University of Southern California Dissertations and Theses

Novelty versus familiarity: divergent effects of low predictability and low personal influence

(USC Thesis Other)

Novelty versus familiarity: divergent effects of low predictability and low personal influence

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Novelty versus Familiarity: Divergent Effects of Low Predictability and Low Personal Influence

by

Bora Min

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)

May 2018

i

Acknowledgements

This dissertation would not have come into the world without my champion, Dr. Norbert
Schwarz. Norbert, I sincerely appreciate your guidance, advice, and support during my Ph.D.
program. I feel extremely fortunate to work with you and be part of your lab where you
constantly remind me of why doing research keeps me inspired and enlightened. Thank you for
being my role model as a passionate and caring researcher as well as a decent human-being.
I would also like to express my gratitude to the dissertation chair, Dr. C.W. Park and my
dearest co-advisor Dr. Cheryl Wakslak. C.W., I would have not started at the USC Marshall,
neither have I completed this dissertation without your support for me. Thank you for sharing
your positive energy when I needed it most, and for being such a hilarious, energetic, and
encouraging dissertation chair and life coach. Cheryl, you are one my favorite people at USC. I
enjoyed every moment we spent working together and learned so much from your unbelievable
insight and amazing humility. You make everyone feel important, and you gave me so much of
your time and shared your wisdom so selflessly that I have become a much stronger researcher
and person.
I thank everyone in the marketing department for their advice and support throughout the
Ph.D. program. I would especially thank Anthony Dukes, Sivaramakrishnan Siddarth, Valerie
Folkes, and Kristin Diehl for their sincere care and kind support.
I am lucky to have gone through my Ph.D. program with my dear friends and colleagues.
I owe special thanks to my better half, Arianna Uhalde for being such a positive influence and
letting me share the ups and downs during tough times with her. He (Michael) Jia, Xiaoqian Yu,

ii

Yao Yao, Julia Kim, Victoria Kim, and Kevin Choi, thank you all for being such nice souls and
great friends.
Lastly, I would like to dedicate this dissertation to my beloved family. I can’t express my
gratitude enough for their unconditional love, care, and support during this process as well as
throughout my life. I thank and love you so much.

iii

TABLE OF CONTENTS
LIST OF TABLES ...................................................................................................................... v
LIST OF FIGURES .................................................................................................................... vi
ABSTRACT ................................................................................................................................. vii
CHAPTER 1. NOVELTY AS OPPORTUNITY AND RISK: OPPOSITE EFFECTS OF
LOW PREDICTABILITY AND LOW PERSONAL INFLUENCE ON CONSUMER
NOVELTY SEEKING................................................................................................................ 1
Introduction ............................................................................................................................. 1
Theoretical Development .................................................................................................. 3
Novelty Seeking and Control ............................................................................................ 3
Two Aspects of Control .................................................................................................... 4
Overview of Studies ................................................................................................................ 6
Study 1 .................................................................................................................................... 6
Method .............................................................................................................................. 7
Results and Discussion ..................................................................................................... 8
Study 2 .................................................................................................................................... 11
Method .............................................................................................................................. 11
Study 2A ............................................................................................................... 12
Study 2B................................................................................................................ 12
Results and Discussion ..................................................................................................... 12
Study 3 .................................................................................................................................... 17
Method .............................................................................................................................. 18
Results and Discussion ..................................................................................................... 18
Study 4A ................................................................................................................................. 25
Method .............................................................................................................................. 26
Results and Discussion ..................................................................................................... 27
Study 4B.................................................................................................................................. 27
Method .............................................................................................................................. 27

iv

Results and Discussion ..................................................................................................... 28
Single Paper Meta-Analysis .................................................................................................... 35
General Discussion ................................................................................................................. 36
Theoretical Implications ................................................................................................... 39
Practical Implications........................................................................................................ 40
CHAPTER 2. OPPOSITE EFFECTS OF LOW PREDICTABILITY AND LOW
PERSONAL INFLUENCE ON MERE EXPOSURE EFFECT ............................................. 43
Introduction ............................................................................................................................. 43
Study 1 .................................................................................................................................... 46
Method .............................................................................................................................. 46
Results and Discussion ..................................................................................................... 47
Study 2 .................................................................................................................................... 49
Method .............................................................................................................................. 49
Research and Discussion................................................................................................... 50
General Discussion ................................................................................................................. 52
FUTURE DIRECTION .............................................................................................................. 54
REFERENCES ............................................................................................................................ 56
APPENDIX 1: NEWS EXCERPTS SHOWN IN THE PREDICTABILITY (ABOVE)
AND UNPREDICTABILITY (BELOW) CONDITIONS IN CHAPTERS 1 AND 2 ........... 65
APPENDIX 2: SIX CHOCOLATE BARS SHOWN IN STUDY 1 THROUGH STUDY
4B IN CHAPTER 1 —THREE FAMILIAR PRODUCTS (ABOVE) AND THREE
UNFAMILIAR PRODUCTS (BELOW) .................................................................................. 67

v

LIST OF TABLES
TABLE 1. Familiarity Check in Study 2B, CHAPTER 1 ............................................................ 13
TABLE 2. Results of Novelty Seeking in Studies 2A and 2B, CHAPTER 1 .............................. 16
TABLE 3. Manipulation Check in Study 3, CHAPTER 1 ........................................................... 21
TABLE 4. Results of Novelty Seeking in Study 3, CHAPTER 1 ................................................ 22
TABLE 5. Results of Novelty Seeking in Study 4B, CHAPTER 1 ............................................. 33

vi

LIST OF FIGURES
FIGURE 1. Number of Novel Chocolate Bars Chosen Due To Predictability VS. Personal
Influence Manipulations in Studies 2A and 2B, CHAPTER 1 ..................................................... 15
FIGURE 2. Number of Novel Chocolate Bars Chosen Due To Predictability VS. Personal
Influence Manipulations in Study 3, CHAPTER 1 ....................................................................... 24
FIGURE 3. Number of Novel Chocolate Bars Chosen as a Function of Control Manipulations
and Frame Manipulations in Study 4B, CHAPTER 1 .................................................................. 34
FIGURE 4. A Single Paper Meta-Analysis Result, CHAPTER 1 ................................................ 36
FIGURE 5. Liking for Nonsense Words at Four Levels of Exposure Frequency as a Function of
Cause and Level of Control in Study 1, CHAPTER 2 .................................................................. 49
FIGURE 6. Liking for the Ideographs at Four Levels of Exposure Frequency as a Function of
Cause and Level of Control in Study 2, CHAPTER 2 .................................................................. 51

vii

ABSTRACT
Consumers constantly perceive, process, and evaluate products and services that are either
familiar or novel to them. Consistent with the concept of a “warm glow of familiarity” (i.e.,
“chicken soup for the soul” effect), prior research suggests that consumers who lack control over
their circumstances tend to stick to familiar options rather than try unfamiliar options. This
dissertation differentiates two distinct aspects of control: predictability (perceiving the world as
an unpredictable vs. predictable place) and personal influence (having vs. lacking control over
one’s own circumstances). These two aspects have long been considered equivalent in the
control literature and used interchangeably in control experiments. However, this dissertation
shows that they can have different implications for how consumers deal with novelty. From a
consumer’s perspective, novelty presents unknown opportunities as well as unknown risks and
these aspects are differentially highlighted by low predictability and low personal influence.
Specifically, low predictability suggests that the world is uncertain and things may change any
minute. Under these circumstances, consumers are motivated to learn about the full range of
options in preparation for possible surprises, which increases the exploration of novel options. In
contrast, low personal influence is associated with current difficulties and consumers hesitate to
expose themselves to additional risks. This results in a preference for the familiar. This
dissertation addresses these diverging aspects of novelty and control, and examines their
implications for consumer judgment and choice. In Chapter 1, I report six choice studies that
show higher novelty seeking under low (vs. high) predictability, but lower novelty seeking under
low (vs. high) personal influence, and identify the underlying processes. Note, however, that
choice does not necessarily imply liking—under low predictability, consumers may choose a
novel option because it seems useful given the circumstances, even though they would actually

viii

find the familiar option more appealing. Chapter 2 addresses this possibility. Drawing on the
classic mere exposure effect, I find from two studies that low personal influence enhances the
mere exposure effect, as reflected in increased liking of familiar (previously seen) stimuli,
whereas low predictability eliminates the classic mere exposure effect. Overall, this dissertation
identifies different aspects of control, shows their diverging impact on consumers’ preference for
familiar and novel products, and identifies managerial implications for more effective
promotions of both familiar products and novel products.

1

CHAPTER 1
“Novelty as Opportunity and Risk: Opposite Effects of Low Predictability
and Low Personal Influence on Consumer Novelty Seeking”

I n tr od u c ti on
Novel products can present unknown opportunities as well as unknown risks. Exploring
novel products can allow consumers to discover new benefits (Hirschman, 1980), but can also
confront them with considerable uncertainty and serious disappointment (Kalish, 1985;
Schiffman 1972). Since the propensities of consumers to adopt novel products play a critical role
in preference formation, product choice (Bianchi, 2006; Hekkert, Snelders, & Wieringen, 2003),
and the development of brand loyalty (Thompson & Sinha, 2008), understanding when
consumers are likely to see novel products positively or negatively carries considerable weight.
One stream of research on novelty seeking has emphasized the “warm glow” of
familiarity. As James (1890, 1961) and Titchener (1910) suggested, familiarity is associated with
security and safety and consumers tend to stick with familiar products when they confront
problematic situations, which can make familiar products comforting “chicken soup for the soul”
(Troisi & Gabriel, 2011). A key aspect of problematic situations is a lack of control over one’s
outcomes. A large body of research shows that lack of control is aversive and motivates
consumers to regain control (e.g., Janoff-Bulman, 1992; for a review see Landau, Kay, &
Whitson, 2015). Consistent with this general observation, past research showed that consumers
who lack personal control avoid novel products (Faraji-Rad, Melumad, & Johar, 2016).

2

Because it is difficult to exert personal influence in a world that is random and
unpredictable, theorists have commonly treated perceptions of influence over one’s outcomes
and perceptions of (un)predictability as interchangeable determinants of perceived control (e.g.,
Hafer & Bègue, 2005; Rutjens, van Harreveld, & van der Pligt, 2013). It is conceivable,
however, that these determinants have diverging effects when it comes to the exploration of
novel products and circumstances. Thinking of oneself as lacking what it takes to exert influence
over one’s situation may draw attention to the potential risks posed by novel products and the
challenges they may present to one’s own well-being and resources. Thinking of the world as
unpredictable, on the other hand, may highlight the need to be flexible and adaptive, drawing
attention to the potential benefits of being aware of what novel options may have to offer. Such
exploration may seem less rewarding in a world that is highly predictable and unlikely to pose
many surprises. If so, the commonly observed novelty avoidance may only emerge under
conditions of low perceived personal influence, whereas conditions of low perceived
predictability may be associated with high novelty seeking.
Across six studies, I found consistent support for these predictions. Throughout,
consumers who thought about an unpredictable world were more likely to choose novel products,
while those who thought about a lack of their own personal influence were less likely to choose
novel products. Moreover, explicitly framing the choice task in terms of opportunity versus risk
elicited preferences that paralleled the pattern observed under low predictability versus low
personal influence, as theoretically predicted. I first provide relevant conceptual background for
this work and then report the experimental findings.

3

Theoretical Development
Novelty Seeking and Control
Consumers constantly evaluate and make choices between previously adopted products
and novel products. While novel products, defined as products that are unfamiliar to an
individual consumer (Hirschman, 1980), can fulfill consumers’ innate need for novelty (Flavell,
1977), consumers oftentimes stick with familiar products. The tendency to alternate between
familiar and unfamiliar products has to do with the different benefits they offer to consumers.
Previously purchased familiar products can provide consumers with certainty of choice and keep
them from risks of investing in uncertainty (Park & Lessig, 1980). However, previously
unpurchased new products may offer consumers superior opportunities that might not exist in
familiar products (Zirger & Maidique, 1990). As a voluminous literature dedicated to exploring
consumers’ novelty-seeking behavior (e.g., Berning & Jacoby, 1974; Ostlund, 1974; Rogers,
1976) has demonstrated, consumer novelty seeking can be influenced by numerous personal and
situational factors; the former includes trait innovativeness (Im, Bayus, & Mason, 2003), need
for stimulation (Roehrich, 2004), and cognitive style (Midgley & Dowling, 1993), and the latter
includes self-congruence, perceived risk (Cowart, Fox, & Wilson, 2008), categorization
flexibility (Moreau, Markman, & Lehmann, 2001), and temporal frame (Grant & Tybout, 2008).
One variable that figures prominently in accounts for consumers’ preference for familiar
items is the warm glow of familiarity first discussed by James (1890, 1961) and Titchener (1910)
and later explored in many variants, from Zajonc’s (1968) mere exposure effect to Monin’s
(2003) warm glow heuristic. According to this account, familiarity is associated with feelings of
security, safety, and comfort, while a novel stimulus might be perceived as potentially risky.
Indeed, consumers’ preference for familiar stimuli, as reflected in the mere exposure effect

4

(Zajonc, 1968), is enhanced when the situation is experienced as problematic and attenuated
when the situation is experienced as safe and benign (De Vries et al., 2010; Gillebaart, Förster, &
Rotteveel, 2012). One of the difficulties consumers frequently experience is a lack of personal
influence. Consumers are inherently motivated to have mastery of the environment (Berry &
West, 1993). Due to a fundamental human motivation for control (Heckhausen & Schulz, 1995;
Kelly, 1955; White, 1959), situations in which they feel out of control can be threatening to the
self (Chaxel, 2016) and lead consumers to engage in various actions to compensate for the lack
of control (e.g., Cutright & Samper, 2014; Hamerman & Johar, 2013; Vanbergen & Laran,
2016). Given the warm glow of familiarity, one can assume that consumers who experience a
lack of control avoid exposure to novel products and situations that may make things worse.
Indeed, recent research by Faraji-Rad et al. (2016) showed that consumers who perceived low
personal influence were less likely to accept new products.
Two Aspects of Control
Because it is difficult to exert personal influence in a world that is random and
unpredictable, the psychology literature has treated perceptions of predictability and perceptions
of personal influence as closely intertwined. On the one hand, perceptions of randomness and
chaos are assumed to be distressing (Janoff-Bulman, 1992; Lerner, 1980) and contribute to the
motivation for personal influence (Kay et al., 2008), which can help in imposing some order on
one’s world. On the other hand, a lack of personal influence motivates a search for structure that
can reduce perceived randomness and unpredictability (Whitson, Galinsky, & Kay, 2015),
attributes that are antithetical to the opportunity of exerting some influence over one’s life. Given
these complementary relationships, it has long been assumed that perceptions of predictability
and being in control over one’s situation are interchangeable markers of an underlying

5

psychological sense of control. Consistent with this assumption, experimental studies have
proceeded as if perceptions of low personal influence and perceptions of low predictability were
interchangeable. For example, low control has been experimentally manipulated through
exposure to unpredictability (e.g., by describing a natural phenomenon as random; Kay et al.,
2014) or exposure to a lack of personal influence (e.g., by recalling a past incident in which one
did not have any influence over the situation; Faraji-Rad et al., 2016); sometimes, both
manipulations are used within the same article (e.g., Banfield, Nadolny, & Kay, 2013; Cutright,
Bettman, & Fitzsimons, 2013; Friesen et al., 2014). I agree that both manipulations are likely to
have the same effect under many conditions, as some of the cited examples illustrate. I suspect,
however, that they can have diverging effects on novelty seeking because they are likely to draw
differential attention to the opportunities offered and risks posed.
Novel products offer potential benefits as well as potential risks, burdens, and
frustrations. The potential risks, burdens and frustrations are likely to loom larger when
consumers feel that they may be unable to exert influence over their own circumstances. Indeed,
novel products may themselves reduce consumers’ sense of control over the environment
(Skinner, 1996) when their use involves changes to one’s prior routines or behaviors (Hoeffler,
2003). Hence, familiar products may seem more comforting and appealing than novel products
to those who experience a lack of personal influence. The available experimental evidence is
consistent with this prediction. Looking at acceptance of new products, Faraji-Rad and
colleagues (2016) observed lower novelty seeking under conditions of low rather than high
personal influence. Seeing the world as unpredictable, on the other hand, may highlight the need
to respond flexibly to novel developments, which may pose unexpected challenges but also may
offer unexpected opportunities. From this perspective, exploring novel options may be an

6

adaptive response to low predictability and help people to prepare for change. Exploring
unknown opportunities may seem less appealing when the world is seen as highly predictable
and likely to remain largely the same, thus posing neither a great adaptive challenge nor a great
promise of unexpected benefits. These considerations predict higher novelty seeking under
conditions of low rather than high predictability.
Overview of Studies
I test these diverging predictions in six studies. Study 1 provides initial evidence that
consumers seek more novelty when they see the world as unpredictable rather than predictable,
in contrast to what previous theorizing about the influence of low control would suggest.
Subsequently, Studies 2A and 2B directly juxtapose the two control manipulations. Replicating
Faraji-Rad and colleagues (2016), I find lower novelty seeking under conditions of low (vs. high)
personal influence; replicating Study 1, I find higher novelty seeking under conditions of
unpredictability (vs. predictability). The addition of a no-manipulation baseline condition in
Study 3 further highlights the diverging effects of different control manipulations. Thinking
about low personal influence decreases novelty seeking relative to baseline, whereas novelty
seeking does not differ from baseline after thinking about high personal influence. In contrast,
thinking about a predictable world decreases novelty seeking relative to baseline, whereas
novelty seeking does not differ from baseline after thinking about an unpredictable world. If the
diverging effects of low predictability versus low personal influence reflect a focus on
opportunity versus risk, explicitly framing the choice task in terms of opportunity should elicit
the pattern observed under low predictability condition, whereas explicitly framing the choice
task in terms of risk should elicit the pattern observed under low personal influence condition.

7

Studies 4A and 4B show that this is the case. A single-paper meta-analysis (McShane &
Böckenholt, 2017) indicates the robustness of the observed effects.
Study 1
Method
I recruited 180 students (female 45.6%, Mage = 19.95) who took a one hour long
marketing experiment for course credit, with the current study one of several unrelated studies
that they completed. Introduced as a study to explore how memory facilitates information
processing, my study first asked participants to read an alleged news excerpt (from Friesen et al.,
2014) about a recent conference attended by scientists studying causality. In the predictability
(vs. unpredictability) condition, the news excerpt indicated that scientists agreed that the world is
a stable and non-random (vs. unpredictable and random) place (see Appendix 1). Participants
were then asked to recall a life experience that can illustrate the arguments they read; that is, a
personal experience where they felt that the world is predictable (in the predictability condition)
or unpredictable (in the unpredictability condition). They briefly wrote about what happened and
how they felt in up to 100 words.
Next, a product choice task followed as an allegedly separate study. Participants were
presented with the images of six chocolate bars and were asked to make three choices to be
consumed over the next three days. The task was incentive compatible and they received the
chosen chocolate bars at the end of the experimental session. Participants’ choices were
unconstrained and they could choose any mix, from different chocolate bar for each day to the
same bar for each day. The six chocolate bars consisted of three domestic products, which were
familiar to participants (3 Musketeers, Milky Way, Twix), and three international products,

8

which were unfamiliar to participants (Double Decker, Moro, Lion Bar) (see Appendix 2). The
choice alternatives were presented in randomized order. Note that the number of familiar and
unfamiliar options is sufficient to allow participants to make all selections from only the familiar
or only the unfamiliar category, were they inclined to do so.
After making three choices, participants indicated how familiar they were with each
chocolate (1 = not familiar at all, 4 = moderately familiar, 7 = very much familiar) and whether
they had ever eaten it. I also measured the level of general familiarity with the product category
of chocolate bars (i.e., “How familiar does the product category of chocolate bars in general feel
to you?”, “How often do you buy chocolate bars?”, and “How often do you eat chocolate bars?”)
on 7-point Likert scales, with higher values indicating higher familiarity. Next, I assessed how
easy or difficult it was for participants to make choices and how much effort they put forth to
make choices on 7-point Likert scales, with higher values indicating higher effort. Finally, I did a
manipulation check (i.e., “You previously described a past experience after reading the excerpt.
Which statement most represented your opinion about the state of the world then?”; 0 = the
world is stable and predictable, 100 = the world is volatile and unpredictable) (from Friesen et
al., 2014) and their current mood (1 = I am in a very bad mood, 7 = I am in a very good mood).
Results and Discussion
Familiarity Check. The familiarity manipulation was successful. Most participants had
eaten the domestic products (3 Musketeers 78.6%, Milky Way 90.2%, Twix 91.9%), but not the
international products (Double Decker 3.5% Moro 3.5%, Lion bar 5.2%). Likewise, participants
found the domestic products familiar and the international products unfamiliar (M3musketeers =
5.68, SD = 1.96; MMilkyway = 6.32, SD = 1.28; MTwix = 6.63, SD = .99; significantly greater than 4,

9

the midpoint of the scale 4, p’s < .001; MDoubleDecker = 1.37, SD = 1.09; MMoro = 1.4, SD = 1.09;
MLionBar = 1.6, SD = 1.46; significantly smaller than 4, the midpoint of the scale, p’s < .001).
Seven participants reported that all products were familiar (equal to or greater than 4) or
unfamiliar (smaller than 4), and I excluded them from analyses since their choices are
uninformative with regard to novelty seeking. Twenty-nine participants were familiar with some
of the international products (giving them a rating equal to or greater than 4) and 28 participants
were unfamiliar with some of the domestic products (giving them a rating below 4). I took this
into account in the analyses of novelty seeking by recoding their choices to correspond with their
own product familiarity. This ensures that all analyses of novelty seeking are consistent with the
respective participant’s personal familiarity with the choice alternatives.
Additional analyses showed that the predictability and unpredictability conditions did not
differ in the percentage of participants who had eaten each product (p’s > .238) and the level of
familiarity with each product (p’s > .398). The conditions also did not differ on a three-item
index for general familiarity (α = .699) with the product category, time spent choosing, reported
effort invested in choice, participants’ age, gender ratio, and current mood. Reported difficulty of
choice was marginally greater in the predictability condition than in the unpredictability
condition (Mpredictability = 3.05 vs. Munpredictability = 2.59; t(171) = 1.87, p = .063). I explored the
possible influence of this difference by including it as a covariate in supplementary analyses
(reported in footnotes), which showed only negligible differences in the results.
Manipulation Check. Participants’ reported examples of experienced predictability
included: regular routines in everyday life, travels that unfolded as planned, and results of sports
games between a strong team and a weak team. Examples of experienced unpredictability
included: unexpected weather conditions (e.g., hurricane), delayed flights, and car accidents.

10

According to the manipulation check, participants who had thought about a unpredictable world
perceived the world as more volatile and unpredictable than those who had thought about a
predictable world (Mpredictability = 40.15, SD = 28.55 vs. Munpredictability = 77.99, SD = 16.59; t(171)
= -10.61, p < .001).
Novelty Seeking. I operationalized novelty seeking in three ways. First, how many
unfamiliar chocolate bars did the participant choose? I refer to this variable as novel chocolate
bars. Second, how many unfamiliar brands did these choices include? I refer to this variable as
novel brands. Both variables range from 0 to 3. A person who picks three unfamiliar bars, each
from a different brand, would receive a score of 3 on both measures; a person who picks the
same unfamiliar bar three times would receive a score of 3 on the novel chocolate bars measure
and a score of 1 on the novel brands measure. Third, a dichotomous measure reflects whether or
not a participant’s choices included any unfamiliar chocolate bars (coded 0 or 1). The means of
this measure indicate the proportion of participants who chose any novel product.
As predicted, participants who had thought about an unpredictable world chose more
unfamiliar chocolate bars than participants who had thought about a predictable world
(Mpredictability = .74, SD = .93 vs. Munpredictability = 1.15, SD = 1.13; t(171) = -2.64, p = .009, d = .4)
1

and their choices included more unfamiliar brands (Mpredictability = .64, SD = .78 vs. Munpredictability =
1.05, SD = 1; t(171) = -2.98, p = .003, d = .46). Overall, 62.4% of those who had thought about
unpredictability chose at least one unfamiliar chocolate bar, compared to 47.7% of those who
had thought about predictability (β = -.596, Wald = 3.71, p = .054).

1
Controlling for choice difficulty, I found that the results remained the same for both novel chocolate bars (t(170) =
3.12, p = .002) and novel brands (t(170) = 3.49, p = .001).

11

These results provide initial evidence that not every manipulation assumed to induce a
sense of “low control” reduces consumers’ novelty seeking, in contrast to what previous
literature would predict. To the contrary, thinking about unpredictability in the world elicited
higher novelty seeking than thinking about predictability. Studies 2A and 2B provide a
replication of this finding and directly juxtapose thinking about the predictability (vs.
unpredictability) of the world with thinking about a personal situation in which one did (vs. did
not) have control over the outcome.
Study 2
Studies 2A and 2B directly test whether thinking about unpredictability (vs.
predictability) leads to higher novelty seeking, as observed in Study 1, whereas thinking about
low (vs. high) personal influence leads to lower novelty seeking, as observed in prior literature
(Faraji-Rad et al., 2016). Both studies follow a 2 (type of control considered: predictability vs.
personal influence) x 2 (level of control: high vs. low) between-subjects design. The two studies
followed the same procedures and used the same materials, except that Study 2A was conducted
online and Study 2B in the laboratory.
Method
Study 2A. I recruited 298 participants (female 45%, Mage = 34.78) via Amazon
Mechanical Turk and paid them 40¢ for their participation. The manipulation of high versus low
predictability followed the same procedures as Study 1 and used the same materials. The
manipulation of high versus low control followed Faraji-Rad et al. (2016). Participants in the
high (low) personal influence condition recalled a situation in which something happened and
they felt a total (did not have any) control over the situation. That is, they described the situation

12

in which they felt a complete sense (lack) of control. In all conditions, participants were given
three minutes for their respective writing task.
All the procedures and measures were identical to Study 1, except that the choice task
was hypothetical (i.e., participants did not get actual chocolate bars) and the manipulation check
read, “You previously described a past experience in an earlier study. In doing so, to what extent
did you feel like life is predictable/controllable or unpredictable/uncontrollable?” (0 = I felt like
life is predictable and controllable, 100 = I felt like life is unpredictable and uncontrollable). I
also added an attention check item (i.e., “Please place the bar on 70” on a slider format question)
as in the subsequent studies that were conducted online. Given Study 1 showed no differences on
measures of ease/difficulty of choice and effort expended on choice, I dropped those questions.
Study 2B. Study 2B was identical to Study 2A in every aspect except that I obtained my
data in the laboratory with a college student population (N = 289, female 43.3%, Mage = 20.57) as
part of a series of unrelated marketing experiments.
Results and Discussion
In Study 2A, four participants reported they could not think of a specific event; they were
removed. Eight participants did not pass my attention check item so they were excluded, but all
the following results remain the same with them included. Among a total of 286 participants,
seven participants indicated that all the presented products were familiar or unfamiliar, thus they
were excluded from analyses. Forty-two participants were familiar with some of the international
products and 13 participants were unfamiliar with some of the domestic products. I recoded their
choices to correspond with their own product familiarity in the same way as Study 1.

13

In Study 2B, one participant failed to come up with a specific event so I excluded this
person. Among 288 participants, 10 participants considered all products as familiar or unfamiliar
so I excluded them from analyses. Forty-one participants were familiar with some of the
international products and 53 participants were unfamiliar with some of the domestic products. I
recoded their choices to correspond with their own product familiarity.
Familiarity Check. The familiarity manipulation was successful in both studies. In
Study 2A, the majority of participants had eaten the domestic chocolate bars (3 Musketeers
97.1%, Milky Way 96.4%, Twix 95%), whereas few participants had eaten the international ones
(Double Decker 6.5%, Moro 5.7%, Lion Bar 6.1%). Reported familiarity with the domestic
products (M3musketeers = 6.53, SD = 1.03; MMilkyway = 6.63, SD = .89; MTwix = 6.66, SD = 1.11) was
above the scale midpoint 4 (p’s < .001), whereas reported familiarity with the international ones
(MDoubleDecker = 1.37, SD = 1; MMoro = 1.32, SD = .96; MLionBar = 1.43, SD = 1.27) was below the
scale midpoint 4 (p’s < .001). Product familiarity, time spent choosing, participants’ age, gender
ratio, current mood did not differ across conditions (p’s > .104 in Study 2A; p’s > .2 in Study
2B). Study 2B mirrored the results (see Table 1).

Table 1. Familiarity Check in Study 2B
Domestic Chocolate Bars International Chocolate Bars
3 Musketeers Milky Way Twix Double Decker Moro Lion Bar
% Eaten
Before
79.1% 91.4% 96% 2.5% 2.2% 6.1%
Familiarity
Rating
5.41
(2.07)
6.07
(1.5)
6.51
(1.03)
1.3
(.96)
1.22
(.74)
1.51
(1.29)
Greater than 4 (p’s < .001) Smaller than 4 (p’s < .001)
Note.—Mean (Standard Deviation).

14

Manipulation Check. The reported examples of experiences of high personal influence
in Study 2A included: organizing an event, being in charge at work, driving, and paying bills.
Examples of low personal influence incidents included: someone’s death, 2016 US presidential
election, and troublesome interpersonal relationships. The manipulation check showed that
participants who had thought about unpredictability (vs. predictability) (Mpredictability = 34.44, SD
= 31.61 vs. Munpredictability = 75.54, SD = 23.65; F(1, 275) = 75.47, p < .001) and those who
thought about low (vs. high) personal influence (MHighPersInfluence = 42.58, SD = 28.09 vs.
MLowPersInfluence = 72.54, SD = 26.8; F(1, 275) = 42.13, p < .001) indicated that they felt life was
more unpredictable and uncontrollable. I found a marginally significant 2-way interaction (F(1,
275) = 2.84, p = .093) indicating that the difference between high versus low control was bigger
when control was manipulated by predictability (vs. personal influence).
Study 2B yielded comparable results. Participants who had thought about unpredictability
(vs. predictability) (Mpredictability = 52.75, SD = 23.66 vs. Munpredictability = 70.86, SD = 21.77; F(1,
274) = 18.75, p < .001) and those who had thought about low (vs. high) personal influence
(MHighPersInfluence = 56.2, SD = 26.54 vs. MLowPersInfluence = 67.65, SD = 23.85; F(1, 274) = 8.18, p
= .005) indicated that they felt life was more unpredictable and uncontrollable. There was no 2-
way interaction (F = 1.32, p > .25).
Novelty Seeking. My key prediction holds that thinking about an unpredictable world
elicits higher novelty seeking than thinking about a predictable world (as observed in Study 1),
whereas thinking about an episode of low personal influence elicits lower novelty seeking than
thinking about an episode of high personal influence (as observed by Faraji-Rad et al., 2016).
The resulting interaction of type of control and level of control was observed on all measures of
novelty seeking in both Studies 2A and 2B as shown in Table 2.

15

In Study 2A, participants who thought about an unpredictable (vs. predictable) world
chose more unfamiliar chocolate bars (Mpredictability = .55 vs. Munpredictability = .89; F(1, 275) = 3.92,
p = .049). In contrast, those who thought about an instance of low (vs. high) personal influence
chose fewer unfamiliar chocolate bars (MHighPersInfluence = 1.04 vs. MLowPersInfluence = .59; F(1, 275)
= 7.21, p = .008). These divergent effects are reflected in a significant 2-way interaction (F(1,
275) = 10.84, p = .001, η
2
p
= .038) (see Figure 1). Study 2B replicated the results as presented in
Table 2.

Figure 1. Number of Novel Chocolate Bars Chosen Due To Predictability VS. Personal
Influence Manipulations in Studies 2A and 2B

Also as predicted, participants who had thought about unpredictability (vs. predictability)
chose more unfamiliar brands (Mpredictability = .42 vs. Munpredictability = .76; F(1, 275) = 5.46, p
= .002). In contrast, those who had thought about low (vs. high) personal influence chose fewer
0
0.5
1
1.5
2
Predictability Personal Control
Study 2A: Number of Novel
Chocolate Bars Chosen
High Low
0
0.5
1
1.5
2
Predictability Personal Control
Study 2B: Number of Novel
Chocolate Bars Chosen
High Low
0.55
0.89
1.04
0.59 0.58
1.07
0.99
0.61
Personal Influence Personal Influence

16

unfamiliar brands (MHighPersInfluence = .85 vs. MLowPersInfluence = .57; F(1, 275) = 3.98, p = .047).
These diverging effects are again reflected in a significant 2-way interaction effect (F(1, 275) =
9.4, p = .002, η
2
p
= .033). I found the same results in Study 2B as shown in Table 2.
Table 2. Results of Novelty Seeking in Studies 2A and 2B
Novel Chocolate Bars
Predictability Manipulation Personal Influence Manipulation
2-way
Interaction
High
Pred.
Low
Pred.
F
(simple)
p High P
Influ.
Low P
Influ.
F
(simple)
p
S2A .55
(.93)
.89
(1.03)
3.92 .049 S2A 1.04
(1.12)
.59
(.94)
7.21 .008 F(1,275)=10.84
p = .001,
η
2
p
= .038
S2B .58
(.83)
1.07
(1.13)
9.2 .003 S2B .99
(.96)
.61
(.79)
5.86 .016 F(1,274)=14.94
p < .001,
η
2
p
= .052
Novel Brands
Predictability Manipulation Personal Influence Manipulation
2-way
Interaction
High
Pred.
Low
Pred.
F
(simple)
p High P
Influ.
Low P
Influ.
F
(simple)
p
S2A .42
(.69)
.76
(.88)
5.46 .002 S2A .85
(.91)
.57
(.89)
3.98 .047 F(1,275)=9.4
p = .002,
η
2
p
= .033
S2B .47
(.62)
.96
(1)
10.81 .001 S2B .96
(.94)
.59
(.77)
6.64 .011 F(1,274)=17.27
p < .001,
η
2
p
= .059
% Participants Choosing At Least One Novel Chocolate Bar
Predictability Manipulation Personal Influence Manipulation
2-way
Interaction
High
Pred.
Low
Pred.
χ
2
p High P
Influ.
Low P
Influ.
χ
2
p
S2A 32.8% 52.8% 5.5 .019 S2A 56.7% 34.2% 7.3 .007 β = 1.75
Wald = 12.49
p < .001
S2B 42.2% 59.4% 3.95 .047 S2B 57.9% 43.5% 3 .083 β = 1.28
Wald = 6.88
p = .009
Note.—Mean (Standard Deviation).

17

Finally, 52.8% of the participants who had thought about unpredictability chose at least
one unfamiliar chocolate bar, whereas 32.8% of the participants who had thought about
predictability did so (χ
2
(1) = 5.5, p = .019). In contrast, 34.2% of the participants who had
thought about low personal influence included at least one unfamiliar chocolate bar in their
choice, whereas 56.7% of those who thought about high personal influence did so (χ
2
(1) = 7.3, p
= .007). These diverging effects resulted in a significant 2-way interaction in a binary logistic
regression (β = 1.75, Wald = 12.49, p < .001). Study 2B replicated the results as presented in
Table 2.
In sum, Studies 2A and 2B provide consistent support for my hypothesis that perceptions
of unpredictability (vs. predictability) elicit higher novelty seeking, whereas perceptions of low
(vs. high) personal influence elicit lower novelty seeking, despite the common assumption that
low predictability and low personal influence are substitutable manipulations of a general control
construct. The observed impact of un/predictability replicates Study 1, whereas the observed
impact of personal influence replicates Faraji-Rad et al. (2016), based on the same high/low
personal influence manipulation they used.
Study 3
Studies 1 and 2 contrasted low and high levels of personal influence and predictability to
shed light on their respective influence on novelty seeking. They highlight that novelty seeking is
lower under low than high personal influence, but higher under low than high predictability.
However, these patterns do not allow me to evaluate which of these conditions differs from what
consumers would spontaneously do, without being exposed to one of these manipulations. Study
3 adds this information by including no-manipulation baseline conditions. The previous studies
did not include no-manipulation conditions because they are oftentimes ambiguous with regard

18

to theory testing—how much control and predictability participants perceive without any
manipulation will depend on whatever “manipulation” daily life just administered to them.
Method
A total of 369 Mechanical Turk workers (female 43.1%, Mage = 35.81) participated for a
compensation of 40¢. Study 3 followed the same design as Studies 2A and 2B, except that I
added two neutral conditions to provide a baseline. In one of these conditions, participants
completed the choice task without any preceding manipulation (choice only condition); in the
other condition, participants read a paragraph about vitamins and wrote a 2-3 sentence summary
(neutral writing condition). In addition, I had two manipulation check items in Study 3: “To
what extent did you feel like the world is predictable/controllable or
unpredictable/uncontrollable?” (0 = I felt like the world is predictable/controllable, 100 = I felt
like the world is unpredictable/uncontrollable) and “To what extent did you feel like life is
predictable/controllable or unpredictable/uncontrollable?” (0 = I felt like life is
predictable/controllable, 100 = I felt like life is unpredictable/uncontrollable). Further, I added a
single item to measure the level of anxiety among participants in the control manipulation
conditions (“To what extent did you feel anxious?”; 0 = not anxious at all, 100 = extremely
anxious). Previous literature suggested that it may be because consumers try to eliminate anxiety
that they engage in compensatory actions after being exposed to control threats (Kay et al.,
2009).
Results and Discussion
Five participants did not pass my attention check; however, their exclusion did not
change the results. Among 364 participants, 10 participants indicated that all chocolate bars were

19

either familiar or unfamiliar to them; they were excluded. Thirty-one participants were familiar
with some of the international products and eight participants were unfamiliar with some of the
domestic products. I recoded their choices to correspond with their own product familiarity in the
same way as earlier studies.
Familiarity Check. As in all previous studies, the majority of participants had eaten the
domestic products (3 Musketeers 96.3%, Milky Way 98.3%, Twix 94.9%), whereas few had
eaten the international products (Double Decker 4.5% Moro 4%, Lion bar 3.7%). Likewise,
reported familiarity with the domestic products (M3musketeers = 6.59, SD =.92; MMilkyway = 6.63, SD
= .92; MTwix = 6.71, SD = .78) was above the scale midpoint 4 (p’s < .001), whereas reported
familiarity with the international ones (MDoubleDecker = 1.27, SD = .84; MMoro = 1.21, SD = .81;
MLionBar = 1.25, SD = .95) was below the scale midpoint 4 (p’s < .001). Product familiarity, time
spent choosing, gender ratio, participants’ age, current mood did not differ between conditions
(p’s > .131).
Manipulation Check. Both control manipulations affected perceptions of
predictability/controllability of the world as well as one’s life in the expected direction (see
Table 3). Participants who had thought about unpredictability (vs. predictability) (Mpredictability =
35.18 vs. Munpredictability = 76.02; F(1, 220) = 85.4, p < .001) and those who had thought about low
(vs. high) personal influence (MHighPersInfluence = 42.57 vs. MLowPersInfluence = 75.34; F(1, 220) =
54.56, p < .001) reported the world was perceived as more unpredictable and uncontrollable.
There was no 2-way interaction (F = 1.66, p = .19). Similarly, participants who had thought
about unpredictability (vs. predictability) (Mpredictability = 41.02 vs. Munpredictability = 64.5; F(1, 220)
= 20.56, p < .001) and those who had thought about low (vs. high) personal influence
(MHighPersInfluence = 45.02 vs. MLowPersInfluence = 62.67; F(1, 220) = 11.52, p = .001) indicated that

20

life was perceived as more unpredictable and uncontrollable. There was no 2-way interaction (F
= .63, NS).
Participants in the choice only condition (M = 58.72, SD = 28.69) and neutral writing
condition (M = 63.48, SD = 33.03) were not different in perceptions of
predictability/controllability of the world, (t(128) = -.87, NS), therefore I combined the two
conditions to compare to control manipulation conditions (Mneutral = 61.25, SD = 31.04). Planned
contrasts showed that those who thought about an unpredictable world or low personal influence
perceived the world as more unpredictable/uncontrollable than those in the neutral conditions
(p’s < .001), and those who thought about a predictable world or high personal influence
perceived the world as more predictable/controllable than those in the neutral conditions (p’s
< .001). I found the consistent results in perceptions of personal influence. Participants in the
choice only condition (M = 50.84, SD = 32.57) and neutral writing condition (M = 55.57, SD =
33.1) were not different in perceptions of predictability/controllability of one’s life, (t(128) =
-.82, NS), hence I combined the two conditions to compare to control manipulation conditions
(Mneutral = 53.35, SD = 32.81). Planned contrasts showed that those who thought about an
unpredictable world or low personal influence perceived their life as more
unpredictable/uncontrollable than those in the neutral conditions (p’s < .043), and those who
thought about a predictable world or high personal influence perceived their life as more
predictable/controllable than those in the neutral conditions (p’s < .089) (see Table 3).

21

Table 3. Manipulation Check in Study 3

Note.—Mean (Standard Deviation).
*** p < .01, ** p < .05, * p < .1; ns: not significant.

Novelty Seeking. Study 3 replicates the findings of Studies 2A and 2B on all measures of
novelty seeking. Because the two baseline conditions did not differ in perceived
control/predictability, novelty seeking, product familiarity, time spent choosing, participants’
age, gender ratio, and current mood (p’s > .146), they were combined to form one neutral
condition.
Participants who had thought about an unpredictable (vs. predictable) world chose more
unfamiliar chocolate bars (Mpredictability = .45 vs. Munpredictability = .95; F(1, 220) = 8.72, p = .003). In
contrast, those who thought about an incident of low (vs. high) personal influence chose fewer
unfamiliar chocolate bars (MHighPersInfluence = .94 vs. MLowPersInfluence = .54; F(1, 220) = 5.54, p
Perception of Predictability of the World
Predictability Manipulation Personal Influence Manipulation
2-way
Interaction

F = 1.66
ns

High
Pred.
Low
Pred.
F
(simple)
High P
Influ.
Low P
Influ.
F
(simple)
35.18
(24.46)
76.02
(21.2)
85.4*** 42.57
(27.05)
75.34
(18.48)
54.56***
Neutral 61.25 (31.04)
t(349) -6.16*** 3.49*** -4.27*** 3.43***
Perception of Controllability of One’s Life
Predictability Manipulation Personal Influence Manipulation
2-way
Interaction

F = .63
ns

High
Pred.
Low
Pred.
F
(simple)
High P
Influ.
Low P
Influ.
F
(simple)
41.02
(26.62)
64.5
(29.63)
20.56*** 45.02
(26.72)
62.67
(26.54)
11.52**
Neutral 53.35 (32.81)
t(349) -2.61*** 2.36** -1.71* 2.04**

22

= .019). These diverging effects resulted in a significant 2-way interaction (F(1, 220) = 14.07, p
< .001, η
2
p
= .06). Comparisons with the baseline condition showed that participants in the
neutral conditions chose about as many unfamiliar chocolate bars (Mneutral = .96) as those who
had thought about an unpredictable world (t(349) = -.1, NS) or high personal influence (t(349) =
-.13, NS), and significantly more unfamiliar chocolate bars than those who had thought about a
predictable world (t(349) = -3.42, p = .001) or low personal influence (t(349) = -2.88, p = .004)
(see Table 4 and Figure 2).

Table 4. Results of Novelty Seeking in Study 3
Novel Brands
Predictability Manipulation Personal Influence Manipulation
2-way
Interaction
F =
13.5***
High
Pred.
Low
Pred.
F
(simple)
High P
Influ.
Low P
Influ.
F
(simple)
.45
(.69)
.89
(.95)
7.25*** .94
(1.05)
.52
(.81)
6.27**
Neutral .89 (.95)
t(349) -3.08*** .004
ns
.33
ns
-2.62***
% Participants Choosing At Least One Novel Chocolate Bar
Predictability Manipulation Personal Influence Manipulation
2-way
Interaction
β =
1.57***
High
Pred.
Low
Pred.
χ
2
High P
Influ.
Low P
Influ.
χ
2

35.7% 55.4% 4.36** 54.9% 36.1% 3.99**
Neutral 56.9%
t(349) -2.68*** -.19
ns
-.25
ns
-2.72***
Note.—Mean (Standard Deviation).
*** p < .01, ** p < .05, * p < .1; ns: not significant.

Novel Chocolate Bars
Predictability Manipulation Personal Influence Manipulation
2-way
Interaction
F =
14.07***
High
Pred.
Low
Pred.
F
(simple)
High P
Influ.
Low P
Influ.
F
(simple)
.45
(.69)
.95
(1)
8.72*** .94
(1.05)
.54
(.83)
5.54**
Neutral .96 (1.02)
t(349) -3.42*** -.1
ns
-.13
ns
-2.88***

23

Further replicating the preceding studies, those who had thought about unpredictability
(vs. predictability) chose more unfamiliar brands (Mpredictability = .45 vs. Munpredictability = .89; F(1,
220) = 7.25, p = .008), whereas those who had thought about low (vs. high) personal influence
chose less unfamiliar brands (MHighPersInfluence = .94 vs. MLowPersInfluence = .52; F(1, 220) = 6.27, p
= .013). These opposite effects yielded a significant 2-way interaction (F(1, 220) = 13.5, p
< .001, η
2
p
= .058). Participants in neutral conditions (Mneutral = .89) chose about as many
unfamiliar brands as those who had thought about an unpredictable world (t(349) = .004, NS) or
high personal influence (t(349) = .33, NS), but significantly more unfamiliar brands than those
who had thought about predictability (t(349) = -3.08, p = .002) or low personal influence (t(349)
= -2.62, p = .009).
Finally, 55.4% of the participants who had thought about an unpredictable world chose
any unfamiliar chocolate bar, whereas 35.7% of those who had thought about a predictable world
did so (χ
2
(1) = 4.36, p = .037). Conversely, 36.1% of those who had thought about low personal
influence chose any unfamiliar chocolate bar, whereas 54.9% of those who had thought about
high personal influence did so (χ
2
(1) = 3.99, p = .046). This again replicates the previous studies
(β = 1.57, Wald = 8.23, p = .004, 2-way interaction). The proportion of participants who chose
any novel chocolate bar did not differ between the neutral (56.9%) and unpredictability
conditions (t(349) = -.19, NS) or low personal influence condition (t(349) = -.25, NS). However,
those in neutral conditions were more likely to choose any unfamiliar chocolate bar than those
who had thought about predictability (t(349) = -2.68, p = .008) or low personal influence (t(349)
= -2.72, p = .007)
2
.

2
In addition, I found that the control manipulations influenced participants’ level of anxiety. Participants who
thought about unpredictability (vs. predictability) felt more anxious (M predictability = 30.36, SD = 29.35 vs.

24

Figure 2. Number of Novel Chocolate Bars Chosen Due To Predictability VS. Personal
Influence Manipulations in Study 3

In sum, Study 3 replicated the diverging effects of unpredictability and low personal
influence. Going beyond Studies 1 and 2, Study 3 allowed for comparisons with two conditions
in which participants’ thoughts about predictability and personal influence were not manipulated.
These comparisons showed that thinking about low personal influence or high predictability
decreased novelty seeking relative to baseline. Note, however, that comparisons to no-
manipulation conditions are fraught with more uncertainty than comparisons between
manipulated conditions when it comes to the influence of current subjective experiences and
accessible thoughts. Whatever people think or feel in the absence of an assigned manipulation is

M unpredictability = 54.59, SD = 31.93; F(1, 220) = 17.16, p < .001), as did those who thought about low (vs. high)
personal influence (M HighPersInfluence = 38.86, SD = 28.92 vs. M LowPersInfluence = 57.92, SD = 33.04; F(1, 220) = 10.53, p
= .001). However, anxiety did not mediate the effects of low control on novelty seeking in either case (95% CI
[-.128, .064] for the predictability conditions, 95% CI [-.116, .043] for the personal influence conditions). This
indicates that coping with anxiety derived from low control does not explain the higher (lower) novelty seeking due
to perceptions of unpredictability (low personal influence).
0
0.5
1
1.5
2
Predictability Personal Control Neutral
Study 3: Number of Novel
Chocolate Bars Chosen
High Low
Neutral
0.45
0.95
0.94
0.54
0.96
Personal Influence

25

a function of unknown contextual influences that can shift the location of a no-manipulation
condition from study to study. I therefore focus on the theoretically more informative
manipulated conditions.
Study 4A
The consistent findings of Studies 1 to 3 highlight that two manipulations of
psychological control that have often been considered substitutable operate in distinctively
different ways. From the perspective of a general control construct, perceiving low personal
influence or low predictability instantiates a low level of perceived control, whereas perceiving
high personal influence or high predictability instantiates a high level of perceived control.
Nevertheless, consumers are more likely to choose novel products when they perceive high (vs.
low) personal influence or think of the world as unpredictable (vs. predictable). I suggest that
this is the case because the two control manipulations draw attention to different aspects of novel
products, namely the novel risks they may entail versus the novel opportunities they may offer.
The risks loom larger when consumers are reminded of a situation in which they lacked personal
influence and things did not go their way. This makes the exploration of novel products less
appealing than when one feels in control. In contrast, potential novel opportunities loom larger
when consumers think about an unpredictable world, which may require familiarity with many
different options and rapid adaptation, benefits that may seem less appealing in a stable and
predictable world. This makes the exploration of novel products more likely when one sees the
world as unpredictable rather than predictable. This rationale implies that the observed effects
are driven by a differential focus on risks versus opportunities. If so, framing choice of novel
products as risks or opportunities should have less influence when manipulating personal
influence or un/predictability itself draws attention to risks or opportunities.

26

Study 4A tests whether explicitly framing the choice task in terms of risks versus
opportunities influences novelty seeking. Study 4B uses the framing developed in Study 4A to test
whether the framing manipulation attenuates the otherwise observed effects of personal influence
and un/predictability.
Method
I recruited 200 MTurk workers (female 60%, Mage = 37.58) for a compensation of 40¢.
All participants completed the writing task used in the neutral writing condition of Study 3. The
subsequent chocolate bar choice task was framed in three different ways. A no-framing version
was identical to the original instruction used in preceding studies (“In this part, we are
investigating the way consumers make choices. Please take a moment and try to imagine that you
will choose chocolate bars for the next three days…”). An opportunity-framing version added
two sentences about the value of new products as opportunities; “In this part, we are
investigating the way consumers make choices. Especially, we are looking at how consumers
deal with new products. Exploring novel products can bring consumers new opportunities they
might not have with existing products. New products can help consumers become better adapted
to the environment by providing the opportunity to diversity their knowledge and experiences.
Please take a moment and try to imagine…” A risk-framing version added two sentences about
the possible risk posed by novel products in the same place, “Especially, we are looking at how
consumers deal with new products. Exploring novel products can bring consumers risks they
might not have with existing products. New products can expose consumers to unpleasant
experiences and invite a waste of resources without offering much utility.” Participants were
randomly assigned to one of the three conditions.

27

Results and Discussion
As expected, opportunity framing increased novelty seeking and risk farming decreased
novelty seeking. Participants for whom novelty was framed as opportunity chose more
unfamiliar chocolate bars than those in the no framing condition (Mnoframing = 1 vs. Mopportunity =
1.33; t(197) = 1.97, p = .05) and more unfamiliar brands (Mnoframing = 1.03 vs. Mopportunity = 1.39;
t(197) = 2.05, p = .042). In contrast, participants for whom novelty was framed as risk chose
fewer unfamiliar chocolate bars than those in the no framing condition (Mnoframing = 1 vs. Mrisk
= .59; t(197) = -2.38, p = .018) and fewer unfamiliar brands (Mnoframing = 1.03 vs. Mrisk = .59;
t(197) = -2.51, p = .013). Finally, 59.3% of the participants in the no framing condition chose at
least one unfamiliar chocolate bar, whereas 76.4% in the opportunity framing condition (t(197) =
2.06, p = .04) and 39.1% in the risk framing condition did so (t(197) = -2.42, p = .016).
These findings confirm that explicit risk and opportunity framing can elicit responses that
parallel the patterns I observed with manipulations of low perceived personal influence and
unpredictability. In Study 4B, I cross these frames with the manipulations used in the preceding
experiments. I expect that explicitly framing novel products as opportunities (risks) enhances
(lowers) novelty seeking under all conditions, except those under which my rationale predicts
that the respective frame was already instantiated by the preceding manipulation of personal
influence or predictability.
Study 4B
Method
I recruited 705 MTurk workers (female 56.9%, Mage = 36.79) for a compensation of 40¢.
I administered a 2 (type of control considered: predictability vs. personal influence) x 2 (level of

28

control: high vs. low) x 3(novelty framing: no framing vs. opportunity framing vs. risk framing)
between-subjects design. The control manipulations are identical with Studies 2 and 3 (except as
noted below) and the framing manipulations are identical with Study 4A.
I made a minor change in the un/predictability manipulation, which I had adopted from
Friesen et al. (2014). The original instructions, which I used in Studies 1 through 3, included
references to ‘unpredictable (predictable)’ as well as ‘uncontrollable (controllable).’ I deleted the
references to controllability in Study 4B
3
.
Results and Discussion
Eleven participants reported that they could not think of a particular incident so they were
excluded. Six participants did not pass the attention check item so they were excluded, although
including them does not change any of the following results. Among 688 participants, 21
participants indicated they found all products were either familiar or unfamiliar, so they were
removed from analyses, leading to 667 valid responses. Sixty-nine participants were familiar
with some of the international products and 24 participants were unfamiliar with some of the
domestic products. I recoded their choices to correspond with their own product familiarity in the
same way as previous studies.
Familiarity Check. The familiarity manipulation was successful. Most participants had
experiences with the domestic products (Musketeers 95.4%, Milky Way 96.6%, Twix 95.2%),
whereas few did with the international products (Double Decker 5.8%, Moro 4%, Lion 4.8%).
The percentage of consumers who had eaten each product did not differ across conditions (p’s

3
Note that the simultaneous use of references to un/controllability and un/predictability works against the diverging
effects of the predictability versus personal influence manipulations observed in Studies 2 and 3.

29

> .392). Reported familiarity with domestic products (M3musketeers = 6.54, SD = 1.03; MMilkyway =
6.64, SD = .89; MTwix = 6.69, SD = .89) was above the scale midpoint 4 (p’s < .001), and that
with international ones (MDoubleDecker = 1.3, SD = .93; MMoro = 1.24, SD = .85; MLionBar = 1.35, SD
= 1.08) was below the scale midpoint 4 (p’s < .001). Reported familiarity with each product did
not differ across conditions (p’s > .175). Product familiarity, time spent choosing, participants’
age, gender ratio, and current mood did not differ across conditions (p’s > .334).
Manipulation Check. I ran a 2 (type of control: predictability vs. personal influence) x 2
(level of control: high vs. low) x 3 (novelty framing: no framing vs. opportunity framing vs. risk
framing) ANOVA. The different novelty frames did not influence participants’ perceptions of
predictability (F(2, 655) = .38, NS) or personal influence (F(2, 655) = 1.53, p > .2). Novelty
frames did not interact with type of control or level of control (F’s < 1.9, p’s > .14).
In terms of perceptions of predictability, there was a significant main effect of level of
control (F(1, 655) = 197.1, p < .001). Participants who had thought about unpredictability (vs.
predictability) saw the world as more unpredictable (Mpredictability = 38.02, SD = 29.47 vs.
Munpredictability = 78.3, SD = 20.15; F(1, 655) = 189.73, p < .001). Moreover, those who had
thought about low (vs. high) personal influence also saw the world as more unpredictable
(MHighPersInfluence = 53.12, SD = 28.29 vs. MLowPersInfluence = 68.83, SD = 23.87; F(1, 655) = 33.89, p
< .001). Despite a significant difference within each type of control, manipulating
unpredictability (vs. predictability) produced a greater difference in perceptions of a predictable
world than manipulating low (vs. high) personal influence, as indicated by a significant 2-way
interaction (F(1, 655) = 37.06, p < .001).
Similarly, participants who had thought about unpredictability (vs. predictability)
(Mpredictability = 52.75, SD = 23.66 vs. Munpredictability = 70.86, SD = 21.77; F(1, 655) = 43.23, p

30

< .001) reported they felt like life was more uncontrollable. Also, those who had thought about
low (vs. high) personal influence (MHighPersInfluence = 56.2, SD = 26.54 vs. MLowPersInfluence = 67.65,
SD = 23.85; F(1, 655) = 64.1, p < .001) indicated they felt like life was more uncontrollable.
Unlike perceptions of predictability, there was no 2-way interaction effect on perceived personal
influence (F(1, 655) = .44, p > .5). Manipulating unpredictability (vs. predictability) produced
comparable differences in perceptions of personal influence to manipulating low (vs. high)
personal influence.
Novelty Seeking. As shown in the no-framing columns of Table 5, the results of the no-
framing conditions replicated the findings of Studies 2 and 3 on the number of novel chocolate
bars chosen. An ANOVA indicated that participants who had thought about an unpredictable (vs.
predictable) world chose more novel chocolate bars (F(1, 222) = 4.35, p = .038), whereas
participants who had thought about low (vs. high) personal influence chose fewer novel
chocolate bars (F(1,222) = 7.43, p = .007), yielding a significant 2-way interaction (F(1, 222) =
11.31, p = .001, η
2
p
= .048).
Table 5 further shows how risk and opportunity framing impacts the previously observed
effects. For participants who had thought about a predictable world, opportunity framing
significantly increased the number of unfamiliar chocolate bars chosen (Mnoframing = .68 vs.
Mopportunity = 1.14; t(655) = -2.36, p = .019), whereas risk framing decreased it (Mnoframing = .68 vs.
Mrisk = .23 ; t(655) = 2.28, p = .023). However, among those who had thought about
unpredictability, opportunity framing did not increase the number of unfamiliar chocolate bars
chosen (Mnoframing = 1.09 vs. Mopportunity = 1.19; t(655) = -.54, p > .5), while risk framing decreased
it (Mnoframing = 1.09 vs. Mrisk = .56; t(655) = 2.63, p = .009) (see Figure 3). That is, explicitly

31

framing novelty as opportunity did not facilitate novelty seeking among those who thought of the
world as unpredictable, presumably because this frame was already instantiated as a function of
their earlier thoughts about an unpredictable world.
The personal influence conditions paralleled these findings. Among participants who had
thought about high personal influence, opportunity framing marginally significantly increased
the number of unfamiliar chocolate bars chosen (Mnoframing = 1.1 vs. Mopportunity = 1.41; t(655) = -
1.72, p = .086), but risk framing decreased it (Mnoframing = 1.1 vs. Mrisk = .64; t(655) = 2.53, p
= .012). However, among those who had thought about low personal influence, opportunity
framing increased the number of unfamiliar bars chosen (Mnoframing = .63 vs. Mopportunity = 1.15;
t(655) = -2.85, p = .004), whereas risk framing did not decrease it (Mnoframing = .63 vs. Mrisk
= .67 ; t(655) = -.19, NS). That is, explicitly framing novelty as risk did not impair novelty
seeking among those who had thought about low personal influence, presumably because this
frame was already instantiated as a function of their earlier thoughts about low personal
influence.
I found consistent results in the second novelty seeking measure (see Table 5).
Participants in no-framing conditions reproduced the previous 2-way interaction; those who had
thought about unpredictability (vs. predictability) chose more unfamiliar brands (F(1, 222) =
4.73, p = .031), but those who had thought about low (vs. high) personal influence chose fewer
unfamiliar brands (F(1, 222) = 5.19, p = .024; for the 2-way interaction; F(1, 222) = 9.83, p
= .002, η
2
p
= .042). Concerning framing effects, among participants who had thought about
predictability, opportunity framing significantly increased the number of unfamiliar brands
chosen (t(655) = -2.21, p = .028), whereas risk framing decreased it (t(655) = 2.17, p = .03). But,
among those who had thought about unpredictability, opportunity framing did not increase the

32

number of unfamiliar brands chosen (t(655) = -.06, NS), while risk framing decreased it (t(655) =
2.91, p = .004). The personal influence conditions generated the same findings. Among
participants who had thought about high personal influence, opportunity framing marginally
significantly increased the number of unfamiliar brands chosen (t(655) = -1.85, p = .065), but
risk framing decreased it (t(655) = 2.27, p = .024). However, among those who had thought
about low personal influence, opportunity framing increased the number of unfamiliar bars
chosen (t(655) = -2.53, p = .012), whereas risk framing did not decrease it (t(655) = .02, NS).
Finally, the same pattern is observed on the last novelty seeking measure. Among
participants in the no-framing conditions, those who had thought about unpredictability (vs.
predictability) were more likely to choose any unfamiliar cholate bars (χ
2
(1) = 4.18, p = .041),
but those who had thought about low (vs. high) personal influence were less likely to do so (χ
2
(1)
= 4.52, p = .034; (β = 1.6, Wald = 8.51, p = .004 for the 2-way interaction). Regarding framing
effects, among those who had thought about predictability, opportunity framing increased (t(655)
= -2.06, p = .04) but risk framing decreased (t(655) = 2.14, p = .032) the percentage of
participants who chose any unfamiliar chocolate bars.
However, among those who thought about unpredictability, opportunity framing did not
increase (t(655) = -.07, NS) while risk framing decreased (t(655) = 2.1, p = .036) the percentage
of participants who chose any unfamiliar chocolate bars. Similarly, among participants who had
thought about high personal influence, opportunity framing increased, despite statistical
nonsignificance, (t(655) = -1.59, p = .113) but risk framing decreased (t(655) = 2.28, p = .023)
the percentage of participants who chose any unfamiliar chocolate bars. But, among participants
who had thought about low personal influence, opportunity framing increased (t(655) = -2.05, p

33

= .041) but risk framing did not decrease (t(655) = -.1, NS) the percentage of participants who
chose any unfamiliar chocolate bars.
Table 5. Results of Novelty Seeking in Study 4B
Novel Chocolate Bars
Predictability Manipulation
High Predictability
Personal Influence Manipulation
High Personal Influence
No
Framing
Opportunity
Framing
Risk
Framing
t No
Framing
Opportunity
Framing
Risk
Framing
t
.68 (.92) 1.14 (1.14) -2.36** 1.1 (1.07) 1.41 (1.16) -1.72*
.68 (.92) .23 (.47) 2.28** 1.1 (1.07) .64 (.91) 2.53**
Low Predictability Low Personal Influence
No
Framing
Opportunity
Framing
Risk
Framing
t No
Framing
Opportunity
Framing
Risk
Framing
t
1.09 (1.06) 1.19 (1.16) -.54
ns
.63 (.85) 1.15 (1.17) -2.85***
1.09 (1.06) .56 (.81) 2.63*** .63 (.85) .67 (.95) -.19
ns

F(simple) =
4.35**
F(simple) =
7.43***

Novel Brands
Predictability Manipulation
High Predictability
Personal Influence Manipulation
High Personal Influence
No
Framing
Opportunity
Framing
Risk
Framing
t No
Framing
Opportunity
Framing
Risk
Framing
t
.62 (.84) 1.02 (1.01) -2.21** .98 (.97) 1.29 (1.08) -1.85*
.62 (.84) .23 (.47) 2.17** .98 (.97) .6 (.86) 2.27**
Low Predictability Low Personal Influence
No
Framing
Opportunity
Framing
Risk
Framing
t No
Framing
Opportunity
Framing
Risk
Framing
t
1.02 (1.01) 1.03 (1.02) -.06
ns
.62 (.83) 1.04 (1.08) -2.53**
1.02 (1.01) .49 (.64) 2.91*** .62 (.83) .61 (.88) .02
ns

F(simple) =
4.73**
F(simple)=
5.19**

% Participants Choosing At Least One Novel Chocolate Bar
Predictability Manipulation
High Predictability
Personal Influence
Manipulation
High Personal Influence
No
Framing
Opportunity
Framing
Risk
Framing
t No
Framing
Opportunity
Framing
Risk
Framing
t
41.5% 61.2% -2.06** 60% 74.1% -1.59
ns

41.5% 20.8% 2.14** 60% 39.7% 2.28**
Low Predictability Low Personal Influence
No
Framing
Opportunity
Framing
Risk
Framing
t No
Framing
Opportunity
Framing
Risk
Framing
t
62.2% 62.9% -.07
ns
41.2% 59.3% -2.05**
62.2% 41.8% 2.1** 41.2% 40.4% .1
ns

χ
2
= 4.18** χ
2
= 4.52**
Note.—Mean (Standard Deviation).
*** p < .01, ** p < .05, * p < .1; ns: not significant.

34

In sum, I found that explicitly framing novel products as opportunities (risks) enhanced
(lowered) novelty seeking under all conditions, except those under which my rationale predicts
that the respective frame was already instantiated by the preceding manipulation. Specifically,
opportunity framing did not increase novelty seeking among participants who had thought about
an unpredictable world, which I assume to draw attention to opportunities, whereas risk framing
did not decrease novelty seeking among those who had thought about low personal influence,
which I assume to draw attention to risk. In combination, these findings support my rationale that
thinking about unpredictability and thinking about a lack of personal influence draw attention to
different aspects of novelty, namely the unknown opportunities and unknown risks associated
with unfamiliar products.

Figure 3. Number of Novel Chocolate Bars Chosen as a Function of Control Manipulations
and Frame Manipulations in Study 4B

0
0.5
1
1.5
2
Predictability Unpredictability High Personal Control Low Personal Control
Study 4. Number of Novel Chocolate Bars Chosen
No Framing Opportunity Framing Risk Framing
0.68
1.14
0.23
1.09
1.19
0.56
1.1
1.41
0.64 0.63
1.15
0.67
High Predictability Low Predictability High Personal Infl. Low Personal Infl.

35

Single Paper Meta-Analysis
I conducted a single paper meta-analysis (SPM) of five studies including manipulations
of predictability or personal influence, following the methodology devised by McShane and
Böckenholt (2017). An SPM of my studies estimates the simple effect of thinking about
unpredictability versus predictability as .43 (SE .04; Studies 1, 2A, 2B, 3, and 4B) and the simple
effect of thinking about low versus high personal influence as -.4 (SE .04; Studies 2A, 2B, 3 and
4B). Based on Studies 2A, 2B, 3, and 4B the interaction effect was estimated at -.83 (SE .06). I
2

was estimated at 41% (95% CI: 0%-68%), suggesting that heterogeneity is low but could range
from zero to medium. This estimate, along with the visual convergence of effects shown in
Figure 4 indicates their robustness across studies.

36

Figure 4. A Single Paper Meta-Analysis Result

Note.— Effect estimates are given by the squares for single-study estimates and the vertical bars
for SPM estimate; 50% and 95% intervals are given by the thick and thin lines, respectively. The
average sample size per condition in each study is given by the size of the squares.

General Discussion
Past research has suggested that perceptions of a predictable world, with a clear order and
structure, and feelings of personal influence are two substitutable aspects of control (Krantz,
High vs. Low
Predictability
High vs. Low
Personal Influence
Interaction
Effect
Effect Estimate
Study 1
Study 2A
Study 3
Study 2B
Study 4B

37

1998; Kruglanski & Webster, 1996). Consistent with this assumption, my own data indicate that
thinking about an unpredictable world decreases perceived predictability as well as perceived
personal influence, just as thinking about a lack of personal influence decreases perceived
personal influence as well as perceived predictability (Study 4B). Nevertheless, these different
instantiations of a global “control” construct may draw attention to different aspects of
unfamiliar products, with important implications for consumers’ novelty seeking. As Hirschman
(1980), Kalish (1985) and other consumer researchers noted, novel products confront consumers
with unknown risks as well as unknown opportunities. Not surprisingly, one may expect that
focusing on the potential risks impairs novelty seeking, whereas focusing on the potential
benefits increases novelty seeking. The framing manipulations of Studies 4A and 4B explicitly
test this implication and provide compelling support.
From this perspective, the observation that a lack of personal influence elicits lower
novelty seeking (as shown in the current Studies 2, 3, and 4B, as well as Faraji-Rad et al., 2016)
suggests that low personal influence draws attention to the potential risks imposed by unfamiliar
products. When consumers feel that they have little influence over their situation, they stick to
the familiar and avoid exposure to unknown risks that they may not be able to control. Consistent
with this analysis, explicitly framing the choice task in terms of risk reduced novelty seeking in
all experimental conditions of Study 4B except the low personal influence condition.
Presumably, risk was already on the minds of these participants, thus rendering the framing
manipulation redundant with the effect of the preceding control manipulation. Further supporting
this rationale, the negative impact of low (vs. high) personal influence was eliminated when the
choice task was explicitly framed in terms of the opportunities novel products may afford.

38

The observed pattern was reversed when consumers’ sense of control was manipulated by
thinking about the predictable versus unpredictable nature of the world. Although thinking about
an unpredictable world decreased participants’ sense of personal influence (as indicated by the
manipulation check of Study 4B), these consumers chose more novel products and brands than
consumers who thought about a predictable world (Studies 1, 2, 3, and 4B). This presumably
reflects that an unpredictable world entails a need to flexibly respond to rapid changes, which is
facilitated by being aware of one’s options. This awareness is less relevant when the world is
seen as predictable and unlikely to undergo rapid change. Consistent with this analysis, framing
the choice task in terms of opportunities increased novelty seeking in all conditions of Study 4B
except the high unpredictability condition. Presumably, exploring opportunities was already on
the minds of these participants, thus rendering the framing manipulation redundant with the
effect of the preceding unpredictability manipulation. Further supporting this rationale, the
positive impact of unpredictability (vs. predictability) was eliminated when the choice task was
explicitly framed in terms of the risks novel products may impose.
Lastly, it is worth noting the relationship between novelty seeking and variety seeking
(Menon & Kahn, 1995). On the one hand, consumption of a novel product always increases a
consumer’s variety experience; on the other hand, novelty is not needed to satisfy a desire for
variety, which can be met by frequent switching among familiar products. In all of the present
studies, the number of familiar and novel choice alternatives was sufficient to allow for three
distinct choices by drawing solely on novel products, solely on familiar products or on a mix of
both. To explore variety seeking, I conducted supplementary analyses of how many different
products participants chose, independent of whether the products were novel or familiar. These
analyses showed that variety seeking was not affected by control manipulations in any of my

39

studies. This also implies that the observed diverging effects of low personal influence and
unpredictability are specific to novelty seeking and do not apply to stimulation seeking in
general, consistent with my theoretical rationale.
Theoretical Implications
Research in psychology as well as consumer behavior typically conceptualizes
individuals’ sense of control as entailing both; a sense of personal influence over one’s own
situation and a sense that the world in general is predictable (Antonovsky, 1979; Kay et al.,
2009). This rationale justifies treating manipulations of personal influence and manipulations of
predictability as mutually substitutable manipulations of the general control construct (e.g., Kay
et al., 2008). Consistent with this rationale, my own data show that either manipulation affects
both measures (Study 4B). However, a given subjective experience—such as low control—is
insufficient to predict a person’s behavior without taking contextual variables into account.
Merely knowing that a person experiences fear, for example, does not allow me to predict
whether this person will start running, hit the brakes, grab a fire extinguisher, sell her stocks, or
hurry to the emergency room. To make such predictions, I need to understand what the person
perceives as the source of her fear and as plausible responses, given the constraints and
affordances of her current situation (for a discussion, see Schwarz & Clore, 1996). The same
applies to people’s sense of influence. Even when different manipulations induce comparable
levels of perceived personal influence and predictability, perceiving different sources of threat is
likely to draw attention to different challenges, which may call for different responses. Whether
this results in differences in observable behaviors depends on the person’s goals and the
constraints and affordances of the situation.

40

From this perspective, novelty seeking is particularly likely to show differential effects of
different control manipulations. Novel products come with unknown opportunities and risks and
which of these aspects looms larger will, in part, depend on what is on consumers’ mind at the
time of choice. Hence, my findings do not imply that different control manipulations will result
in differential effects on all dependent variables; nor do they imply that consumers cannot
compensate for a perceived lack of personal influence with heightened efforts to see the world as
a clearly structured and predictable place, or vice versa (e.g., Cutright et al., 2013). My findings
do highlight, however, that predicting the behavioral effects of perceived control requires close
attention to the challenge conveyed by the specific control experience, the range of responses
that may address the respective challenge, and the opportunity to engage in one of these
responses afforded by the situation. My analysis of novelty seeking illuminates some of these
complexities and future research may fruitfully address the malleability of control responses in
other domains.
Practical Implications
The current findings show that perceived unpredictability (vs. predictability) stimulates
higher novelty seeking. These findings go hand in hand with a current practice in the
marketplace as well as a consumer behavior trend. Unexpected environmental change in the
marketplace requires firms and organizations to adopt innovations quickly to retain their
competitive advantages (Damanpour & Gopalakrishnan, 1998). In particular, in today’s
increasingly global marketplace, environmental uncertainty and unpredictable change prompt
greater emphasis on innovation of product and marketing process (Bourgeois, 1985; Miller &
Friesen, 1983). Moreover, from individual consumers’ perspective, unpredictable change in
one’s life environment (e.g., marital, residential, or occupational status) inspires adjustment of

41

their brand preferences and greater endorsement of novel options (Andreasen, 1984). This is also
in line with Wood (2010) that showed that consumers who experience times of change or
upheaval are more likely to choose novel options, in stark contrast to their own lay theory which
predicts a higher consumption of comfort food. Echoing this finding, my results suggest that
marketers may benefit from promoting novel brands and products to consumers who are
experiencing new life changes (e.g., moving to a new town). In addition, it may be worthwhile to
consider macro-level changes in economic, societal, political, or environmental conditions to
decide the most effective timing of marketing actions. Since the timing of marketing actions is
crucial for their success (Assmus, Farley, & Lehmann, 1984), marketers may benefit from
analyzing the degree of stability, predictability, and orderliness that consumers in the target
market are likely to experience. These possibilities provide fruitful avenues for future research.
On the other hand, experiencing low (vs. high) personal influence can lead to lower
novelty seeking. As Faraji-Rad et al. (2016) pointed out, consumer segments that have a higher
desire for personal influence (e.g., political conservatives), or consumers in a culture that highly
values personal influence (e.g., Indian market), would be more reluctant to accept novel
products. In environments that entail frequent experiences of personal influence deprivation,
marketers may be well advised to invest in traditional and well-known bestsellers. As observed
in numerous studies, social position (i.e., employment grade) and gender interact with the
environment at home and work to affect the level of experienced control deprivation (Griffin et
al., 2002), often resulting in depression and anxiety (Folkman, 1984). For example, women in the
lowest or middle employment grades, who also experience little personal influence at home or
work, are most likely to experience depression. For such consumer segments a classic well-
known product can be more appealing and comforting. It thus would be more beneficial for

42

marketers to downsize the investment in novel or less known products when targeting these
consumers.
Moreover, marketers can selectively decide the way they frame their products depending
on the familiarity of the products. Marketers may emphasize the newness of their products more
effectively by generating a context that illuminates the unpredictable aspect of the world (e.g., a
copy of ‘Expect the unexpected’ presented with a series of absurd images for advertising
innovativeness of Intrust Bank Arena; a sandal placed instead of the third cut in a 4-cut cartoon
depicting an unpredictable situation for promoting new shoes of Rider Shoes). In contrast,
leading brands would be better off if they featured their capability of supporting consumers’ own
choices to corroborate their competitive position in the market. For example, Burger King has
used ‘Have it your way’ as their motto for the past three decades. They declare “You have the
right to have what you want, exactly when you want it...” and this can successfully associate the
familiarity of their brand with consumer needs for control over their choice and life.

43

CHAPTER 2
“Opposite effects of Low Predictability and Low Personal Influence on Mere Exposure Effect”

Introduction
4

The desire for control figures prominently in many theories of human behavior and
emotion (for reviews, see Heckhausen & Schulz, 1995; Kay, Sullivan, & Landau, 2015; Morling
& Evered, 2006; Skinner, 1996). Because it is difficult to control one’s situation in a world that
is random and unpredictable (Janoff-Bulman, 1992; Lerner, 1980), the control literature has long
considered perceptions of predictability (i.e., order and structure in the world) and perceptions of
the core components of psychological control. It assumes that “large-scale events (e.g., natural
disasters, economic crises, wars, or terrorism), but also more personal events and incidents (e.g.,
sudden illness or loss of one’s job), can threaten perceptions of personal influence as well as
beliefs in an orderly and predictable world” (Rutjens, van Harreveld, & van der Pligt, 2013, p.
253). Manipulations of personal influence and predictability have been employed
interchangeably in experiments (e.g., Banfield, 2011; Cutright, Bettman, & Fitzsimons, 2013;
Friesen, Kay, Eibach, & Galinsky, 2014). Consistent with the assumption that both contribute to
a sense of control, perceptions of low predictability increase the motivation to exert personal
influence (e.g., Kay, Gaucher, Napier, Callan, & Laurin, 2008) and a lack of personal influence
increases the motivation to search for predictability (e.g., Whitson, Galinsky, & Kay, 2015).
Nevertheless, these two components of control may differ in ways that have important

4
Chapter 1 and 2 are two separate manuscripts under review at two different journals. For this reason, the
introduction of Chapter 2 is somewhat repetitive.

44

implications for a key domain of human behavior, namely people’s willingness to explore novel
options and their desire to stick to the familiar.
Low personal influence is associated with current difficulties—something does not go as
desired. Previous research (e.g., Horswill & McKenna, 1999; Maner & Schmidt, 2006; Skinner,
1996) suggests that consumers hesitate to expose themselves to additional risks under such
conditions—when given a choice, they prefer the familiar over the novel, as observed in studies
of product choice (Faraji-Rad, Melumad, & Johar, 2016; Min & Schwarz, 2017
5
). In contrast, the
perception that the world is unpredictable implies that things may change any moment. Under
such circumstances, it may be useful to learn about the full range of possibilities to be prepared
for what may come. If so, low predictability may increase people’s willingness to expose
themselves to novelty. This rationale predicts diverging effects of low personal influence and
low predictability in the domain of novelty seeking, in contrast to the usual equation of both
variables as instantiations of the control construct.
I found support for this prediction in an incentive compatible product choice paradigm
(Min & Schwarz, 2017). Participants who had been induced to see the world as unpredictable
selected more unfamiliar products than participants who had been induced to see the world as
predictable. In contrast, participants who had been induced to perceive low personal influence
selected fewer unfamiliar products than participants who had been induced to perceive high
personal influence, consistent with earlier research (Faraji-Rad et al., 2016). These diverging
patterns indicate that control is not a unitary variable—low predictability and low personal

5
It refers to Chapter 1 in this dissertation.

45

influence have opposite effects on people’s preference for novel versus familiar choice
alternatives.
However, that choice does not necessarily reflect liking. Participants who saw the world
as unpredictable may have preferred a familiar product while nevertheless exploring an
unfamiliar one to broaden their horizons in preparation for an unpredictable world. Hence,
participants’ differential choices may not reflect a change in the “warm glow” (Titchener, 1910)
of familiarity but a change in strategic behavior that overrides the temptations of the familiar in
favor of an adaptive exploration of the unknown. To address this ambiguity, the present chapter
draws on the most basic procedure used to assess preference for the familiar; Zajonc’s (1968)
mere exposure paradigm.
Repeated exposure to a stimulus increases liking of the stimulus, even in the absence of
any reinforcement (for reviews, see Bornstein, 1989; Montoya, Horton, Vevea, Citkowicz, &
Lauber, 2017). This effect is more pronounced when the person sees the situation as problematic,
e.g., due to an incidental sad mood (De Vries, Holland, Chenier, Starr, & Winkielman, 2010) or
induced prevention focus (Gillebaart, Förster, & Rotteveel, 2012). From the perspective of
psychological control theory, any state of low control is considered problematic, independent of
whether it has been induced through a lack of predictability or a lack of personal influence. If so,
both manipulations of low control should increase liking of the familiar. In contrast, my choice
findings (Min & Schwarz, 2017) suggest that low predictability motivates people to explore
unknown opportunities, whereas low personal influence discourages people from exploring
novel options that may present unwanted risks. I therefore expect to observe a mere exposure
effect under low personal influence, but not under low predictability. Two studies test these

46

predictions by exposing participants to nonsense stimuli at different frequency levels. The
institutional board review (IRB)/ethics committee approved both studies (UP-16-00518).
Study 1
Method
College students (N = 191; mean age 20.1; 61.8% female) were randomly assigned to the
conditions of a 2 (type of control: personal influence vs. predictability) x 2 (level of control: high
vs. low) x 4 (repetition: 0, 5, 15, 40 times)-factorial design, with the last factor manipulated
within participants. The N was determined by the number of participants available in the pool.
An alleged attention task exposed participants to three of four nonsense words (vloo, thoi,
crou, spei), which were selected on the basis of pretest data to be similarly positive and easy to
pronounce. On each trial, selected words popped up in three different positions (left, center,
right) on the computer monitor and participants marked all positions in which a word was shown.
I administered two orders; in the first (second) order, spei (crou) was repeated 40 times, thoi
(vloo) 15 times, vloo (thoi) 5 times, and crou (spei) 0 time, all in random order for a total of 20
trials. The exposure frequencies were adopted from Gillebaart et al. (2012).
Next, participants completed a reading-and-writing task that served as a manipulation of
control. In the (un)predictability conditions, they read an alleged news excerpt (from Friesen et
al., 2014) about a recent conference of scientists studying causality. In the predictability
(unpredictability) condition, the excerpt indicated that scientists agreed that the world is a
predictable and non-random (unpredictable and random) place (see Appendix 1). Participants
were asked to recall and describe a personal experience where they felt that the world is
predictable (unpredictable). In the high (low) personal influence condition participants recalled a

47

situation in which something happened and they felt they had total control (no control) over the
situation (adopted from Faraji-Rad et al., 2016). In all conditions, participants were given three
minutes for their respective writing task.
Finally, participants rated, in random order, how much they like each of the four
nonsense words (1 = not at all, 7 = very much), how familiar each word felt (1 = not at all, 7 =
extremely), and whether they recalled having seen it in the attention task (1 = 100% sure I have
not seen it, 7 = 100% sure I have seen it). Participants also responded to two manipulation
checks (“You previously described a past experience in an earlier study. In doing so, to what
extent did you feel like the world is predictable or unpredictable?” 0 = predictable, 100 =
unpredictable; “In doing so, to what extent did you feel like life is controllable or
uncontrollable?” 0 = controllable, 100 = uncontrollable) and provided demographic information.
Results and Discussion
Confirming the control manipulations, participants saw the world as less predictable after
thinking about unpredictability (M = 72.13) than predictability (M = 31.96; F(1,187) = 119.97, p
< .001) and as less controllable after thinking about low (M = 67.9) than high personal influence
(M = 41.94; F(1,187) = 39.56, p < .001).
Replicating the classic mere exposure effect, participants overall liked a nonsense word
more, the more often they had seen it, F(3, 561) = 4.02, p = .008, η
2
p
= .021. This main effect of
exposure frequency was qualified by the predicted 3-way interaction with type and level of
control, F(3, 561) = 4.59, p = .003, η
2
p
= .024. As shown in Figure 5, liking increased with the
frequency of exposure under high predictability, F(3, 561) = 4.64, p = .003, η
2
p
= .024, for the
simple effect, but not under low predictability, F(3, 561) = .02, NS, for the simple effect. This

48

pattern is reflected in a simple interaction of exposure frequency and level of predictability, F(3,
567) = 2.18, p = .089, η
2
p
= .011. This interaction effect is again due to differential responses to
novel items. The two levels of predictability differ at zero exposures, F(1, 92) = 5.44, p = .022,
but not at 40 exposures, F(1, 92) = .6, p > .4, for the simple effects. These results are consistent
with Min and Schwarz’s (2017) observation that unpredictability increases consumers’ choice of
unfamiliar options, but incompatible with the logic of De Vries et al. (2010) and Gillebaart et al.
(2012). They highlight that not every problematic situation makes us turn to the safety of the
familiar. When the world is seen as unpredictable, exploring the unfamiliar may appear
beneficial.
As predicted, the opposite pattern emerged when control was manipulated through
personal influence. In this case, liking increased with the frequency of exposure under low
personal influence, F(3, 561) = 4.75, p = .003, η
2
p
= .025, for the simple effect, but not under
high personal influence, F(3, 561) = .08, NS, for the simple effect. This pattern is reflected in a
simple interaction of exposure frequency and level of personal influence, F(3, 567) = 2.28, p
= .078, η
2
p
= .012. This interaction effect is due to differential responses to novel items. The two
levels of personal influence differ at zero exposures, F(1, 95) = 6.27, p = .014, but not at 40
exposures, F(1, 95) = .92, p > .3, for the simple effects. Assuming that low personal influence is
a marker of a problematic situation, these results are consistent with the rationale of De Vries
and colleagues (2010) and Gillebaart and colleagues (2012)—people strongly prefer familiarity
over novelty when the world is a problematic place, but not when it is a safe place.

49

Figure 5. Liking for Nonsense Words at Four Levels of Exposure Frequency as a Function
of Cause and Level of Control in Study 1

Study 2
Method
Study 2 is a pre-registered replication (http://aspredicted.org/blind.php/?x=vp3uw2) of
Study 1 with two changes. First, data were collected on Amazon Mechanical Turk (N = 312;
mean age 38.63 years; 52.6% female) and participants were compensated with $0.45. Second,
following Zajonc (1968), I used four Chinese ideographs ( 章, 空, 弩, 痱) as stimuli. To ensure
sufficient power, I aimed for N = 280 (70 per cell) completed cases.
3.9
4.3
4.7
5.1
5.5
0 5 15 40
High Personal Control
Low Personal Control
Liking for Stimuli (1-7)
3.9
4.3
4.7
5.1
5.5
0 5 15 40
High Predictability
Low Predictability
Exposure Frequency
Exposure Frequency
High Personal Influence
Low Personal Influence

50

I added an attention check item, which 20 participants did not pass. I also asked
participants whether they had known any of the ideographs prior to the experiment; 8
participants reported they knew at least one. Excluding these participants did not change the
results and all were retained. Three participants could not think of any specific event that fit their
control manipulation assignment and were excluded, resulting in a total N = 309 for analyses.
Results and Discussion
Confirming the control manipulations, participants saw the world as less predictable after
thinking about unpredictability (M = 80.74) rather than predictability (M = 40.34; F(1, 305) = 84,
p < .001) and as less controllable after thinking about low (M = 66.02) rather than high personal
influence (M = 48.68; F(1, 305) = 15.62, p < .001).
Replicating the classic mere exposure effect, participants overall liked each ideograph
more, the more often they had seen it, F(3, 915) = 23.24, p < .001, η
2
p
= .071. This main effect of
exposure frequency was again qualified by the predicted 3-way interaction with cause and level
of control, F(3, 915) = 26.79, p < .001, η
2
p
= .081. As shown in Figure 6, liking increased with
the frequency of exposure under high predictability, F(3, 915) = 24.56, p < .001, η
2
p
= .075, for
the simple effect, but not under low predictability, F(3, 915) = 1. 89, p = .13. This pattern is
reflected in a significant simple interaction of exposure frequency and level of predictability,
F(3, 921) = 11.71, p < .001, η
2
p
= .037 and replicates Study 1. Again, level of predictability
influenced responses to novel items, F(1,150) = 58.23, p < .001, but not to familiar items,
F(1,150) = .02, p > .8, for the simple effects at zero and 40 exposures.

51

As predicted, the opposite pattern again emerged when control s manipulated through
personal influence. In this case, liking increased with frequency of exposure under low personal
influence, F(3, 915) = 26.09, p < .001, η
2
p
= .076, for the simple effect, but not under high
personal influence, F(3, 915) = .19, NS, for the simple effect. This pattern replicates Study 1 and
is reflected in a significant simple interaction of exposure frequency and level of personal
influence, F(3, 921) = 10.97, p < .001, η
2
p
= .034. As in Study 1, level of personal influence
influenced responses to novel items, F(1,155) = 26.39, p < .001, but not responses to familiar
items, F(1,155) = 1.07, p > .3, for the simple effects at zero and 40 exposures.

Figure 6. Liking for the Ideographs at Four Levels of Exposure Frequency as a Function of
Cause and Level of Control in Study 2

3.5
3.9
4.3
4.7
5.1
5.5
0 5 15 40
High Predictability
Low Predictability
3.5
3.9
4.3
4.7
5.1
5.5
0 5 15 40
High Personal Control
Low Personal Control
Liking for Stimuli (1-7)
Exposure Frequency
Exposure Frequency
High Personal Influence
Low Personal Influence

52

General Discussion
Past research on psychological control treated perceptions of predictability and
perceptions of personal influence as interchangeable instantiations of the control construct.
Challenging this assumption, research into novelty seeking found that perceptions of low
personal influence decreased novelty seeking (Faraji-Rad et al., 2016; Min & Schwarz, 2017),
whereas perceptions of low predictability increased novelty seeking (Min & Schwarz, 2017). The
latter observation may reflect an increased preference for the novel and/or strategic exploration
in preparation for an unknown future, which people may engage in even if they would truly
prefer the familiar. To shed light on this issue, I used Zajonc’s (1968) classic mere exposure
paradigm to explore whether different types of control differentially affect how much people like
novel and familiar stimuli.
As expected, people who had thought about a situation of low personal influence liked a
given stimulus more, the more often they had seen it, exhibiting a mere exposure effect. This
preference for the familiar was not observed for people who had thought about situations in
which they had high personal influence; they liked novel stimuli as much as familiar ones. These
observations are compatible with previous research that obtained mere exposure effects under
sad mood (De Vries et al., 2010) or prevention focus conditions (Gillebaart et al., 2012) but not
under happy mood or promotion focus conditions. Throughout, the novel is disliked when the
world is seen as problematic.
However, biological (e.g., Huston, 1994; McNeely, 1988) and social (e.g., Girard, 2010;
Huckfeldt, Johnson, & Sprague, 2004) theorizing suggests that a persistent avoidance of the
novel would have adverse implications for adaptation and survival in the face of ever-changing
conditions. Indeed, I obtained the reverse pattern when control was manipulated by perceived

53

unpredictability, which presumably highlights the need for rapid and frequent change. People
who had thought about unpredictable events in their own lives liked novel stimuli as much as
familiar ones, eliminating the mere exposure effect. However, people who had thought about
predictable events preferred the familiar, presumably because perceived predictability reduces
the need to adapt and discourages exploration of the unfamiliar.
These findings answer the question posed by my choice studies (i.e., Chapter 1), (Min &
Schwarz, 2017): does low predictability enhance novelty seeking because consumers understand
that it may be strategically beneficial to know one’s full range of options in an unpredictable
world or because they find novel options appealing? In the present Chapter 2, participants had no
opportunity to learn more about a stimulus by choosing it, which eliminates any potential
strategic incentive. Seeing the world as unpredictable nevertheless made novel stimuli more
appealing as reflected in increased liking ratings. This increased appeal of the unfamiliar
presumably fosters its exploration, with potentially beneficial effects for adaptation under
conditions of unpredictability.
These observations shed doubt on the usual treatment of personal influence and
predictability as interchangeable components of a general control construct and call for a broader
exploration of their differential implications. As a general principle, I conjecture that perceptions
of low personal influence draw attention to the risks involved in the unknown, which discourages
exploration. In contrast, perceptions of low predictability draw attention to the potential benefits
of knowing what one may face, which encourages exploration. If so, perceptions of low
predictability (vs. low personal influence) may also facilitate (vs. impair) the generation and/or
acceptance of novel solutions on creativity and problem solving tasks. Similarly, perceptions of
low predictability (vs. low personal influence) may increase (vs. impair) people’s willingness to

54

interact with unfamiliar others. Adding complexity to these predictions, perceptions of low
predictability will sometimes entail perceptions of low personal influence, raising questions of
their relative impact and prioritization that go beyond the current consensus of treating the two as
interchangeable components of control.
FUTURE DIRECTION
My dissertation shows that psychological control can have divergent effects on consumer
novelty-seeking behavior depending on how it is instantiated. Despite the conventional
assumption that personal influence and predictability are equivalent, I found that they lead to
opposite downstream consequences on consumers’ evaluation of novelty. One of the next
questions worth asking would be whether novelty seeking is only one or one of many domains
where personal influence and predictability do not result in similar effects. To expand the
understanding of psychological control, I claim that it critical to investigate when and why
psychological control has convergent versus divergent influences on consumer behavior.
The second question that awaits future research is the nature of orthogonality of personal
influence and predictability. In the current framework, I looked at four separate cases where only
one of the two becomes salient in consumers’ mind in the experiments. However, from the
theoretical as well as realistic standpoint, it is plausible to question if the two can coexist
orthogonally in one’s mind. For instance, one can perceive the world to be fairly random in
general, but she might still feel in control of her current situation or her life. On the other hand,
one can feel out of control over the current event, whereas at the same time she might believe she
nevertheless lives in a pretty predictable world. In such an experience in which predictability and
personal influence do not go hand in hand in one’s mind, how would they interact with each
other and influence her perception and evaluation? Further investigation of the nature of these

55

two components of psychological control will contribute to deepening our understanding of the
construct.
Lastly, my dissertation looked at consumer choices and liking of novel products and
stimuli. I argue that novelty seeking can have implications for other conceptually related
constructs. For example, to what extent one is open to exploring something new may predict the
way she interacts with things or people that are prototypical versus atypical. Considering the
current findings, would those who perceive low (vs. high) predictability and who feel high (vs.
low) personal influence show greater preference for products with an atypical design for the
product category? Furthermore, it is well known that fluently processed information feels more
familiar and thus more preferable than disfluently processed information (e.g., Reber, Schwarz,
& Winkielman, 2004; Reber, Winkielman, & Schwarz, 1998). Drawing on my current findings,
would those who have (vs. lack) personal influence and who think the world is unpredictable (vs.
predictable) like the product in a rather disfluently displayed ad more? I believe addressing these
intriguing follow-up questions will advance the literature of both psychological control and
consumer novelty seeking.

56

REFERENCES
Andreasen, A. R. (1984). Life status changes and changes in consumer preferences and
satisfaction. Journal of Consumer Research, 11(3), 784-794.
Antonovsky, A. (1979). Health, stress, and coping. California: Jossey Bass.
Assmus, G., Farley, J. U., & Lehmann, D. R. (1984). How advertising affects sales: Meta-
analysis of econometric results. Journal of Marketing Research, 65-74.
Banfield, J. C. (2011). Helping in a random world: Evidence that prosocial intentions and
behavior can satiate compensatory control needs (Doctoral dissertation). Retrieved from
UWSpace. (http://hdl.handle.net/10012/6025)
Banfield, C., Nadolny D., & Kay, A. C., (2013). Helping in a random world: Evidence that
prosocial intentions and behavior can satiate compensatory control needs. Unpublished
manuscript.
Berning, C. A. K., & Jacoby, J. (1974). Patterns of information acquisition in new product
purchases. Journal of Consumer Research, 1(2), 18-22.
Berry, J. M., & West, R. L. (1993). Cognitive self-efficacy in relation to personal mastery and
goal setting across the life span. International Journal of Behavioral Development, 16(2),
351-379.
Bianchi, M. (Ed.). (2006). The active consumer: Novelty and surprise in consumer choice (Vol.
20) .London: Routledge.
Bornstein, R. (1989). Exposure and affect: Overview and meta-analysis of research, 1968-1987.

57

Psychological Bulletin, 106, 265-289.
Bourgeois, L. J. (1985). Strategic goals, perceived uncertainty, and economic performance in
volatile environments. Academy of Management Journal, 28(3), 548-573.
Chaxel, A. S. (2016). Why, when, and how personal control impacts information processing:
A framework. Journal of Consumer Research, 43(1), 179-197.
Cowart, K. O., Fox, G. L., & Wilson, A. E. (2008). A structural look at consumer innovativeness
and self ‐congruence in new product purchases. Psychology & Marketing, 25(12), 1111-
1130.
Cutright, K. M., Bettman, J. R., & Fitzsimons, G. J. (2013). Putting brands in their place: How a
lack of control keeps brands contained. Journal of Marketing Research, 50, 365-377.
Cutright, K. M., & Samper, A. (2014). Doing it the hard way: How low control drives
preferences for high-effort products and services. Journal of Consumer Research, 41(3),
730-745.
Damanpour, F., & Gopalakrishnan, S. (1998). Theories of organizational structure and
innovation adoption: the role of environmental change. Journal of Engineering and
Technology Management, 15(1), 1-24.
De Vries, M., Holland, R. W., Chenier, T., Starr, M. J., & Winkielman, P. (2010). Happiness
cools the warm glow of familiarity: Psychophysiological evidence that mood modulates
the familiarity-affect link. Psychological Science, 21, 321-328.

58

Faraji ‐Rad, A., Melumad, S., & Johar, G. V. (2017). Consumer desire for control as a barrier to
new product adoption. Journal of Consumer Psychology, 27(3), 347-354.
Flavell, J., H. (1977). Cognitive development, Englewood Cliffs, New Jersey: Prentice-Hall.
Folkman, S. (1984). Personal control and stress and coping processes: A theoretical
analysis. Journal of Personality and Social Psychology, 46(4), 839-852.
Friesen, J. P., Kay, A. C., Eibach, R. P., & Galinsky, A. D. (2014). Seeking structure in social
organization: Compensatory control and the psychological advantages of hierarchy.
Journal of Personality and Social Psychology, 106, 590-609.
Gillebaart, M., Förster, J., & Rotteveel, M. (2012). Mere exposure revisited: the influence of
growth versus security cues on evaluations of novel and familiar stimuli. Journal of
Experimental Psychology: General, 141, 699-714.
Girard, V. (2010). Diversity is me (survival guide for mixed race people). LA: Lulu.com.
Grant, S. J., & Tybout, A. M. (2008). The effect of temporal frame on information considered in
new product evaluation: The role of uncertainty. Journal of Consumer Research, 34(6),
897-913.
Griffin, J. M., Fuhrer, R., Stansfeld, S. A., & Marmot, M. (2002). The importance of low control
at work and home on depression and anxiety: do these effects vary by gender and social
class?. Social Science & Medicine, 54(5), 783-798.
Hafer, C. L., & Begue, L. (2005). Experimental research on just-world theory: problems,
developments, and future challenges. Psychological Bulletin, 131(1), 128-167.

59

Hamerman, E. J., & Johar, G. V. (2013). Conditioned superstition: Desire for control and
consumer brand preferences. Journal of Consumer Research, 40(3), 428-443.
Heckhausen, J., & Schulz, R. (1995). A life-span theory of control. Psychological
Review, 102(2), 284-304.
Hekkert, P., Snelders, D., & Wieringen, P. C. (2003). ‘Most advanced, yet acceptable’:
Typicality and novelty as joint predictors of aesthetic preference in industrial
design. British Journal of Psychology, 94(1), 111-124.
Hirschman, E. C. (1980). Innovativeness, novelty seeking, and consumer creativity. Journal of
Consumer Research, 7(3), 283-295.
Hoeffler, S. (2003). Measuring preferences for really new products. Journal of Marketing
Research, 40(4), 406-420.
Horswill, M. S., & McKenna, F. P. (1999). The effect of perceived control on risk
taking1. Journal of Applied Social Psychology, 29, 377-391.
Huckfeldt, R., Johnson, P. E., & Sprague, J. (2004). Political disagreement: The survival of
diverse opinions within communication networks. UK: Cambridge University Press.
Huston, M. A. (1994). Biological diversity: the coexistence of species. UK: Cambridge
University Press.
Im, S., Bayus, B. L., & Mason, C. H. (2003). An empirical study of innate consumer
innovativeness, personal characteristics, and new-product adoption behavior. Journal of
the Academy of Marketing Science, 31(1), 61-73.

60

James, W. (1890). The consciousness of self, The Principles of Psychology, Vol. 1,
New York: Henry Holt.
________ (1961), Psychology: The Briefer Course, New York: Harper & Row.
Janoff-Bulman, Ronnie (1992), Shattered Assumptions: Towards a New Psychology of Trauma.
New York: Free Press.
Kalish, S. (1985). A new product adoption model with price, advertising, and
uncertainty. Management Science, 31(12), 1569-1585.
Kay, A. C., Gaucher, D., Napier, J. L., Callan, M. J., & Laurin, K. (2008). God and the
government: Testing a compensatory control mechanism for the support of external
systems. Journal of Personality and Social Psychology, 95, 18-35.
Kay, A. C., Laurin, K., Fitzsimons, G. M., & Landau, M. J. (2014). A functional basis for
structure-seeking: Exposure to structure promotes willingness to engage in motivated
action. Journal of Experimental Psychology: General, 143(2), 486-491.
Kay, A. C., Sullivan, D., & Landau, M. J. (2015). Psychological importance of beliefs in control
and order: Historical and contemporary perspectives in social and personality
psychology. In M. Mikulincer & P. R. Shaver (Eds.), APA handbook of personality and
social psychology. Vol. 1: Attitudes and social cognition (Vol. 1, pp. 309–337).

61

Kay, A. C., Whitson, J. A., Gaucher, D., & Galinsky, A. D. (2009). Compensatory control:
Achieving order through the mind, our institutions, and the heavens. Current Directions
in Psychological Science, 18(5), 264-268.
Kelly, G. (1955). The psychology of personal constructs, New York: Norton.
Krantz, D. L. (1998). Taming chance: Social science and everyday narratives. Psychological
Inquiry, 9(2), 87-94.
Kruglanski, A. W., & Webster, D. M. (1996). Motivated closing of the mind:" Seizing" and"
freezing." Psychological Review, 103(2), 263-283.
Landau, M. J., Kay, A. C., & Whitson, J. A. (2015). Compensatory control and the appeal of a
structured world. Psychological Bulletin, 141(3), 694.
Lerner, M. J. (1980). The belief in a just world: A fundamental delusion. New York:
Plenum.
Maner, J. K., & Schmidt, N. B. (2006). The role of risk avoidance in anxiety. Behavior
Therapy, 37, 181-189.
McNeely, J. A. (1988). Economics and biological diversity: developing and using economic
incentives to conserve biological resources. Switzerland: IUCN.
McShane, B. B., & Böckenholt, U. (2017). Single-paper meta-analysis: Benefits for study
summary, theory testing, and replicability. Journal of Consumer Research, 43(6), 1048-
1063.

62

Menon, S., & Kahn, B. E. (1995). The impact of context on variety seeking in product
choices. Journal of Consumer Research, 22(3), 285-295.
Midgley, David F. and Grahame R. Dowling (1993), "A Longitudinal Study of Product form
Innovation: The Interaction between Predispositions and Social Messages," Journal of
Consumer Research, 19(4), 611-25.
Miller, Danny and Peter H. Friesen (1983), "Strategy ‐making and Environment: The Third
Link," Strategic Management Journal, 4(3), 221-35.
Min, B., & Schwarz, N. (2017). Novelty as risk and opportunity: Opposite effects of low
personal control and perceived unpredictability on novelty seeking. Unpublished
manuscript.
Monin, B. (2003). The warm glow heuristic: when liking leads to familiarity. Journal of
Personality and Social Psychology, 85(6), 1035-1048.
Montoya, R. M., Horton, R. S., Vevea, J. L., Citkowicz, M., & Lauber, E. A. (2017). A re-
examination of the mere exposure effect: The influence of repeated exposure on
recognition, familiarity, and liking. Psychological Bulletin, 143, 459-498.
Moreau, C. P., Markman, A. B., & Lehmann, D. R. (2001). “What is it?” Categorization
flexibility and consumers' responses to really new products. Journal of Consumer
Research, 27(4), 489-498.
Morling, B., & Evered, S. (2006). Secondary control reviewed and defined. Psychological

63

Bulletin, 132, 269-296.
Ostlund, L. E. (1974). Perceived innovation attributes as predictors of innovativeness. Journal of
Consumer Research, 1(2), 23-29.
Park, C. W., & Lessig, V. P. (1981). Familiarity and its impact on consumer decision biases and
heuristics. Journal of Consumer Research, 8(2), 223-230.
Reber, R., Schwarz, N., & Winkielman, P. (2004). Processing fluency and aesthetic pleasure: Is
beauty in the perceiver's processing experience?. Personality and social psychology
review, 8(4), 364-382.
Reber, R., Winkielman, P., & Schwarz, N. (1998). Effects of perceptual fluency on affective
judgments. Psychological Science, 9(1), 45-48.
Roehrich, G. (2004). Consumer innovativeness: Concepts and measurements. Journal of
Business Research, 57(6), 671-677.
Rogers, E. M. (1976). New product adoption and diffusion. Journal of consumer Research, 2(4),
290-301.
Rutjens, B. T., van Harreveld, F., & van der Pligt, J. (2013). Step by step: Finding compensatory
order in science. Current Directions in Psychological Science, 22, 250-255.
Schiffman, L. G. (1972). Perceived risk in new product trial by elderly consumers. Journal of
Marketing Research, 9(1), 106-108.
Schwarz, N., & Clore, G. L. (1996). Feelings and phenomenal experiences. Social psychology:
Handbook of Basic Principles, 2, 385-407. ed. E. Tory Higgins and Arie Kruglanski,
Hillsdale, New Jersey: Erlbaum.

64

Skinner, E. A. (1996). A guide to constructs of control. Journal of Personality and Social
Psychology, 71, 549-570.
Titchener, E. B. (1918). A Textbook of Psychology, New York: Macmillan.
Thompson, S. A., & Sinha, R. K. (2008). Brand communities and new product adoption: The
influence and limits of oppositional loyalty. Journal of Marketing, 72(6), 65-80.
Troisi, J. D., & Gabriel, S. (2011). Chicken soup really is good for the soul: “Comfort food”
fulfills the need to belong. Psychological Science, 22(6), 747-753.
VanBergen, N., & Laran, J. (2016). Loss of control and self-regulation: The role of childhood
lessons. Journal of Consumer Research, 43(4), 534-548.
White, R. W. (1959). Motivation reconsidered: The concept of competence. Psychological
Review, 66(5), 297-333.
Whitson, J. A., Galinsky, A. D., & Kay, A. (2015). The emotional roots of conspiratorial
perceptions, system justification, and belief in the paranormal. Journal of Experimental
Social Psychology, 56, 89-95.
Wood, S. (2009). The comfort food fallacy: Avoiding old favorites in times of change. Journal
of Consumer Research, 36(6), 950-963.
Zajonc, R. B. (1968). Attitudinal effects of mere exposure. Journal of Personality and
Social Psychology, 9 (2), 1-27.
Zirger, B. J., & Maidique, M. A. (1990). A model of new product development: An empirical
test. Management Science, 36(7), 867-883.

65

APPENDIX 1. News Excerpts shown in the Predictability (above) and Unpredictability
(below) conditions in Chapters 1 and 2
Is Everything Under Control? A Harvard Conference Reveals the Answer
“The world really is an orderly place,” said Thomas Cornwallis, a statistics professor at Oxford.
Cornwallis made the comments at a conference hosted by Harvard University in January. The
conference, titled “Understanding the World” was aimed at trying to understand the causes of
events in the world. Cornwallis was one of several panellists who agreed that the world mostly
operates in stable, understandable patterns. At the same conference, Marten Keese, a professor at
Utrecht University in the Netherlands, spoke about an article he published in the renowned
journal Science. Keese claimed that people's behaviour has clear causes. “There are good reasons
for people to believe that the world is orderly and non-random,” said Keese. “Most people
believe their outcomes are under control, and our data support that belief.”

Is Everything Under Control? A Harvard Conference Reveals the Answer
“The world really is a random place,” said Thomas Cornwallis, a statistics professor at Oxford.
Cornwallis made the comments at a conference hosted by Harvard University in January. The
conference, titled “Understanding the World” was aimed at trying to understand the causes of
events in the world. Cornwallis was one of several panellists who agreed that the world mostly
operates in erratic, unpredictable ways. At the same conference, Marten Keese, a professor at
Utrecht University in the Netherlands, spoke about an article he published in the renowned
journal Science. Keese claimed that people's behaviour does not have clear causes. Although
people may believe that the world is orderly and non-random, Keese says our perceptions are

66

flawed. “Unperceived factors determine what happens to us. Most people believe their outcomes
are under control, but our data suggest that random fluctuations have greater effects.”

67

APPENDIX 2. Six Chocolate Bars shown in Study 1 through Study 4B in Chapter 1 —
Three Familiar Products (above) and Three Unfamiliar Products (below)

Abstract (if available)

Abstract Consumers constantly perceive, process, and evaluate products and services that are either familiar or novel to them. Consistent with the concept of a “warm glow of familiarity” (i.e., “chicken soup for the soul” effect), prior research suggests that consumers who lack control over their circumstances tend to stick to familiar options rather than try unfamiliar options. This dissertation differentiates two distinct aspects of control: predictability (perceiving the world as an unpredictable vs. predictable place) and personal influence (having vs. lacking control over one’s own circumstances). These two aspects have long been considered equivalent in the control literature and used interchangeably in control experiments. However, this dissertation shows that they can have different implications for how consumers deal with novelty. From a consumer’s perspective, novelty presents unknown opportunities as well as unknown risks and these aspects are differentially highlighted by low predictability and low personal influence. Specifically, low predictability suggests that the world is uncertain and things may change any minute. Under these circumstances, consumers are motivated to learn about the full range of options in preparation for possible surprises, which increases the exploration of novel options. In contrast, low personal influence is associated with current difficulties and consumers hesitate to expose themselves to additional risks. This results in a preference for the familiar. This dissertation addresses these diverging aspects of novelty and control, and examines their implications for consumer judgment and choice. In Chapter 1, I report six choice studies that show higher novelty seeking under low (vs. high) predictability, but lower novelty seeking under low (vs. high) personal influence, and identify the underlying processes. Note, however, that choice does not necessarily imply liking—under low predictability, consumers may choose a novel option because it seems useful given the circumstances, even though they would actually find the familiar option more appealing. Chapter 2 addresses this possibility. Drawing on the classic mere exposure effect, I find from two studies that low personal influence enhances the mere exposure effect, as reflected in increased liking of familiar (previously seen) stimuli, whereas low predictability eliminates the classic mere exposure effect. Overall, this dissertation identifies different aspects of control, shows their diverging impact on consumers’ preference for familiar and novel products, and identifies managerial implications for more effective promotions of both familiar products and novel products.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Manipulating consumer opinion on social media using trolls and influencers

PDF

Performance and attention novelty slows hedonic adaptation during habit formation

PDF

Consumers' subjective knowledge influences evaluative extremity and product differentiation

PDF

Scale construction and effects of brand authenticity

PDF

Technology, behavior tracking, and the future of work

PDF

How product designs and presentations influence consumers’ post-acquisition decisions

PDF

Better now than later: the cost of victims' delayed accusations

PDF

The effects of a customer's comparative processing with positive and negative information on product choice

PDF

Increasing capabilities or decreasing cost: Fairness perceptions of job displacement due to automation and outsourcing

PDF

Essays on the luxury fashion market

PDF

Essays on consumer returns in online retail and sustainable operations

PDF

The effects of curiosity-evoking events on consumption enjoyment

PDF

Creation and influence of visual verbal communication: antecedents and consequences of photo-text similarity in consumer-generated communication

PDF

Shrinkage methods for big and complex data analysis

PDF

The curse of loyalty: interdependent self construal and support for corruption

PDF

Essays on the role of entry strategy and quality strategy in market and consumer response

PDF

Understanding virality of YouTube video ads: dynamics, drivers, and effects

PDF

Marketing strategies with superior information on consumer preferences

PDF

Essays on revenue management with choice modeling

PDF

Essays on information and financial economics

Asset Metadata

Creator Min, Bora (author)

Core Title Novelty versus familiarity: divergent effects of low predictability and low personal influence

School Marshall School of Business

Degree Doctor of Philosophy

Degree Program Business Administration

Publication Date 04/12/2018

Defense Date 03/19/2018

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

Tag Control,familiarity,novelty,OAI-PMH Harvest,personal influence,predictability

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Park, C.W. (committee chair), Schwarz, Norbert (committee member), Wakslak, Cheryl (committee member)

Creator Email bestbora@gmail.com,boramin@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-7060

Unique identifier UC11670716

Identifier etd-MinBora-6235.pdf (filename),usctheses-c89-7060 (legacy record id)

Legacy Identifier etd-MinBora-6235.pdf

Dmrecord 7060

Document Type Dissertation

Rights Min, Bora

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 a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA