benefit = 5.30 vs. Mfu n d = 4.76, F(l, 157)
= 4.47, p < 0.05, b = 0.029), confirming that the benefit focus manipulation worked.
The control group, who did not receive any regulatory focus manipulation, showed
an average of 4.71, which is close to Mfu n d where no benefit focus manipulation was
given as well. The comparison to the control group further confirms that the
reported difference is due to the benefit focus manipulation.
Regarding the fund focus manipulation, the subjects were asked if, ‘I was
mainly thinking if I was able to afford the prices of the alternatives.’ Again, the
manipulation was confirmed to be successful (Mbenefit= 4.26 vs. Mfu n d = 5.20, F(l,
157) = 14.46, p < 0.001, s2 = 0.089). Despite the significant difference between two
regulatory focus manipulations, the control group reported an average of 5.26, which
is close to MfU n d rather than M benefit- This is interesting because it implies that the
manipulation maintains the fund focus by the subjects instead of enhancing it. In
other words, benefit focus can be enhanced by manipulations while fund focus
73
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
manipulation keeps the chronic fund focus level, which gets weaker under benefit
focus manipulation. We will further discuss the implication of this difference
between two regulatory focus manipulations in the next chapter.
The regulatory reference had two questions for its manipulation check. The
subjects were asked if, ‘I was mainly thinking about whether I have to buy the
inexpensive alternative,’ a manipulation check question for not-buying reference,
and if, ‘I was mainly thinking about whether I have to buy the expensive alternative,’
for buying reference. Although both questions resulted in a directional support for
the manipulation (Maying = 3.71 vs. Mn o t.b u y in g = 3.85 for not-buying question and
Maying = 4.03 vs. Mn o t-b u y in g = 3.77 for buying), neither reached the level of
significance. The control group fell right between the figures presented above,
showing averages of 3.78 (not-buying reference question) and 3.97 (buying reference
question). It seems that it is difficult to manipulate regulatory reference and generate
enough separation between buying and not-buying references, and the directional
support indicates that the manipulation worked to some degree. However, the
weakness of the regulatory reference manipulation has to be taken into account when
analyzing and interpreting the data in the regulatory reference dimension.
Overspending Behavior The proposed conceptual model postulates that
making a self-regulatory state salient is likely to enhance the probability of
overspending compared to the control group. The control group had 35.5%
likelihood of overspending while the test groups had 62.5%, 85.4%, 40.0%, and
55.3% in benefit gain from buying, benefit nongain from not buying, fund loss from
74
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
buying, and fund nonloss from not buying conditions, respectively. The ANOVA
including the control group showed significant main effects for both regulatory focus
and reference (F(l, 183) = 12.34, p < 0.005, s2 = 0.063 for regulatory focus
■ y
dimension and F(l, 183) = 7.15, p < 0.01, s = 0.038 for regulatory reference), but
the interaction was not significant. Need for cognition turned out to be insignificant
'y
while buying impulsiveness was significant (F(l, 183) = 6.09, p < 0.05, sz = 0.032).
The contrast analysis indicated that the overspending probabilities in benefit
gain from buying (Hjf) and benefit nongain from not buying (Hu,) conditions were
significantly different from the probability of the control group (p < 0.05 and p <
0.001, respectively). Fund nonloss from not buying (Hic ) condition showed
marginally significant difference from the control group (p < 0.10), but fund loss
from buying (Hid) condition failed to show significant difference. In short, Hja and
Hib are supported, Hic is marginally supported, and Hid is not supported. This
indicates that, even after the effect of buying impulsiveness was teased out, there are
significant differences in overspending probability among the self-regulatory states.
The benefit focus conditions especially were found to have higher probabilities for
overspending than the control group (see Figure 6 for the summary of overspending
probabilities for control and test groups).
75
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Figure 6. Overspending Probabilities in Study 1
0.9
Not Buying
0.7
J Q
( 0
- Q
O
L .
Buying
| 0.5
S 2
0 )
O 0.4
Control
0.3
Fund Benefit
Regulatory Focus
Following up on the results that most of the test groups reported higher
probabilities for overspending in the analysis above, the differences among the test
groups were analyzed. The ANOVA without the control group showed main effects
by both regulatory focus and reference dimensions (F(l, 153) = 12.88, p < 0.001, s
= 0.078, F(l, 153) = 6.60, p < 0.05, s2 = 0.041, respectively). A contrast analysis
(see Table 5) showed that benefit gain from buying and fund nonloss from not buying
is the pair that fails to be significantly different from each other, and the difference
between fund loss from buying and fund nonloss from not buying is only marginally
significant (p < 0.15). All the other pairs were found to be significantly different
from each other. The main effect of regulatory focus was significant, but it failed to
support H2a, which predicted the benefit gain from buying condition to be the highest
in overspending probability. Benefit gain from buying condition generated less cases
76
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
of overspending than benefit nonloss from not buying condition, > which
predicted higher overspending probability in fund focus than in benefit focus, was
also not supported because the main effect of regulatory focus favored benefit focus,
not fund focus. However, f/^c, which suggests a main effect on regulatory reference
favoring not-buying, is supported by the main effect presented earlier. H2d, which
further posits fund focus and not-buying reference as the highest in overspending
probability, failed to reach significance.
Table 5. Contrast Analysis of Overspending Probabilities in Study 1
Benefit gain
from
buying
Benefit nongain
from
not buying
Fund loss
from
buying
Fund nonloss
from
not buying
Overspending
Probability
62.5% 85.4% 40.0% 55.3%
Benefit gain
From
Buying
- - - -
Benefit nongain
from
not buying
-22.9%a
(p < 0.05)
- - -
Fund loss
from
buying
22.5%
(p < 0.05)
45.4%
(p< 0.001)
- -
Fund nonloss
from
not buying
7.2%
(n.s.)
30.1%
(p < 0.01)
-15.3%
(p <0.15)
-
a. The difference is computec by (column - row).
In short, the exclusivity hypothesis (IFfi was not supported, but one of the
three addition hypotheses (H2C ) was supported. It is difficult to conclude if self-
regulatory states are additive based on these results. However, the results indicate
that it would be the case that only regulatory reference is additive, not the regulatory
77
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
focus. The hypotheses H2 b - Ihd are developed based on the postulation of both
dimensions being additive, with main effects (for regulatory focus and reference in
H2 b and H2 c, respectively) and an interaction effect (H2J), but only the hypothesis
regarding the main effect of self-regulatory reference was supported. This indicates
that consumers are unlikely to consider both foci simultaneously when making a
purchase decision while it is likely that consumers are capable of considering both
buying and not-buying references in an additive fashion but not exclusively. This
finding provides an important theoretical implication in applying self-regulation
theory in consumer choice context. Further discussion will be presented later in
Chapter V (Discussion) on this issue.
Process M easures The proposed conceptual model posits four decision
variables to have a major impact on consumer choice depending on the manipulated
self-regulatory state. These key decision variables are measured after the subjects
completed the task and are compared across the self-regulatory states. First,
regarding reservation price, which was hypothesized to be the major decision
variable for the benefit gain from buying condition (//ja), the subjects were asked,
‘whether I was willing to pay that much for a digital camera or not was the key issue
for my choice decision.’ The mean values clearly supported with M b en efit/b u y in g =
4.38 Compared to M b en efit/n o t-b u y in g — 3.17, M fu n d /b u y in g — 3.73, and M fu n d /n o t-b u y in g ~ ~ 3.97
(see Figure 7). The ANOVA showed a main effect by regulatory reference (F(l,
152) = 5.51, p < 0.05, b2 = 0.035) and an interaction effect by regulatory focus and
reference (F(l, 152) = 11.77, p < 0.005, s2 = 0.072). The contrast analysis
78
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
comparing the responses from the benefit gain from buying condition to the other
conditions pairwise showed significant differences except for the fund nonloss from
not buying condition, which reported marginal difference with p < 0.15 (see Table 6
for pairwise comparisons).
Figure 7. Importance Ratings of Reservation Price in Study 1
Buying
U )
c
0 ) 4
( 0
< D O
0 ( / >
1 ^
I r- 3.5
a
E
2.5
Benefit Fund
Regulatory Focus
Table 6. Contrast Analysis of Reservation Price Importance Ratings in Study 1
Benefit gain
from
buying
Benefit nongain
from
not buying
Fund loss
from
buying
Fund nonloss
from
not buying
Importance
Rating
4.38 3.17 3.73 3.97
Difference
from
benefit gain
from
buying
-
1.21
(p< 0.001)
0.65
(p < 0.05)
0.41
(p < 0.15)
79
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
The second key decision variable in the proposed conceptual model is fund
availability that was hypothesized to be a major factor in the fund loss from buying
condition (//?/,). The subjects were asked if, ‘whether I can afford it or not was the
key issue for my choice decision.’ Fund loss from buying condition reported the
highest importance rating score among the conditions (Mbe n e fit/b u y in g = 4.38, M benefit/not-
b u y in g — 5.00, hffu n ( j/b u y in g — 5.20, and M fu n d /n o t-b u y in g — 4.29, see Figure 8), and the
ANOVA reported a significant interaction effect between regulatory focus and
reference (F(l, 153) = 7.41, p < 0.01, s2 = 0.046). The pairwise contrasts of the fund
loss from buying condition showed significant differences in the importance ratings
for all pairs but with benefit nongain from not buying (see Table 7) showing partial
support for Fla.
The key decision variable for the benefit nonloss from not buying condition
was predicted to be anticipated regret in the proposed conceptual model (H3 c ). The
subjects were asked if ‘I was mainly thinking about how much I would regret my
choice later.’ The mean scores were M b en efit/b u y in g = 4.13, Mb e n e fit/n o t-b u y in g = 5.59,
M fun d /b u y in g = 3.98, and Mfu n d /n o t-b u y in g = 4.71 (see Figure 9) with a significant main
effect of regulatory reference (F(l, 153) = 17.78, p < 0.001, s2 = 0.104), a marginally
significant main effect of regulatory focus (F(l, 153) = 3.20, p < 0.10, s2 = 0.021),
and a marginally significant interaction effect between regulatory focus and
reference (F(l, 153) = 2.22, p < 0.15, s2 = 0.014). The p airw ise co m p a riso n reported
significant contrasts between the benefit nongain from not buying condition and all
the other conditions supporting Flic (see Table 8).
80
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Figure 8. Importance Ratings of Fund Availability in Study 1
O
c
2
o
Q .
E
5.4
5.2
N ot Buying
4.4
Buying
4.2
Fund Benefit
Regulatory Focus
Figure 9. Importance Ratings of Anticipated Regret in Study 1
O )
c
’ •M . — ,
(T J A)
* %
Q ) o
o (/)
s -
■c '
O T "
a.
E
6
Not Buying m .
5.5
5
4.5
Buying
4
3.5
Fund Benefit
Regulatory Focus
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Table 7. Contrast Analysis of Fund Availability Importance Ratings in Study 1
Benefit gain
from
buying
Benefit nongain
from
not buying
Fund loss
from
buying
Fund nonloss
from
not buying
Importance
Rating
4.38 5.00 5.20 4.29
Difference
from
fund loss
from
buying
0.82
(p < 0.05)
0.20
(n.s.)
-
0.91
(p < 0.05)
The fourth key decision variable in the proposed conceptual model is
perceived saving for the fund nonloss from not buying condition {H3fi. The subjects
were asked if, ‘I was mainly thinking about how much money I would still have after
buying a digital camera. The mean scores were hlben efit/b u y in g 3.60, N 4 b c n c tu /n o t-b u y in g
= 3.39, M fund/buying — 4.23, and M fu n d /n o t-b u y in g — 4.63 (see Figure 10). The ANOVA
reported only a significant main effect of regulatory focus (F(l, 153) = 10.99, p <
0.001, s2 = 0.067), and the pairwise contrast scores has the fund nonloss from not
buying condition showing significant differences from the other conditions except
the fund loss from buying condition (see Table 9) showing partial support for H3d.
In general, the four key decision variables suggested in the proposed
conceptual model for consumer choice under budget constraint received support
from the empirical data. This confirms that consumers do activate one of these key
decision variables depending on the self-regulatory state they are in.
82
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Figure 10. Importance Ratings of Perceived Saving in Study 1
B Not Buying
4.5
O )
c
Buying
0 )
(0
o o
0 ( / }
1 ^
o T-.
Q .
E
3.5
Benefit Fund
Regulatory Focus
Table 8. Contrast Analysis of Anticipated Regret Importance Ratings in Study 1
Benefit gain
from
buying
Benefit nongain
from
not buying
Fund loss
from
buying
Fund nonloss
from
not buying
Importance
Rating
4.13 5.59 3.98 4.71
Difference
from
benefit
nongain
from
not buying
1.46
(p< 0.001)
-
1.61
(p < 0.001)
0.88
(p < 0.05)
83
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Table 9. Contrast Analysis of Perceived Saving Importance Ratings in Study 1
Benefit gain
from
buying
Benefit nongain
from
not buying
Fund loss
from
buying
Fund nonloss
from
not buying
Importance
Rating
3.60 3.39 4.23 4.63
Difference
from fund
nonloss
from
not buying
1.03
(p<0.01)
1.24
(p < 0.01)
0.40
(n.s.)
-
Budget Revision vs. Violation A set of hypotheses labeled as H4 deals with
the budget revision and violation process in consumer choice context. To test these
hypotheses, the subjects were asked to, ‘indicate how much you would set as your
budget if you have to plan your digital camera purchase all over again.’ The
ANOVA reported a significant main effect of regulatory focus (F(l, 148) = 10.01, p
^ 0.01, £ 0.063) with the means of M b en efit/b u y in g $409.19, M b en efit/n o t-b u y in g
$395.68, Mfu n d /b u y in g = $366.54, and Mfu n d /n o t-b u y in g = $366.76. However, it failed to
show any significant difference between two regulatory references as hypothesized
in H 4. Another ANOVA was conducted to examine the effect of choice on the new
budget. Since only the subjects who decided to overspend would had to go through
either budget revision or violation process, the choice was thrown into the analysis as
an independent variable. The result showed that there was a significant main effect
by choice (F(l, 144) = 34.90, p < 0.001, s2 = 0.195) and regulatory focus (F(l, 144)
= 4.52, p < 0.05, £2 = 0.030), and a significant three-way interaction among
regulatory focus, reference, and choice (F(l, 144) = 4.41, p < 0.05, £2 = 0.030). The
84
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
two-way interaction between regulatory reference and choice was marginally
significant (F(l, 144) = 3.86, p < 0.10, s2 = 0.013) (see Figure 11).
450
420
% 390
a t
-a
3
m
5 360
C D
Z
330
300
Figure 11. New Budget After Choice
No Overspending
(Choice = 0)
Not Buying ^
s
s
s
s
s
V
s
Buying ♦--------------------------
n ♦
s
Benefit Fund
Regulatory Focus
450
Buying
420
Not Buying
390
a )
o i
■o
3
C O
3 360
o
z
330
Overspending
(Choice = 1)
300 -
Benefit Fund
Regulatory Focus
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
These results indicate that the subjects who chose to overspend are more
likely to raise the budget when they are in benefit focus with buying reference than
in the other self-regulatory state, and therefore, H 4a is supported. To confirm this
result, a one-way ANOVA only with the subjects who overspent was conducted with
regulatory reference as the only independent variable. The result provided a
significant support for H 4a (M a y in g = $432.13 and M not-buymg = $399.96, F(l, 91) =
4.46, p < 0.05, s = 0.047) confirming that consumers do revise their budget under
buying reference.
Budget violation process proposed to be linked to not-buying reference (H 4b)
failed to receive support. Although not as much as in buying reference condition,
not-buying reference still showed a major increase in the new budget figures
($399.96 in average compared to the original budget of $300). More discussion on
this issue is presented in the next chapter.
Time Another behavioral measure recorded was the time spent by the
subjects on each and every step of the task. There were 5 stages of the task: 1)
reading the instructions, 2) first manipulation, 3) evaluating the alternatives, 4)
second manipulation, and 5) making a choice. A MANOVA with all these 5 time
measures as dependent variables was conducted and an extended set of covariates
was included in the analysis because of their potential influence on the time
recorded. The covariates w ere: product category (d igital cam era) fam iliarity, brand
(Olympus) familiarity, brand quality perception, interest level in the product
category, perceived task difficulty, and interest level in the task, along with need for
86
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
cognition and buying impulsiveness that are always included. Need for cognition
and interest level in the task were found to have an impact on the evaluation time
(F(l, 145) = 10.07, p < 0.005, s2 = 0.065 and F(l, 145) = 4.90, p < 0.05, s2 = 0.033,
respectively). Category familiarity and perceived task difficulty influenced the time
subjects spent on the second manipulation, which follows the evaluation stage (F(l,
145) = 5.86. p < 0.05, e2 = 0.039 and F(l, 145) = 6.48, p < 0.05, e2 = 0.043,
respectively).
After sorting out the effects from the covariates, the remaining variances
reported two main effects. First, the regulatory focus was found to affect the time
spent on the second manipulation page (F(l, 145) = 6.89, p < 0.05, s = 0.045) where
the means were, M benefit = 31.7 and Mfu n d = 27.5 seconds. Evaluation stage by itself
failed to show any significant differences among the self-regulatory states, but the
composite measure for the total time spent on evaluating the alternatives (evaluation
plus second manipulation) supported the hypothesis that predicts longer time spent
on evaluating alternatives under benefit focus (7/ja). The means of the composite
evaluation time were Mbenefit = 65.9 and M fu n d = 61.2 seconds (F(l, 145) = 3.95, p <
0.05, s2 = 0.027).
Another hypothesis (Hit,) suggests that the time spent on making a choice
should be longer under benefit focus. It was found that there is a marginal effect of
regulatory focus on the time spent on the choice stage of the task (F(l, 1 4 5 ) = 3 .0 9 , p
< 0 .1 0 , e 2 = 0 .0 2 1 ) , but the direction was the opposite to what H 5b predicts. The fund
focus actually showed longer time spent ( M f und = 1 8 . 0 seconds) than the benefit
87
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
focus (Mbenefit =14.6 seconds). It is possible that the subjects who spent a longer
time evaluating the products would need less time to make a choice because they
already arrived at a decision entering the choice stage. To test this argument, an
ANOVA with the sum of the times spent on all three stages of evaluation and choice
(evaluation, second manipulation, and choice stages) was performed. The result
showed no significant main effect by regulatory focus, which, combined with the
main effect in evaluation time, leads to a conclusion that the positive impact of
benefit focus on time spent is only on evaluation stage and has a reverse effect on
choice stage due to the fact that the longer it takes to evaluate the alternatives, the
easier it is to make the decision.
Another unexpected result that was not hypothesized was the main effect by
regulatory reference on the evaluation time. The composite measure introduced
above as a sum of evaluation and second manipulation stages failed to show any
effect by regulatory reference, but the individual evaluation time (stage 3) was lower
with not-buying than with buying reference (F(l, 145) = 4.25, p < 0.05, s2 = 0.028).
The means are, Mbuying = 36.2 and M n o t-buying = 31.6 seconds. It can be argued that
buying reference is a deviation from the status quo (i.e., staying within the budget)
and encourages the subjects to take more time to evaluate the alternatives.
Emotional Responses Based on the social psychology literature on self
regulation, it w a s h y p o th esized that b en efit fo cu s is lik ely to in d u ce d ejection -related
emotions (e.g., happy, satisfied, and encouraged) {H r,a) and fund focus agitation-
related emotions (e.g., calm, relaxed, and relieved) (H(,[,). After making a choice, the
88
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
subjects were asked to rate their emotional states in 6 bipolar scale questions: sad—
happy, dissatisfied—satisfied, discouraged—encouraged, nervous—calm, tense—
relaxed, and upset—relieved. The first three fall into the category of dejection-
related emotions, and the next three are a subset of agitation-related emotions. A
factor analysis of these six measures supported the classification of the emotions into
dejection and agitation related loading happy, satisfied, and encouraged to dejection
and calm and relaxed to agitation dimensions (see Table 10). The upset—relieved
measure was loaded to both dimensions, failing to be a clear agitation-related
emotion as predicted.
Table 10. Factor Scores for Emotional Responses
Measures
Component
Dejection Agitation
Sad—Happy -0.003 0.811
Dissatisfied—Satisfied -0.000 0.844
Discouraged—Encourages 0.286 0.802
Nervous—Calm 0.943 -0.007
Tense—Relaxed 0.937 -0.009
Upset—Relieved 0.666 0.509
Note:
1. Extraction method: Principal component analysis
2. Rotation method: Varimax with Kaiser normalization
3. Eigenvalues: 2.846 (1 component), 1.731 (2 components), and 0.503
(components).
Based on the factor analysis, composite scores were computed from the
measures: one representing dejection-related and another representing agitation-
related emotion. The dejection-related emotion score was computed by taking the
average of happy, satisfied, and encouraged measures and the agitation score by the
89
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
average of calm and relaxed. The measure for ‘relieved’ was dropped in the
analysis.
No significant effect of regulatory focus on the type of emotional response
was found from the analysis, and therefore, neither Hea nor H ^b were supported. The
insignificance found in this analysis can be attributed to the complexity of the task
compared to the tasks used in social psychology literature. Anagram solving was
often used to test the effect of regulatory focus on emotional responses in social
psychology (e.g., Higgins, Bond, Klein, and Strauman 1986; Strauman 1996;
Strauman and Higgins 1987), and the outcome of the anagram solving task was
controlled in the experiments. In other words, the task did not involve any decision
making by the subjects. In contrast, the task given to the subjects in the present
study had two different possible outcomes: stay within the budget or overspend.
Therefore, the emotional response should be found different not as a main effect of
self-regulatory state, but an interaction between self-regulatory state and choice. In
other words, a consumer with benefit focus would feel differently when he chooses
to stay within the budget than when he overspends.
A MANOVA with both dejection and agitation scores was conducted to test
the interaction between self-regulatory state and choice. Main effects on the
dejection score were found with both regulatory focus (F(l, 148) = 10.07, p < 0.005,
s2 = 0.064) and reference (F(l, 148) = 4.00, p < 0.05, s2 = 0.026), but no main effect
was found on the agitation score. The choice turned out to have a significant impact
on the dejection score (F(l, 148) = 12.30, p < 0.005, s2 = 0.077) and a marginal
90
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
impact on the agitation score (F(l, 148) = 2.85, p < 0.10, z = 0.019). Regulatory
focus was also found to have a marginal interaction with choice for the dejection
score (F(l, 148) = 3.83, p < 0.10, s2 = 0.025).
To understand the interaction effect of regulatory focus and choice on
emotional responses, another M A N O V A was conducted with agitation and dejection
score as dependent variables and regulatory focus and choice as independent
variables. The analysis failed to report any significance for the interaction effect, but
the main effects of choice on both dependent variables were found significant (F(l,
152) = 9.42, p < 0.005, s2 = 0.058 for dejection score and F(1,152) = 4.55, p < 0.05,
e2 = 0.029 for agitation score). This means that the main effect of choice overrides
the differential impact of regulatory focus on emotional responses. The means show
that the subjects experienced more positive dejection related emotions
(Mdejection/overspent = 5.16, M dejection/stayed = 4.74) and less positive agitation related
emotions (Ma g itatio n /o v ersp en t— 4.61, Ma g jta tio n /sta y e d — 5.19) when they overspent (see
Figure 12). In other words, the subjects were more likely to feel happy, satisfied,
encouraged and less likely to feel calm and relaxed when they chose to overspend
than when they stayed within the budget.
2. Study 2: Interaction Between Self-regulatory State and Absolute Budget
Level
Study 2 adds a situational variable—absolute budget level—as another
independent variable and tests its interaction with the regulatory focus and reference.
91
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
The results failed to support a set of hypotheses labeled as Hy predicting more
variance in overspending probability when absolute budget level is high. Only a
main effect of regulatory focus on overspending probability was found significant
(F(l, 290) = 29.60, p < 0.001, 82 = 0.093), which is consistent with the finding in
Study 1.
Figure 12. Emotional Responses Depending on Choice in Study 1
0 )
>
Dejection
related
emotions
(A
0
a
1
h -
6
>
4 - )
CT 5
O )
0 )
c
I
Agitation
related
'■ emotions
4.5
Overpsent Stayed withing the budget
Choice
Even though it failed to reach significance, the comparison between the
overspending probability patterns in low and high absolute budget level showed that
only the low absolute budget level condition resembles the findings in Study 1. High
absolute budget level condition did not show any difference between the buying and
not-buying conditions (see Figure 13). Recall that Study 1 found regulatory focus to
be exclusive and regulatory reference to be additive.
92
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Figure 13. Overspending Probabilities in Study 2
0.5
Low Absolute Budget Level
£ 04
!5
0.3
o >
0.2
Not Buying
0.1
Buying
0
Benefit Fund
Regulatory Focus
0.5
High Absolute Budget Level
0.4
0.3
C L
O )
c
' ■ £
c
4 )
Q .
1 2
< u
>
O
Not Buying
Buying
Fund Benefit
Regulatory Focus
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The findings in Study 2 were consistent with the findings in Study 1
indicating that the subjects activated both buying and not-buying references in the
high absolute budget condition. The null findings of the manipulation checks for the
regulatory reference manipulation in this study also supported this argument.
The failure to support the hypotheses in Study 2 could be explained by two
reasons. First, it could indicate that there are differences in the decision making
processes in consumer choice with high vs. low absolute budget level. The
hypotheses were developed based on the notion of indifference (low absolute budget
level) vs. ambivalence (high absolute budget level), and being ambivalent might
interact with self-regulatory states, especially regulatory reference. In other words,
when consumers are ambivalent among choices, it would be difficult to put them in
one regulatory reference when they evaluate the alternatives and make a choice.
Consequently, the manipulation failed and led to null results. An alternative
explanation is the level of manipulation. The stimulus presented the subjects with
$150 and $200 alternatives for low absolute budget level conditions and $599 and
$799 alternatives for high conditions. It could have not been enough to differentiate
the absolute budget level, which resulted in similar patterns of overspending in both
conditions. A future study to further understand the role of absolute budget level
will be presented in the last chapter.
94
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
3. Study 3: Interaction Between Self-regulatory State and Core Attribute
Type
Study 3 adds a situational variable—core attribute type (functional vs.
hedonic)—as another independent variable and tests its interaction with the
regulatory focus and reference. It was hypothesized in the previous chapter that the
salience of hedonic attributes contributes to higher overspending. The results of an
ANOVA failed to report any main or interaction effect of the core attribute type on
the overspending probability. Buying impulsiveness and need for cognition
explained much of the variance (F(l, 253) = 11.70, p < 0.005, £2 = 0.044, and F(l,
253) = 10.88, p < 0.005, s2 = 0.041, respectively). The only effect that was found to
be significant was the interaction between regulatory focus and reference (F(l, 253)
= 7.72, p < 0.01, £2 = 0.030), and the main effect of regulatory focus was found to be
marginally significant (F(l, 253) = 3.70, p < 0.10, s2 = 0.014). Figure 14 shows the
overspending probabilities of each condition. The fund loss from buying condition
reported an inconsistent result with Study 1 by displaying the highest overspending
probability (Mbenefit/buying ~' 0.37, Mbenefit/not-buying — 0.34, Mfund/buying “ 0.13, and
Mfund/not-buying — 0 .3 8 ) .
Further investigation of the effect of core attribute type reported that it has
significant main and interaction effects on one of the manipulation check
questions—the importance rating of affordability (‘I was mainly thinking if I was
able to afford the prices of the alternatives’). Affordability was rated relatively
lower for hedonic attribute condition (M hedonic = 4.98) than in functional condition
95
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
(Mfu n c tio n a i = 5.53) (F(l, 253) = 6.66, p < 0.05, s2 = 0.026), and the three-way
interaction with regulatory focus and reference showed that the manipulation only
worked for functional attribute conditions (F(l,253) = 4.65,p< 0.05,e = 0.018, see
Figure 15). This partially explains the insignificant result in overspending behavior
by indicating that the subjects were not properly manipulated under the hedonic core
attribute condition. More specifically, the subjects in the hedonic core attribute
condition apparently had a difficult time focusing on the fund side of the purchase.
Figure 14. Overspending Probabilities in Study 3
0.4
Buying
0.3
0.2
> Not Buying
0.1
Benefit Fund
Regulatory Focus
96
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Figure 15. Fund Focus Manipulation Check for Study 3.
U )
.E
03 A)
* 3
Q ) O
O (/)
I t
o n.
Q .
E
6.5
Functional Buying
6
5.5
5
-a Not Buying
4.5
Benefit Fund
Regulatory Focus
o>
e
+ 3
( 0 < 1 >
* «
a > o
£
o
Q .
E
o ^
Q .
6.5
Hedonic
6
5.5
M Not Buying
5
Buying
4.5
Benefit Fund
Regulatory Focus
A similar result was found for the importance of the budget during the
evaluation stage. In Study 1, the importance of the budget was higher in fund focus
and buying reference. However, Study 3 reported a three-way interaction that the
97
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
subjects were less concerned about the budget in fund than in benefit focus and in
buying than in not-buying reference when the core attribute was hedonic (F(l, 253) =
4.13, p < 0.05, s = 0.016, see Figure 16), which was the opposite trend of the
findings of Study 1. This result is consistent with the failed manipulation check
presented above indicating that putting subjects into fund focus is more difficult
when the core attribute is hedonic. This explains the failure to report any
significance in the other manipulation checks than the affordability manipulation.
For example, the importance rating of anticipated regret failed to show any
difference among conditions when the core attribute was hedonic, and it can be
attributed to the failure to manipulate fund focus.
Another notable finding in Study 3 was the significant differences among the
conditions for the reservation price measured after the choice. A two-way
interaction was reported between regulatory focus and core attribute type (F(l, 250)
'y
— 4.15, p < 0.05, £ = 0.016, see Figure 17). More specifically, the functional core
attribute condition showed a similar pattern to Study 1, but the hedonic core attribute
condition failed to show any difference between benefit and fund focus. Considering
the budget was given at $300 for Study 3, the subjects raised the reservation price
regardless of the regulatory focus manipulation (Mbenefit/hedonic = $354.77, M fund/hedonic
= $354.48), which is again consistent with the findings presented above concerning
the failed fund focus manipulation.
98
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Figure 16. Budget Importance Rating During The Evaluation Stage in Study
6.5
Buying
o >
T -
5.5
H Not Buying
Functional
Fund Benefit
Regulatory Focus
6.5
Hedonic
a )
re
0
tn
1
5.5
Not Buying
Buying
Benefit Fund
Regulatory Focus
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Figure 17. Revised Reservation Price After Choice in Study 3
360
355
m Hedonic
350
345
340
335
330
Functional 325
320
Fund Benefit
Regulatory Focus
In sum, the null finding in Study 3 in terms of overspending probability is not
discouraging. It is rather encouraging that the results shed insights into how the
situational variables influence the overspending behavior. More specifically, it is
apparent that the nature of the core attribute interacts with the activation of the self-
regulatory state, rather than the choice process. This means that there is a certain
self-regulatory state with which consumers would feel comfortable in making a
choice depending on the level of a situational variable. Study 3 is believed to be an
example of hedonic core attributes making it difficult to manipulate consumers to
activate fund focus.
100
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
V. G e n e r a l D i s c u s s i o n
The previous chapter presented the results of the empirical tests that
examined the proposed conceptual model for consumer choice under budget
constraint. The proposed conceptual model received strong support from the pilot
studies and Study 1, confirming that overspending can be explained with four key
decision variables—reservation price, fund availability, anticipated regret, and
perceived saving—depending on the self-regulatory state activated by the consumer.
1. Exclusivity vs. Addition Hypothesis
An important question that lies behind the theoretical foundation provided by
the past literature on self-regulation theory in social psychology is whether the
different self-regulatory states are exclusive or additive. In other words, can
consumers activate benefit and fund foci or buying and not-buying references
simultaneously while making a purchase decision? Social psychology literature fails
to provide any conceptual framework on this question partly because self-regulatory
states are often measured as a personality trait where one state dominates when not
manipulated. When it is manipulated as done in the present study, it is possible for
the personality trait and the manipulated self-regulatory state to interact or conflict
with each other.
Compared to the fact that Pilot Study 2 found a chronic preference of benefit
gain from buying to the other self-regulatory states, the reported overspending
101
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
probabilities can provide some insights on the issue of exclusivity vs. addition
hypothesis. Study 1 found that buying focus, which is found to be chronic
preference in Pilot Study 2, is more effective in inducing overspending than fund
focus, and not-buying reference, which is not the chronic preference, is more
effective than buying reference. This indicates that activating multiple regulatory
foci does not add up and inflate the overspending probability, but multiple regulatory
references do. In other words, a consumer with a chronic preference of benefit focus
and buying reference would be likely to overspend when exposed to a marketing
message designed around benefit focus and not-buying reference but not when
exposed to a message pitching fund focus and buying reference. Therefore,
regulatory focus is exclusive, but regulatory reference is additive.
The fact that regulatory reference is additive in its nature might be able to
explain the weak results found in its manipulation checks. Due to its exclusive
nature, regulatory focus manipulation should work well as found in Study 1 because
the subjects choose one focus over the other, not both. On the other hand, as
described in the previous chapter, the manipulation checks for regulatory reference
failed to reach significance even though the direction was consistent with the
manipulations. It is somewhat mysterious that the manipulation did not work
cleanly, but most dependent variables were found to be different depending on the
regulatory reference as hypothesized. This can be attributed to the additive nature of
regulatory reference. More specifically, the additive nature enabled the subjects to
use the regulatory reference that was not manipulated along with the manipulated
102
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
reference with ease, which created large variances for both manipulation check
measures. As a result, it became difficult to reach significance with the inflated
variances. Therefore, the directional support for the manipulation can be a fair
indicator to identify which regulatory reference was the dominant one in the decision
making process. At the same time, it should be admitted that the separation between
two regulatory references was very small, indicating that many subjects did consider
both references to some extent, which discounts any findings related to regulatory
reference.
One of the findings that should be examined carefully is the budget revision
vs. violation. The result showed a significant difference between buying and not-
buying references but when including in the analysis only the subjects who
overspent. It can be argued that the subjects who were successfully manipulated
were more likely to overspend than the ones who were not, so that the effect of the
regulatory reference manipulation was only observable from the subjects who
overspent. This supports the idea that the manipulation was successful to some
extent. Furthermore, the finding that the subjects still raised the budget compared to
the original level under budget violation process should be examined carefully in
view of the additive nature of regulatory reference. It can be interpreted that even
the subjects assigned to the not-buying reference conditions were able to consider the
buying reference simultaneously and make the decision accordingly. If this was the
case, the hypothesis that predicted no budget increase under not-buying conditions
103
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
should not be rejected. In other words, the possibility still remains that the
hypothesis would be supported when only not-buying reference is in effect.
The null results in Study 2 can also be attributed to the additive nature of
regulatory reference. The main effect of regulatory focus that is consistent with
Study 1 was found, but no main effects of regulatory reference and absolute budget
level were significant. As described in the previous chapter, further investigation
identified that high absolute budget level conditions showed no difference between
buying and not-buying references, which implies that the subjects considered both
buying and not-buying references at the same time when the absolute budget level
was high. It can be interpreted that high absolute budget level forced the subjects to
consider both references to be more elaborate in their decision making process.
When the absolute budget level is low, it does not require elaborate processing, and
as a result, both references are less likely to be considered simultaneously. On the
other hand, regulatory focus manipulation still worked and showed its effect on
overspending behavior because of its exclusive nature.
In short, the evidences found in Studies 1 and 2 support the notion that
regulatory focus works exclusively and regulatory reference additively in a consumer
choice context. It is also suggested the higher the absolute budget level, the higher
the likelihood of activating both references is. This has strong theoretical and
managerial implications. Theoretically, it extends the self-regulation theory by
identifying the interactive natures of regulatory foci and references. In a consumer
choice context, it should be noted that consumers are capable of comparing
104
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
alternatives with multiple reference points but not with multiple foci (benefit and
fund). Therefore, marketing messages should be designed around one regulatory
focus at a time to be effective but may emphasize both regulatory references
simultaneously. However, it is possible that consumers incorporate multiple foci in
different stages. For example, a consumer would apply fund focus while narrowing
down the alternatives to a few and apply benefit focus when choosing one for
purchase. It is recommended to identify which focus is more effective in each stage
of the choice process for a marketing communication program to be effective. It is
also necessary to understand the role of the relevant situational variables such as
absolute budget level because they interact with triggering one or both regulatory
references.
2. Emotional Responses
Self-regulation theory predicts that promotion focus induces more dejection
related emotions, and prevention focus induces more agitation related emotions.
Adopting these predictions, Study 1 was expected to show similar trends in
emotional responses from benefit (promotion) and fund (prevention) focus.
However, as reported in the previous chapter, the hypotheses in Study 1 related to
the emotional responses failed to post significance, but the behavioral outcome
(choice) was found to interact with the emotional responses. More specifically, the
subjects who overspent showed stronger positive dejection related emotions (happy,
satisfied, and encouraged), and the ones who stayed within the budget showed
105
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
stronger positive agitation related emotions (calm and relaxed). Applying the
predictions of self-regulation theory to these findings in Study 1, one can draw a
conclusion that overspending behavior itself is a promotion focus and staying within
the budget a prevention focus.
There are two possible explanations for this finding. First, it is possible that
regardless of the manipulation, the subjects in promotion focus overspend, and the
ones in prevention focus do not. Then the subjects who overspent show dejection
(agitation) related emotions not because of the overspending (staying within the
budget) behavior but because of the promotion (prevention) focus that induced the
behavior. The second explanation suggests that the emotion induced from the
behavioral outcome overrides the emotional responses from the self-regulatory state.
More specifically, the self-regulatory state induces a set of emotions following the
predictions of self-regulation theory, and the behavioral outcome induces another.
The literature in social psychology often employs an anagram solving task in self-
regulation research, in which the subjects do not have a choice in terms of their
behavioral outcome. Therefore, the emotional responses recorded after the task can
be attributed to the self-regulatory state. On the contrary, the consumer choice task
employed in Study 1 involved two options—stay within the budget and overspend—
for behavioral response. Therefore, it is possible that the choice of behavioral
response induces its own set of emotions regardless of the emotional responses
induced by the self-regulatory state. In sum, the emotional responses recorded in
106
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Study 1 are not the equivalent measures to the ones reported in social psychology.
Instead, it is an independent set of emotions induced by the behavioral response.
The first explanation challenges the conceptual foundation of the present
study’s proposed model. However, it lacks the support from other measures such as
the manipulation checks and time variables. For example, social psychology
literature predicts a longer time spent on the task in promotion focus than in
prevention focus. The finding in Study 1 is consistent with the above prediction
implying that benefit and fund foci do represent promotion and prevention foci,
respectively. Therefore, a main effect of regulatory focus should have been reported
for the emotional responses, which was not found in Study 1, and the second
explanation would be more adequate to explain the findings in Study 1. In other
words, it is not the emotional responses from the manipulation or the process that are
recorded in Study 1 but the emotional responses to the behavior, which was the last
stage of the task. Therefore, it would be inaccurate to infer anything from the
emotional responses other than their relationship with the behavioral responses.
3. Interaction with Situational Variables
Studies 2 and 3 both found null results in terms of the effects of the
situational variables on overspending behavior. It was hypothesized in Chapter III
that these situational variables were expected to interact with the consumer choice
processes after a self-regulatory state was activated through manipulations.
However, Study 2 found that absolute budget level interacted with the regulatory
107
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
reference manipulation, showing that high absolute budget level induced both
regulatory references to be activated. Study 3 showed similar findings regarding the
fund focus manipulation. The fund focus conditions showed erratic manipulation
check results under hedonic core attribute, while functional core attribute conditions
showed consistent manipulation check figures with Study 1. The importance of
budget in Study 3 also showed a significantly lower figure in hedonic core attribute
conditions, which implies with the problematic manipulation check that it is difficult
to put the subjects in fund focus when the core attribute is hedonic.
These problems in Studies 2 and 3 provide us with a valuable insight into the
role of situational variables in consumer choice. Instead of interacting with the
processes, it is apparent that these situational variables influence the activation of the
self-regulatory states. For example, high absolute budget level activates both buying
and not-buying references, and hedonic core attribute pushes consumers toward
benefit focus. This has a strong theoretical implication that it is possible not to
observe any effects even though the conceptual model proposed in the present study
is valid because the context in which the model is tested can limit the effectiveness
of the manipulations. Fortunately, the context used in Study 1 turned out to be
neutral enough to observe the effects of the manipulated self-regulatory states to
consumer choice. The managerial implication of this finding is that marketers
should not simply rely on the findings in Study 1 to enhance or reduce the
overspending probability because spending behaviors depend on context factors such
as the price level and the nature of the product core attribute.
108
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
4. Validation of Proposed Conceptual Model Using Path Analysis
It has been discussed in various aspects that the manipulations in the
empirical testing of the proposed conceptual model are subject to context factors.
Even though the model received strong support in Study 1, testing the model without
the influence of the context factors would be valuable to validate the model. To do
so, a path analysis using a structural equation model was performed with the
manipulation check measures as the antecedents instead of the category variables
used as independent variables in the ANOVA and MANOVA reported earlier. In
other words, the assignments of the subjects to certain conditions were ignored, and
the self-report regulatory focus and reference were used as the antecedents in the
model. This eliminated the possible interaction between the context factors and the
self-regulatory state manipulation and provided more rigorous test for the proposed
conceptual model. In addition, it would be possible to observe the effect of
regulatory reference, which had trouble in its manipulation, on consumer choice
more clearly.
More specifically, benefit and fund focus manipulation check measures were
defined as antecedents with an assumption that they are correlated as were buying
and not-buying reference manipulation check measures. These four antecedents
were defined to be the predictors for the importance ratings of the key decision
variables (reservation price, fund availability, anticipated regret, and perceived
saving). Subsequently, these four key decision variables were defined to be the
1 0 9
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
predictors for the choice. An exploratory path analysis was conducted with the data
set of Study 1 to compute the model parameters using EQS version 5.7b, and the
'S
estimates are summarized in Table 12 with the incremental % and p-value for each
estimate. The x2(14) for the model’s goodness of fit was 187.859 (p < 0.001), and
the Comparative Fit Index (CFI) was 0.378. Figure 18 illustrates the model with
standardized parameter estimates and significance level. Due to the small sample
size, a generous threshold of p = 0.25 for marginal significance is adopted in the
following discussion and only the relationships with p < 0.25 are shown in the figure.
The solid lines in the figure illustrate the relationships predicted by the proposed
model, and the dotted lines demonstrate the relationships that were not predicted by
the proposed model but still turned out to be (marginally) significant.
Considering the fact that only 158 cases were used in the analysis, the
structural equation model provides encouraging results supporting the proposed
model. The sample size, smaller than what a structural equation model normally
requires (30 per variable, 270 in the present study), limits the value of the results, but
the path analysis serves as a useful tool to offer additional support for the findings
presented in the previous chapter. Reservation price, as predicted, was influenced
mostly by buying reference (p < 0.05), with the benefit focus being second predictor
even though it was only marginally significant (p < 0.25). Fund availability was not
fou n d to b e co n sisten t w ith the p rop osed m o d e l’s p red iction w ith n o t-b u y in g
reference (p < 0.05) as its sole predictor.
110
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Table 11. Parameter Estimates of Path Analysis Using The Data from Study 1
Variable Pair Parameter Estimate
/ P
Outcome Predictor
Plain
Solution
Normalized
Solution
Choice
Reservation Price 0.056 0.167 5.41 0.020
Fund Availability -0.011 -0.040 0.31 0.577
Anticipated Regret 0.066 0.245 11.52 0.001
Perceived saving -0.080 -0.296 18.96 0.000
Reservation
Price
Benefit Focus 0.088 0.096 1.36 0.244
Fund Focus 0.009 0.010 0.01 0.908
Buying Reference 0.129 0.154 4.14 0.042
Not Buying Reference -0.051 -0.060 0.59 0.442
Fund
Availability
Benefit Focus 0.059 0.055 0.37 0.541
Fund Focus 0.082 0.075 0.38 0.540
Buying Reference 0.004 0.004 0.00 0.963
Not Buying Reference 0.158 0.157 3.95 0.047
Anticipated
Regret
Benefit Focus 0.112 0.099 3.22 0.073
Fund Focus -0.096 -0.085 0.90 0.342
Buying Reference -0.020 -0.020 0.06 0.801
Not Buying Reference 0.108 0.103 1.80 0.180
Perceived
Saving
Benefit Focus -0.150 -0.133 2.62 0.106
Fund Focus 0.336 0.294 24.78 0.000
Buying Reference -0.007 -0.007 0.01 0.927
Not Buying Reference 0.219 0.208 8.36 0.004
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Figure 18. Path Analysis Results for Proposed Conceptual Model in Study 1
-0.483
-0.084
Benefit
Focus
0.096
Reservation
Price
&
Fund
V \ /
\ \ /
v < 9 \ Y
Fund
Focus
N k A
4
/
/
/
Availability
' S '
A '
Buying
X X ' n X
/ \ / N M
f X / \ Anticipated
Reference
' X N ^
/ \ s *
Regret
Not-buying
/
/^ r
--------fc-
Perceived
Reference 0.208***
w
Saving
Choice
Note:
No asterisk: p < 0.25
* : p < 0.10
** : p < 0.05
*** . p < 0.005
Anticipated regret was influenced by benefit focus (p < 0.10) and not-buying
reference (p < 0.20) with marginal significance, as predicted. Perceived saving was
found to be consistent with the proposed model’s prediction with fund focus (p <
0.001) and not-buying reference (p < 0.005) as the predictors. Additional
relationships that were found marginally significant were benefit focus influencing
perceived saving (p < 0.15), which implies that higher benefit focus reduces the
112
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
emphasis on perceived saving and makes intuitive sense. Among the four predictors
of choice, all but fund availability were found to be significant with correct signs.
The null effect of fund availability is not surprising because of its failure to report a
statistical difference from the control group in terms of overspending probability as
shown in Figure 6. Higher importance of reservation price and anticipated regret and
lower emphasis on perceived saving led to higher overspending probability.
The correlation coefficient between benefit and fund focus scores was -
0.483, which supports the argument presented earlier that regulatory focus is
exclusive. In other words, the strong negative correlation means only one of the foci
is activated, not both. On the other hand, the correlation coefficient between buying
and not-buying reference scores was -0.084, which represents almost a random
relationship. It means that activating one regulatory reference does not necessarily
lessen the chance to activate another reference, which supports the earlier argument
that regulatory reference is additive.
In summary, the proposed relationships between self-regulatory states, key
decision variables, and choice turn out to be consistent with the proposed conceptual
model except for the relationships involving fund availability. Considering the fact
that the subjects were college undergraduate students who do not normally have
flexibility in their budget might have limited the effect of fund availability on choice.
In other words, the subjects had a difficult time increasing the fund availability to be
able to overspend. There were no surprising relationships that were found significant
in the path analysis other than the one between not-buying reference and fund
113
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
availability, which is a strong support for the overall structure of the model. In
addition, the path analysis supports the notion that regulatory focus is exclusive in
nature and regulatory reference is additive.
It is important that the proposed model received fairly strong support from both
(M)ANOVA and path analysis. The proposed choice processes involving four key
decision variables were confirmed in both analyses except the process involving fund
focus and buying reference. Furthermore, the suggested exclusive nature of
regulatory focus and additive nature of regulatory reference were also confirmed in
both analyses. Since the path analysis eliminates the intervention of context factors,
the consistent findings imply that the proposed conceptual model is indeed in effect
when consumers make a choice under budget constraint. The failure to verify the
fund loss from buying process calls for further research.
114
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
V I. T h e o r e t ic a l C o n t r ib u t io n s a n d M a n a g e r ia l I m p l ic a t io n s
The present research attempts to understand consumer choice under budget
constraint and identify the key variables for the choice processes. Building upon the
self-regulation theory literature in social psychology, the self-regulatory states that
lead to different motivations to resolve the conflict between staying within the
budget and overspending are defined with benefit and fund foci and buying and not-
buying references. Subsequently, the key decision variables (reservation price, fund
availability, anticipated regret, and perceived saving) that link these self-regulatory
states and choice are identified to explain the processes of consumer choice under
budget constraint.
1. Application of Self-regulation Theory to Consumer Context
The major theoretical contributions of the present study is threefold. First, it
provides the framework for applying self-regulation theory to a consumer behavior
context. As shown throughout the present study, self-regulation, which is a conflict
resolution mechanism, is found to be advantageous to conceptualize purchase
process because purchase is in its nature a conflict resolution task where consumers
have to balance the benefit and cost. By linking benefit to promotion and fund to
prevention focus and buying to desired and not-buying to undesired end-state
reference, the present study provides a self-regulatory framework that is applicable
to various marketing contexts beyond the present study’s scope.
115
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The second theoretical contribution is the identification of the nature of
regulatory focus and reference. As discussed in the last two chapters, regulatory
focus is found to be exclusive and regulatory reference additive. In other words,
consumers are not capable of considering both regulatory foci at the same time and
are capable of considering multiple regulatory references simultaneously, if
necessary. These findings extend our knowledge on self-regulation by specifying the
possible motivational consequences of self-regulation and theorizing their effects on
the behavioral outcome. For example, the conflict resolution becomes more efficient
when a set of motivations originated by multiple regulatory references drives the
behavior than when a set of motivations from multiple regulatory foci is present.
The third contribution is to find that the self-regulatory states are affected by
the context factors. This is important because it sets out how consumer research has
to deal with the self-regulatory state. It is more strongly recommended to posit the
context factors as the antecedents of the self-regulatory states than to posit them as
the moderators of the process initiated by the self-regulatory states. Regarding the
methodology to test theoretical models related to self-regulation, an extensive
investigation of the effects of context factors on the activation probabilities of
various self-regulatory states becomes a prerequisite. A path analysis using
structural equation modeling is proposed and employed in the present study to avoid
the possible confounding of context factors, but conventional (M)ANOVA would
still be appropriate under the condition where context factors are assured not to cause
serious bias in self-regulatory state activation.
116
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
2. Groundwork for Generalized Constraint M anagement Model
Furthermore, the present study provides the groundwork for understanding
the role of other kinds of constraints in consumer decision making. Since budget is
one of many constraints that influences consumer choice behavior, the proposed
conceptual model provides valuable insights into conceptualizing the role of other
constraints such as time pressure in consumer choice. There have been many studies
that examined the effect of time pressure on consumer choice (e.g., Cristol & Sealey,
1996; Dhar and Nowlis 1999), but a single conceptual framework that enables us to
understand the antecedents, psychological processes, and behavioral outcomes in
consumer behavior under time pressure has not been offered. It would be possible
for the proposed conceptual model with budget constraint to be translated into a time
pressure context and provide a conceptual model for understanding the effects of
time pressure. Furthermore, the present study’s conceptual model sheds lights on a
unified conceptual model that explains the consumer choice processes under
constraints in general.
3. Expansion of Proposed Model to Underspending Behavior
Another theoretical implication of this framework is the possibility of
understanding the process of underspending as well. If the gains and losses are
flipped in the model, a conceptual model that represents the processes behind
underspending can also be built. For example, maximizing benefit gain from buying
117
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
is a motivational force that encourages consumers to overspend by buying an
expensive product. On the contrary, to prevent a consumer from overspending, the
model suggests that activating a process, which can be defined as benefit loss from
buying, may be effective. In real day-to-day life, it is not difficult to find cases
where consumers actually lose benefits by choosing an expensive product such as
losing valuable family time after buying an expensive video game console.
Understanding both overspending and underspending behavior would be useful for
public policy makers whose task is to prevent heated overspending by consumers,
especially in the era of credit cards.
4. Managerial Implications
In addition to the theoretical implications discussed above, there are strong
managerial implications of the study. First, a firm that maintains multiple
specifications and models in a single product category would benefit from
understanding the consumer choice process under budget constraint. For example,
Sony has different 32-inch television models, but it also offers all different screen
sizes with multiple alternatives in each size category. There is a strong possibility of
cannibalization among the various Sony models, and promoting a certain
specification or model would be quite challenging. However, applying the proposed
theoretical model of the present study enables the seller to effectively position their
various models to minimize cannibalization and maximize profit. Since the present
model also suggests effective ways of persuading consumers to overspend, the seller
118
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
can design a marketing campaign around the findings from the present study to
maximize its effectiveness for each screen size and price level. Not only could the
sellers with a wide breadth of products enjoy this implication but also the sellers who
focus on a certain price range. If a firm manufactures only high priced items, the
proposed model would be able to provide valuable insights into how to design the
product and what attribute to be communicated. By understanding the consumer
choice process with budget constraints, the firm could also understand the nature of
competition by identifying the important decision variables for the target segment.
Another potential managerial implication is the utility of the present study’s
findings in direct marketing and personal selling. Since it has been reported that the
four routes to overspending can be activated by momentary stimulation, direct
marketers and salespeople can benefit from the present study’s findings to guide
consumers toward a specific product they want to pitch. The notion of exclusive vs.
additive nature of regulatory focus and reference, respectively, becomes practically
important in this perspective because the effectiveness of the sales pitch depends on
the content of the message related to regulatory focus and reference and to the effect
of the context factors as well. For example, mixing two different foci would not be
ideal, but designing a pitch with multiple references would be effective in some
cases. In addition, giving a sales pitch that emphasizes benefit focus while
prohibiting the product characteristics benefit focus from being activated would not
be effective.
119
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
The present study offers a systematic understanding of not only consumer
choice behavior under budget constraint but also what to expect from the seller from
the consumers’ perspective. Various sales tactics based on the proposed model on
overspending have been practiced for a long time in the field, and consumers have
not had a systematic approach that can prove the effectiveness of these sales tactics.
The present framework provides consumers with a better understanding of the
manipulative tactics used by sellers, enabling the consumers to be less influenced by
these tactics. The public policy makers, from the consumer welfare perspective,
would benefit greatly from the present study since it provides them with a scientific
background forjudging the legitimacy of sellers’ practices.
120
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
VII. C o n c l u d in g R e m a r k s , L im it a t io n s , a n d F u t u r e S t u d ie s
The present research investigates the role of budget in consumer choice and
builds a conceptual model that explains the processes. The most important
contribution of the present study includes: 1) operationalization of self-regulatory
states in consumer choice context, 2) identifying the key decision variables
connecting the self-regulatory states and behavioral outcomes, 3) identifying the
differential effects among the self-regulatory states on consumer choice behavior,
and 4) identifying the interrelationship among the self-regulation states. A multi
faceted approach is taken to validate the model and the results from both
(M)ANOVA and path analysis offered support.
The present study also has limitations that lead to future studies. First, the
budget formation process was not examined in the present study. Instead, it was
controlled so that only the effects of the processes could be cleanly observed. It is
intuitive that there exist various budget formation processes, and they should differ
in their influences on consumer choice. Therefore, understanding the budget
formation process is a necessary condition to clearly identify the antecedents of the
whole consumer choice process under budget constraint, and its interaction with the
activation of different self-regulatory states should be investigated as well.
The second limitation of the study was the number of alternatives, limited to
two for practical reasons. This provided clean data for overspending probability and
maintained the subjects’ involvement at a high level. However, it is necessary to
121
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
investigate the role of budget in a consumer choice context with more than two
alternatives. The choice set has to be expanded both horizontally (i.e., adding more
alternatives in the same product class with different prices), vertically (i.e., adding
more alternatives in different product classes with different prices), or both. For
example, the horizontal expansion would add more digital cameras to Study l ’s
stimulus in the same 1.3 to 2.1 megapixel range as the current alternative with
different prices, and the vertical expansion would add 3 megapixel cameras (beyond
the current attribute level) into the choice set. By doing so, the consumers would
have different ways of overspending, if they wish. The same amount can be
overspent by either going over your budget horizontally or vertically. Explaining the
difference between horizontal and vertical overspending would offer a direct
managerial implication to a firm’s product line design and extend the conceptual
model to higher external validity.
The third limitation of the present study involves the role of regulatory
reference in consumer choice. It was argued earlier that regulatory reference is
additive in its nature while regulatory focus is not, and this issue calls for further
investigation. The theoretical implication would be invaluable if a future study
grants strong support for the additive nature of regulatory reference. The study does
not have to be in a consumer choice context because the main goal of the study is to
conceptualize how self-regulatory states interact with each other when
simultaneously activated regardless of the context. It is a more fundamental issue in
self-regulation, and it would be a significant theoretical contribution to social
122
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
psychology and marketing, to design and test specifically the exclusivity and
addition hypotheses.
Another limitation of the present study is the postulation of the situational
variables’ roles in consumer choice process under budget constraint. It is suggested
that situational variables should be regarded as the antecedents for self-regulatory
states, and it calls for further investigation. Identifying the relationships between
situational variables and self-regulatory states would be critical to expand the
proposed model. The analysis for this future study has to posit the activated self-
regulatory state as the dependent variable and the situational variables as the
predictors of the self-regulatory state. The activation bias of each self-regulatory
state depending on various situational variables is important to understanding the
boundary condition of the proposed model. Even though the process is in place,
when the proper self-regulatory state is not activated, the process cannot be effective.
It is observed as if the model did not work, but it does not mean that the model is
wrong. It is simply the case that under certain situational variables, one or more
manipulations do not work, and the effect cannot be observed for those conditions(s).
For example, absolute budget level failed to show discriminant validity for
regulatory reference manipulation when it was high, and core attribute type, when
hedonic, interfered with fund focus manipulation.
123
R eproduced with perm ission o f the copyright owner. Further reproduction prohibited without perm ission.
R e f e r e n c e s
Atkinson, J. W. (1964), An Introduction to Motivation, Princeton, NJ: Van Nostrand.
Bakke, E. W. (1940), The Unemployed Worker, New Haven, CT: Yale University
Press, 142-143.
Baron, J. (1986), “Trade-offs Among Reasons for Action,” Journal o f The Theory o f
Social Behavior, 16 (July), 173-195.
_______ and M. Spranca (1997), “Protected Values,” Organizational Behavior and
Human Decision Processes, 70 (April), 1-16.
Bell, D. E. (1982), “Regret in Decision Making under Uncertainty,” Operations
Research, 330 (September/October), 961-981.
Brendl, C. M., E. T. Higgins, and K. M. Lemm (1995), “Sensitivity to Varying Gains
and Losses: The Role of Self-discrepancies and Event Framing,” Journal o f
Personality and Social Psychology, 69, 1028-1051.
Briesch, R. A., L. Krishunamurthi, T. Mazumdar and S. P. Raj (1997), “A
Comparative Analysis of Reference Price Models,” Journal o f Consumer
Research, 24, 202-214.
Brockner, J. and E. T. Higgins (2001), “Regulatory Focus Theory: Implications for
the Study of Emotions at Work,” Organizational Behavior and Human
Decision Processes, 86 (1), 35-66.
_______ and J. Z. Rubin (1985), Entrapment in Escalating Conflicts: A Social
Psychological Analysis, New York, NY: Springer.
Carver, C. S. and M. F. Scheier (1981), Attention and Self-regulation: A Control-
theory Approach to Human Behavior, New York, NY: Springer-Verlag.
_______ and________ (1990), “Principles of Self-Regulation: Action and Emotion,”
in Handbook o f Motivation and Cognition: Foundations o f Social Behavior
(Vol. 2), eds. E. T. Higgins and R. M. Sorrentino, New York, NY: Guilford
Press, 3-52.
_______ and T. L. White (1994), “Behavioral Inhibition, Behavioral Activation, and
Affective Responses to Impending Reward and Punishment: The BIS/BAS
scales,” Journal o f Personality and Social Psychology, 67, 319-333.
124
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Cacioppo, J. T. and G. G. Bemtson (1994), “Relationship between Attitudes and
Evaluative Space: A Critical Review, with Emphasis on The Separability of
Positive and Negative Substrates,” Psychological Bulletin, 115, 401-423.
_______ , W. L. Gardner, and G. G. Berntson (1997), “Beyond Bipolar
Conceptualizations and Measures: The Case of Attitudes and Evaluation
Space,” Personality and Social Psychology Review, 1 (1), 3-25.
_______ , R. E. Petty and C. F. Kao (1984), “The Efficient Assessment of Need for
Cognition,” Journal o f Personality Assessment, 48, 306-307.
Cristol, S. M. and P. Sealey (1996), “Replacement Marketing,” Marketing
Management, 5 (Spring), 12-22.
Crowe, E. and E. T. Higgins (1997), “Regulatory Focus and Strategic Inclinations:
Promotion and Prevention in Decision-making,” Organizational Behavior
and Human Decision Processes, 69, 117-132.
Dhar, R. and S. M. Nowlis (1999), “The Effect of Time Pressure on Consumer
Choice Deferral,” Journal o f Consumer Research, 25 (March), 369-384.
Dutton, J. E. and E. E. Jackson (1987), “Categorizing Strategic Issues: Links to
Organizational Action,” Academy o f Management Review, 12, 76-90.
Edwards, W. (1955), “The Prediction of Decisions among Bets,” Journal o f
Experimental Psychology, 51, 201-214.
Emery, F. E. (1970), “Some Psychological Aspects of Price,” in Pricing Strategy,
eds. B. Taylor and G. Willis, Princeton, NJ: Brandon Systems, 89-97.
Engel, J. and R. Blackwell, (1982), Consumer Behavior, Chicago, IL: Dryden.
Estes, W. K. (1944), “An Experimental Study of Punishment,” Psychological
Monographs, 57, No. 263.
Feather, N. T. (1982), “Actions in Relation to Expected Consequences: An Overview
of A Research Program,” in Expectations and Actions: Expectancy-value
Models in Psychology, ed. N. T. Feather, Hillsdale, NJ: ErlBaum, 53-95.
Festinger, L. (1957), A Theory o f Cognitive Dissonance, Evanston, IL: Row,
Peterson.
125
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Gray, J. A. (1982), The Neuropsychology o f Anxiety: An Inquiry into The Functions
o f The Septo-hippocampal System, New York, NY: Oxford University Press.
Heath, C. and J. B. Soil (1996), “Mental Budgeting and Consumer Decisions,”
Journal o f Consumer Research, 23 (June), 40-52.
Heider, F. (1958), The Psychology o f Interpersonal Relations, New York, NY:
Wiley.
Higgins, E. T. (1997), “Beyond Pleasure and Pain,” American Psychologist, 52
(December), 1280-1300.
_______ (1998), “From Expectancies to Worldviews: Regulatory Focus in
Socialization and Cognition, ” in Attribution and Social Interaction: The
Legacy o f Edward E. Jones, eds. J. M. Darley and J. Cooper, Washington,
DC: American Psychological Association, 243-309.
_______ (1999), “Promotion and Prevention as A Motivational Duality: Implications
for Evaluative Processes,” in Dual-process Theories in Social Psychology,
eds. S. Chaiken and Y. Trope, New York, NY: The Guilford Press, 503-525.
_______ , R. N. Bond, R. Klein and T. Strauman (1986), “Self-discrepancies and
Emotional Vulnerability: How Magnitude, Accessibility, and Type of
Discrepancy Influence Affect,” Journal o f Personality and Social
Psychology, 51,5-15.
_______ , J. Shah, and R. Friedman (1997), “Emotional Responses to Goal
Attainment: Strength of Regulatory Focus as Moderator,” Journal o f
Personality and Social Psychology, 72, 515-525.
Idson, L. D., N. Liberman and E. T. Higgins (2000), “Distinguishing Gains from
Nonlosses and Losses from Nongains: A Regulatory Focus Perspective on
Hedonic Intensity,” Journal o f Experimental Social Psychology, 36, 252-274.
Iyer, E. S. (1989), “Unplanned Purchasing: Knowledge of Shopping Environment
and Time Pressure,” Journal o f Retailing, 65 (1), 40-56.
Kahneman, D. and A. Tversky (1979), “Prospect Theory: An Analysis of Decision
under Risk,” Econometrica, 47, 263-291.
Kalwani, M. U., C. K. Yim, H. J. Rinne, and Y. Sujita (1990), “A Price Expectation
Model of Consumer Brand Choice,” Journal o f Marketing Research, 27
(August), 251-262.
126
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Keller, P. A. and L. G. Block (1996), “Increasing the Effectiveness of Fear Appeals:
The Effect of Arounsal and Elaboration,” Journal o f Consumer Research, 22
(March), 448-459A
Klein, N. H. and J. E. Oglethorpe (1987), “Cognitive Reference Points in Consumer
Decision Making,” Advances in Consumer Research, 14, 183-187.
Kristensen, H. and T. Garling (1997), “Determinants of Buyers’ Aspiration and
Reservation Price,” Journal o f Economic Psychology, 18, 469-486.
Landman, J. (1987), “Regret and Elation Following Action and Inaction,”
Personality and Social Psychology Bulletin, 13 (4), 524-536.
Lewin K. (1935), A Dynamic Theory o f Personality, New York, NY: McGraw-Hill.
Luce, M. F. J. W. Payne, and J. R. Bettman (1999), “Emotional Trade-off Difficulty
and Choice,” Journal o f Marketing Research, 36 (May), 143-159.
Markus, H. and P. Nurius (1986), “Possible Selves,” American Psychologist, 41,
954-969.
Monger, J. E. and R. A. Feinberg (1997), “Mode of Payment and Formation of
Reference Price,” Pricing Strategy & Practice, 5 (4), 142-147.
Monroe, K. B. (1973), “Buyers’ Subjective Perceptions of Price,” Journal o f
Marketing Research, 10 (February), 70-80.
_______ and S. M. Petroshius (1981), “Buyers’ Perception of Price: An Update of
The Evidence,” in Perspectives in Consumer Behavior, 3rd ed. Eds. H.
Kassarjian & T. S. Robertson, Glenview, IL: Scorr-Foresman, 43-55.
Rainwater, L., R. P. Coleman, and G. Handel (1959), “Behavioral Decision
Research: A Constructive Processing Perspective,” Annual Review o f
Psychology, 43, 87-131.
Rao, V. and D. A. Gautschi (1982), “The Role of Price in Individual Utility
Judgments: Development and Empirical Validation of Alternative Models,”
in Research in Marketing, ed. L. McAlister, Greenwich, CT: JAI Press.
Rohan, M. J. and M. P. Zanna (1998), “The ‘Products of Socialization’: A
Discussion of Self-Regulatory Strategies and Value Systems,” in Attribution
and Social Interaction: The Legacy o f Edward E. Jones, eds. J. M. Darley
and J. Cooper, Washington, DC: American Psychological Association, 279-
292.
127
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Roney, C. J. R., E. T. Higgins, and J. Shah (1995), “Goals and Framing: How
Outcome Focus Influences Motivation and Emotion,” Personality and Social
Psychology Bulletin, 21, 1151-1160.
Rook, D. W. (1987), “The Buying Impulse,” Journal o f Consumer Research, 4, 189-
199.
_______ and R. J. Fisher (1995), “Normative Influences on Impulsive Buying
Behavior,” Journal o f Consumer Research, 22, 305-313.
Russo, J. E. and P. H. Shoemaker (1989), Decision Types: The Ten Barriers to
Brilliant Decision Making and How To Overcome Them, New York:
Doubleday.
Scherer, R. M. (1980), Industrial Market Structure and Economic Performances (2n d
ed.), Chicago, IL: Rand McNally.
Schor, J. (2000) “The New Politics of Consumption,” in Do Americans Shop Too
Much?, ed. J. Schor, Boston, MA: Beacon Press, 3-33.
Sheppard, B. H., J. Hartwick, and P. R. Warshaw (1988), “The Theory of Reasoned
Action: A Meta-Analysis of Past Research with Recommendations for
Modifications and Future Research,” Journal o f Consumer Research, Dec.,
325-342.
Simonson, I. (1992), “Influences of Anticipating Decision Errors and Regret on
Purchase Timing and Choices between Brand Name and Price,” Journal o f
Consumer Research, 19 (June), 133-138.
Skinner, B. F. (1953), Science and Human Behavior, New York, NY: Macmillan.
Strauman, T. J. (1996), “Stability within the Self: A Longitudinal Study of the
Structural Implications of Self-discrepancy Theory,” Journal o f Personality
and Social Psychology, 71,1142-1153.
_______ and E. T. Higgins (1987), “Automatic Activation of Self-discrepancies and
Emotional Syndromes: When Cognitive Structures Influence Affect,”
Journal o f Personality and Social Psychology, 53, 1004-1014.
Tanner, J. F., J. B. Hunt, and D. R. Eppright (1991), “The Protection Motivation
Model: A Normative Model of Fear Appeals,” Journal o f Marketing, 55
(July), 36-45.
128
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Tetlock, P. E., R. S. Peterson, and J. S. Lemer (1996), “Revising The Value
Pluralism Model: Incorporating Social Content and Context Postulates,” in
The Psychology o f Values: The Ontario Symposium on Personality and
Social Psychology (Vol. 8), eds. C. Sleigman, J. M. Olson, and M. P. Zanna,
Mahwah, NJ: Lawrence Erlbaum, 25-51.
Thaler, R. (1980), “Toward A Positive Theory of Consumer Choice,” Journal o f
Economic Behavior & Organization, 1, 36-40.
Thompson, M. M., M. P. Zanna, and D. W. Griffin (1995), “Let’s Not Be Indifferent
about (Attitudinal) Ambivalence,” in Attitude Strength: Antecedents and
Consequences, eds. R. E. Petty and J. A. Krosnick, Hillsdale, NJ: Lawrence
Erlbaum, 361-386.
Thorndike, E. L. (1935), The Psychology o f Wants, Interests, and Attitudes, New
York, NY: Appleton-Century-Crofits.
Weiner, R. S. (1988), “Behavioral Perspectives on Pricing: Buyers’ Subjective
Perceptions of Price Revisited,” in Issues in Pricing, ed. T. M. Divinney,
Lexington, MA: Lexington, 35-57.
White, S. B. and M. A. Neale (1994), “The Role of Negotiator Aspirations and
Settlement Expectancies on Bargaining Outcomes,” Organizational Behavior
and Human Decision Processes, 57, 303-317.
Wilkie, W. L. and E. A. Pessemier (1983), “Issues in Marketing’s Use of
Multiattribute Models,” Journal o f Marketing Research, Nov., 428-441.
Whittier, T. E. (1994), “Eliciting Consumer Choice Heuristics: Sales
Representatives’ Persuasion Strategies,” Journal o f Personal Selling and
Sales Management, 14 (4), 41-53.
Zelizer, Vivana A. (1993), “Making Multiple Monies,” in Explorations in Economic
Sociology, ed. Richard Swedberg, New York, NY: Sage, 193-212.
129
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Developing business communication skills: Leveraging stage versus global processes of change in skills improvement approaches

PDF

An investigation of repeat visit behavior on the Internet: Models and applications

PDF

Displaced aggression as a function of target power

PDF

Inventory and pricing models for perishable products

PDF

Effects of redundancy in media coverage on nonprofessional investors' earnings forecasts

PDF

Attachment, accommodation, and the outcomes of romantic relationships

PDF

"How will I feel about it?": The asymmetric impact of affective misforecasting on consumer satisfaction

PDF

Customer satisfaction, customer bargaining power, and financial performance

PDF

Information systems capabilities and indirect relational value: Sustaining networked organizations

PDF

Information diffusion and the boundary of market efficiency

PDF

Biopharmaceutical strategic alliances: Interorganizational dynamics and factors influencing FDA regulatory outcomes

PDF

Does adding nonfinancial value drivers to a summary financial measure improve the learning and performance of managers?

PDF

Essays on technological evolution and financial returns to innovation

PDF

Diagnosing communication connections: Reaching underserved communities through existing communication ecologies

PDF

Foreign capital inflows and trade policy change: The automotive and consumer electronics industries in Brazil and China

PDF

A Positive Psychology goal intervention using strengths and values to enhance goals and increase well -being

PDF

Determining competent/resilient outcomes across multiple domains in sexually abused and nonabused females

PDF

Age and sex differences in levels of proliferation within the zebra finch telencephalic ventricular zone

PDF

Individual motivation loss in group settings: An exploratory study of the social -loafing phenomenon

PDF

Augmenting knowledge reuse using collaborative filtering systems

Asset Metadata

Creator Whang, Yun-Oh (author)

Core Title Consumer choice under budget constraint: Why consumers overspend

School Graduate School

Degree Doctor of Philosophy

Degree Program Business Administration

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag business administration, marketing,OAI-PMH Harvest,psychology, behavioral

Language English

Contributor Digitized by ProQuest (provenance)

Advisor [illegible] (committee chair), [illegible] (committee member), Cody, Michael J. (committee member), Folkes, Valerie (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-546464

Unique identifier UC11334798

Identifier 3094382.pdf (filename),usctheses-c16-546464 (legacy record id)

Legacy Identifier 3094382.pdf

Dmrecord 546464

Document Type Dissertation

Rights Whang, Yun-Oh

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA