University of Southern California Dissertations and Theses

Attention, retention, and incentive processes in observational learning

(USC Thesis Other)

Attention, retention, and incentive processes in observational learning

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content ATTENTION, RETENTION, AND INCENTIVE PROCESSES
IN OBSERVATIONAL LEARNING
by
Steven Max Vincent
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
(Psychology)
January 1973
INFORMATION TO USERS
This dissertation was produced from a microfilm copy of the original document.
While the most advanced technological means to photograph and reproduce this
document have been used, the quality is heavily dependent upon the quality of
the original submitted.
The following explanation of techniques is provided to help you understand
markings or patterns which may appear on this reproduction.
1. The sign or "target" for pages apparently lacking from the document
photographed is "Missing Page(s)". If it was possible to obtain the
missing page(s) or section, they are spliced into the film along with
adjacent pages. This may have necessitated cutting thru an image and
duplicating adjacent pages to insure you complete continuity.
2. When an image on the film is obliterated w ith a large round black
mark, it is an indication that the photographer suspected that the
copy may have moved during exposure and thus cause a blurred
image. You will find a good image of the page in the adjacent frame.
3. When a map, drawing or chart, etc., was part of the material being
photographed the photographer followed a definite method in
"sectioning" the material. It is customary to begin photoing at the
upper left hand corner of a large sheet and to continue photoing from
left to right in equal sections with a small overlap. If necessary,
sectioning is continued again — beginning below the first row and
continuing on until complete.
4. The majority of users indicate that the textual content is of greatest
value, however, a somewhat higher quality reproduction could be
made from "photographs" if essential to the understanding of the
dissertation. Silver prints of "photographs" may be ordered at
additional charge by writing the Order Department, giving the catalog
number, title, author and specific pages you wish reproduced.
University Microfilms
300 North Zeeb Road
Ann Arbor, Michigan 46106
A Xerox Education Company
T
| 73-14,452
VINCENTt Steyen Wax, 1947-
ATTENTION, RETENTION, AND INCENTIVE PROCESSES
IN OBSERVATIONAL LEARNING.
University of Southern California, Ph.D., 1973
Psychology, experimental
University Microfilms, A XEROX Company, Ann Arbor, Michigan
THIS DISSERTATION HAS BEEN MICROFILMED EXACTLY AS RECEIVED.
UNIVERSITY O F SO UTHERN CALIFORNIA
TH E GRADUATE SCHO OL
U N IV E R S IT Y PARK
LOS A NG ELES. C A L IF O R N IA 0 0 0 0 7
This dissertation, w ritten by
......Steven^Mu^Vlncent
under the direction of A*.®.... Dissertation Com
mittee, and approved by a ll its members, has
been presented to and accepted by T he Graduate
School, in p a rtia l fu lfillm e n t of requirements of
the degree of
D O C T O R O F P H IL O S O P H Y
Dt*u
Date...?®bruaryj^. 19.73
DISSERTATION COMMITTEE
Chtunm s
PLEASE NOTE:
Some pages may have
i nd i st i net print.
Filmed as received.
University Microfilms, A Xerox Education Company
ACKNOWLEDGMENTS
As is the case, I suppose, with most graduate stu
dents at this point in time, I can think of an almost
infinite number of people to whom I genuinely owe a great
deal. To thank all of these people here is much too great
a task, and so I choose to acknowledge only a relative
few, while keeping the multitudinous others clearly in
mind.
I must begin where I began— with my parents, Lloyd
and Lorraine Vincent. They guided when I needed guiding,
pushed when I needed pushing, and supported me at all
times. It is impossible to say enough, so I can only
trust that the meaning and feeling behind what I say is
clear.
The members of my dissertation committee were with
out exception helpful, and remarkably low as anxiety
arousers. Drs. A1 Marston and Norm Tiber combined logical
and procedural advice with a demonstrated desire to help
my graduate training in many different ways. I am pleased
to regard them as friends as well as teachers. Dr. Steve
Prankel has been my major advisor throughout nearly all
of my graduate training in addition to chairing my disser
tation committee. His support, guidance, and concern have
sustained me through all aspects of my training, and it
ii
would be very difficult to exaggerate his value to me
as a model— a model as a therapist, as a researcher, as
a teacher— as a psychologist and a person in toto.
Lastly and most importantly, I want to thank my
wife, Joan, without whom this dissertation would quite
likely have been finished two years from now. In very
tangible ways she contributed a great deal, for example,
serving as the model in this experiment. More critically,
however, she made this research, and the majority of my
training leading to it, a possibility by means of so many
intangible contributions. Mostly she did it by being who
she is, the way she is, and I am only beginning to appre
ciate this. I hope I never finish.
iii
TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS ..................................... ii
LIST OF T A B L E S ..................................... vi
Chapter
I. INTRODUCTION ............................... 1
II. THEORETICAL CONCEPTIONS .................... 3
III. ATTENTION PROCESSES ......................... 15
IV. RETENTION PROCESSES ......................... 26
V. INCENTIVE EFFECTS ........................ 34
VI. STATEMENT OF THE PROBLEM.................. 37
VII. METHOD .................................... 42
Subjects................................. 42
Apparatus................................. 42
Procedure................................. 43
Dependent Measures ...................... 49
VIII. RESULTS ................................. 54
Attention Effects.......................... 55
Retention Effects.......................... 56
Incentive Effects.......................... 68
Sex Differences............................ 70
Interaction Effects ...................... 71
Various Effects on Other Dependent Measures 72
iv
Chapter Page
IX. DISCUSSION................................... 71 *
Attention................................. 75
Retention................................. 79
Incentive................................. 85
APPENDICES 92
REFERENCES ..................................... 116
v
List cf Tables
Page
1. Imitation Scoring Code...................... 5C
2. Analysis of Variance for the Number of Whole
Reproductions of Elements........ 57
3. Analysis of Variance for the Number cf
Partial Reproductions of Elements .... 58
A. Analysis of Variance for the Hunter of Whole
Plus Partial Reproductions ............... 59
5. Analysis of Variance for the Number of Com
pletely Reproduced Sequences ............. 60
6. Analysis of Variance for the Number of Novel
Behaviors Produced ...................... 6l
7. Analysis of Variance for the Number of
Borrowed Behaviors Produced ............. 62
8. Analysis of Variance for the Total Number of
Extraneous Behaviors .................... 63
9. Mean Number of Whole Plus Partial Imitations
as a Function of Type of Code Used .... 65
10. Mean Number of Extraneous Behaviors as a
Function of the Type of Code Used .... 65
11. Mean Number and Mean Percentage of Whole Plus
Partial Imitations as a Function of Atten
tion, Retention, Incentive, and Sex . . . 66
vi
Table Page
12. Mean Number of Extraneous Behaviors as a
Function of Attention, Retention, Incen
tive, and S e x ............................... 67
13* Mean Number of Whole Elements Imitated as a
Function of Incentive....................... 69
14. Mean Number of Whole Plus Partial Elements
Imitated as a Function of Incentive .... 69
15. Mean Number of Completely Reproduced Se
quences as a Function of Incentive .... 69
vii
CHAPTER I
INTRODUCTION
Those variables which govern the acquisition, re
tention, and performance of new behaviors have been among
the most constant topics of psychological theory and re
search for many years. A particular aspect of this general
learning problem is the area known variously as "modeling,"
"imitation," "observational learning," "vicarious learn
ing," "identification," and by several other names. All
these labels refer to the phenomenon of one individual ac
quiring a new behavior seemingly as a function of his
having observed another individual performing the behavior.
A great deal of evidence has accumulated over the
last decade to show that a variety of behavioral changes
are feasible through the use of observational learning.
New and complex patterns of behavior are obtainable through
this process (e.g., Bandura and Kupers, 1964; Gerst, 1971).
Inhibition and disinhibition of already learned behaviors
have also been demonstrated (e.g., Bandura and Mischel,
1965; Bandura, Grusec, and Menlove, 1967), and the facil
itation of previously learned responses, which have rarely
or never been punished, and thus are distinguishable from
1
2
disinhibited responses, is also given empirical support in
the research literature (Bandura and Harris, 1966; Rosen
thal, Zimmerman, and Durning, 1970).
Nevertheless, many significant questions remain
regarding the phenomenon of observational learning. One
of the most important of these questions is that one which
seeks to understand how observational learning takes place,
i.e., the attempt to specify the necessary and sufficient
conditions for the successful completion of this partic
ular learning process.
CHAPTER II
THEORETICAL CONCEPTIONS
The earliest attempts to explain imitation date
back to the turn of the century. Morgan (1896) and
McDougall (1908) spoke of modeling and imitation as innate
propensities, i.e. instincts, and consequently they re
ceived little or no empirical investigation. As the idea
of the role of instincts in human behavior fell in stature
new explanations were offered to account for the imitation.
Allport (1924) and Holt (1931) proposed associative the
ories of modeling which essentially said that temporal con
tiguity between the modeling stimuli and matching re
sponses by the observer were sufficient for imitation.
However, these theories lacked explanatory power in regard
to two commonly observed occurrences of imitation. First,
they offered no mechanism to explain how a novel response,
composed of several already known response elements, might
be acquired in an integrated form during the model-observer
interaction. Secondly, the associative theories were
unable to explain how an observer imitated the behavior
of a model long after he observed the model's performance
and without having practiced the matching response pre
viously.
3
1»
When the importance of reinforcement began to be
come known, and when behaviors began to become predictable
to a certain extent by virtue of one's knowing the pre
vailing reinforcement contingencies, explanations of imi
tation-observational learning were also framed in rein
forcement terminology. The earliest work in imitation
from this point of view was done by Miller and Dollard
(19^1). According to these writers, the necessary condi
tions for observational learning were a motivated subject
who was reinforced for matching the behavior of a model
on an initially trial and error basis. The Miller and
Dollard paradigm was one in which the observer S watched
a model responding to environmental cues which were not
visible to the S. The model was working on a two-choice
discrimination problem, and his responses were consistently
rewarded as were S's responses each time S matched the
model's choice, which he came to do consistently. This
was considered to be a demonstration of observational
learning, however, it appears actually to have been an
example of discrimination learning in which the model's
choice became the discriminative stimulus for S's choice.
Moreover, this conceptualization offered no explanation
for how a novel response might be acquired by S. Bandura
(1965) argued that this account was weak for its failure
to explain instances in which the observer did not perform
the response in the presence of the model, and instances
in which neither the model nor the observer was rewarded
for the response.
The Skinnerian analysis of modeling naturally also
places a great deal of emphasis on reinforcement. Baer
and his associates (Baer and Sherman, 1964; Baer, Peter
son, and Sherman, 1967) have been the main proponents of
this analysis of observational learning. For them, imi
tation is actually the process of an individual matching
his own response to the modeled response. This is a
result of a process of differential reinforcement during
which the observer has been reinforced for imitating and
ignored or punished when he did not imitate. By virtue
of the fact that imitating has been so frequently paired
with reinforcement, the act of matching the behavior of
another comes to have secondary reinforcing properties.
As a result of this, imitation is a reinforcer in and of
itself, and the Skinnerian theories of observational learn
ing use this assumption to explain how learning by observ
ing can occur when S does not receive an external reward
for his matching response.
Numerous experiments have indeed demonstrated that
reinforcing some imitative responses can result in an in
crease in the frequency of non-reinforced imitation (Baer
and Sherman, 1964; Baer, Peterson, and Sherman, 1967;
Bringham and Sherman, 1968). However, this fact does not
prove that reinforcing properties are inherent in
Imitation, since in each of these studies when reinforce
ment was halted imitation of both reinforced and non
reinforced responses (both of which continued to be
modeled) decreased immediately. If imitation per se is
reinforcing, such a decrement should not occur.
Bandura (1969) reviewed the Skinnerian analysis of
modeling, and found it lacking on the grounds that it
specifies as necessary conditions some elements which may
not even bo present. The standard paradigm for the Skin
nerian analysis is Sd—> R — >Sr. In observational learning
S^ is the modeling stimulus, R is the observer's matching
response, and Sr is the reinforcing stimulus. Bandura
criticizes the explanatory potential of this paradigm with
regard to observational learning by arguing that it does
not show how the learning may occur when the observer does
not make the matching response, and therefore is not re
warded, vrhile in the presence of the modeling stimulus
(Sd). Further, when the observer does later reproduce his
newly learned behavior the is often no longer present.
Therefore, according to Bandura, it would appear that dur
ing acquisition of new behavior R and Sr need not be pres
ent, and further when R does occur need not be present.
This conclusion points up an interesting and impor
tant distinction which must be made, with regard to obser
vational learning, namely, the difference between the ac
quisition and performance of new behaviors. It is clear
7
from a great deal of research (Bandura, 1968) that rein
forcement contingencies are very powerful factors in de
termining whether or not already learned behaviors are
actually performed. Further, since learning is typically
inferred from observed changes in performance, reinforce
ment is often assigned an essential role in the process
of learning, as was discussed previously. One experiment
in particular, however, points up the clear difference
that may exist between learning and performance, and also
reveals the role of reinforcement is the control of both
processes.
In this study (Bandura, 1965) children observed a
film of a model who demonstrated novel physical and verbal
aggressive behaviors toward a bozo doll. One group of
children saw the model being generously rewarded for his
aggression; a second group saw him being severely punished
for the same behavior; and a third group saw the model
being aggressive without experiencing any consequences
whatsoever. Rewards and punishments in the first two
groups were both physical and social. A post-exposure
test of aggression was made by counting the number of
imitative aggressive responses made by each S during a
ten-minute play period during which the S was left alone
in a room with the bozo doll and other toys. Results from
this test indicated a significant treatment effect in
which the "model rewarded" and "no consequences group"
displayed significantly more imitative aggressive behav
iors than did the "model punished" group. Following the
post-exposure test, however, children in all three groups
were offered large rewards for imitating the model’s
aggressive behavior. This introduction of incentives
eliminated the previously found differences, and revealed
that the children in all three groups had learned equiva
lent amounts of novel aggressive behavior.
From the evidence furnished by this study by Bandura
the differing importance of the role of reinforcement in
the two processes of acquisition and performance comes to
be clarified. Performance might be conceptualized as the
Individual's willingness to reproduce those behaviors
which he has in his repertoire. This process of repro
duction of learned behaviors is controlled largely by the
reinforcing consequences of the behayior, and therefore
would seem to explain to a large extent a Skinnerian anal
ysis such as was described earlier. However, as was pre
viously pointed out, the standard operant paradigm (S^— *
R — * Sr) does not account for a large class of observa-
tionally learned behaviors in which the Sd is not tempo
rally contiguous with the R and Sr.
Mowrer (I960) and Aronfreed (1968) have offered
another another theory of imitation which places a great
deal of emphasis upon the affect which has conditioned to
imitative responses. However, because this affective con
ditioning is assumed to be a result of the reinforcement
associated either with the model's behavior or the ob
server’s imitative response, this theory has the same
difficulty that other reinforcement theories experience
when faced with explaining observational learning that
takes place without reinforcement (Sheffield and Maccoby,
1961). Nevertheless, these formulations by Mowrer and
Aronfreed do raise the prospect of an interesting new
factor in observational learning, i.e. a cognitive or
symbolic component. Mowrer suggests that the selection of
any one response from among the alternatives available in
a particular situation takes place on a symbolic level.
Aronfreed proposes the existence of "cognitive templates"
of a behavior. Neither of these writers specifies, how
ever, the nature of these symbols or templates, or how they
come to be formed. What they do, though, is lead logi
cally to a consideration of a final theory of observa
tional learning, Bandura's contiguity mediational theory.
Bandura (1965) states that it is possible for a
person to acquire a modeled behavior without making any
other response than simply observing the model perform,
and he calls this no-trial learning. In order to account
for this phenomenon, he proposes a contiguity-mediational
theory which revolves around two representational systems—
10
an imaginal one and a verbal one (Bandura, 1969). These
systems are actually coding systems which transform ob
served behavior into images or words. Later, when it is
desirable to reproduce the observed behavior these codes
can mediate the recall of the behavior in its complete and
integrated form.
In the case of the formation of images a process of
sensory conditioning is assumed to occur. The observation
of a model's behavior presumably results in a series of
perceptual responses on the part of the observer. Because
these responses are temporally contiguous, Bandura asserts
that they become sequentially associated and integrated in
the central nervous system. If the sequence of perceptual
responses is elicited often enough (the necessary number
presumably varying for different sequences at different
times) a process of sensory conditioning takes place so
that the elicitation of any one event in the sequence
automatically results in images of the associated events.
With regard to verbal coding Bandura is less spe
cific about how the process takes place, saying simply:
"After modeled sequences of responses have been translated
into readily utilizable verbal symbols, later performances
of matching behavior can be effectively controlled by co
vert verbal self-directions (1969, p. 13^0 " It is not
clear from this statement whether or not the verbal codes
11
are sequentially associated and centrally Integrated, as
were the Images. It seems logical to assume that verbal
coding does not always take place, I.e. that a person may
observe and make perceptual responses and learn without
labeling the responses. What this seems to imply is that
verbal labeling responses in observational learning act
as an adjunct to perceptual responses, and, as Bandura
suggests, may facilitate long-term retention by being more
effective than imaginal codes.
In this theory of observational learning it is
assumed to be necessary that the stimulus elements of a
new behavior be temporally contiguous. This is not, how
ever, asserted to be a sufficient condition for learning
to take place. Rather, Bandura offers a set of Inter
related subprocesses which are controlled by different
variables, and by which the nature and degree of observa
tional learning are influenced. Although he does not
specify which of these subprocesses do provide the suffi
cient conditions for learning, he does identify them and
attempts to explain their significance.
The first subsystem involves the attention pro
cesses which are going on while a person is observing a
model. It is assumed by the contigulty-mediational theory
that contiguous presentation of stimuli will fail to pro
duce learning if attention is not evoked. Without appro
priate attention observation may be directed to irrelevant
events, and this may work to prevent learning of the rel
evant events for the new behavior by blocking sensory
registration of the stimuli. It may be said, then, that
attention allows for sensory registration of modeling
stimuli, and that such registration is essential for ob
servational learning because perceptual responses may not
occur centrally without it. In Bandura’s words, "To pro
duce learning, therefore, stimulus contiguity must be
accompanied by discriminative observation" (1969, p. 136).
The second subsidiary process in this theory of
modeling is the retention process. Discussion of this
process leads logically to consideration of a distinc
tion between learning and memory. Such a distinction may,
however, be at best superficial, and at worst obstructive.
Learning is always inferred from performance, and since
contiguity-mediational theory attempts to explain how
Imitation may occur after modeling cues have been re
moved (as well as when they are present), it is inferring
learning from delayed performance. For such performance
to occur, it is necessary that remembering or retaining
processes be taking place. Therefore the retention process
may be seen to be a subsidiary of the larger learning pro
cesses. In contiguity-mediational theory this retention
process is assumed to involve symbolic systems, as was
mentioned previously.
13
The third subprocess In the modeling phenomenon as
described by Bandura is the motor-reproduction process.
This is particularly applicable to behaviors which are
comprised of a set of complex motor responses. In such
responses many of the simpler elements are themselves
rather complex, and learning the integrated pattern of the
larger behavior is not possible without knowing the simpler
elements. Additionally some behaviors may simply demand
motor practice to be learned correctly, even though the
overall pattern of the behavior is acquired by observa
tion as is the case in golf or ballet. In cases such as
these, performance of the imitative behavior is guided by
the invoked symbolic processes; however, the intricacy
of the behavior demands that it be practiced in order for
it to be performed accurately. Motor reproduction pro
cesses and capacities are necessary for such practice.
Finally, incentive and motivational processes may
interact with other subprocesses to influence observa
tional learning, and they are particularly important in
regulating performance. Incentive may play a large role
in determining whether or not attention is focused upon
the appropriate stimulus events being modeled. Similarly,
incentive may play a determining role in retention pro
cesses by leading to deliberate efforts to remember what
is being observed. Clearly, it is difficult to separate
these subprocesses, but three of them— attention,
m
retention, and incentive— are of special interest in this
research and will therefore be given additional coverage
at this time.
CHAPTER III
ATTENTION PROCESSES
As was stated previously, attention is required in
order for sensory registration to take place, and there
fore attentional processes are crucial to observational
learning. Incentive and motivational manipulations pre
sumably may interact with attentional processes to focus
and direct the observing response towards the relevant
modeling stimuli. Additionally, characteristics of both
the model and the observer may be argued to influence the
level of attention.
Bandura (1969) attributes particular importance to
the attention arousing level of different models in at
tempting to verify the important role of attention. He
reasons that because differing reinforcement histories
are associated with imitating different models, certain
models come to receive more attention from the observer.
A process of generalization occurs such that after being
reinforced for imitating models who possess certain char
acteristics an individual learns that other models with
similar characteristics may be expected to demonstrate
behaviors which will be rewarded if copied. For this
15
16
reason models who are highly competent at a given task
are more likely to be Imitated than are their less suc
cessful counterparts (Gelfand, 1962; Mausner, 195^a,b;
Mausner and Bloch, 1957; Rosenbaum and Tucker, 1962).
Gelfand (1962) related both premeasures of self
esteem taken from questionnaire responses and self-esteem
manipulated by doing better or worse than a partner on
four tasks with dependent measures of imitating the
partner's choice on 20 picture preference trials and
verbal conditioning. Results showed that the manipula
tions made no difference between groups on the degree to
which they were operantly conditioned to give desired
verbal responses. However, on the picture preference
test Ss who experienced failure relative to their partner
gave significantly more imitative responses than did Ss
who were relatively successful.
Mausner (195^a,b) used length of line Judgments to
test for imitation of estimates as a function of rein
forcement history (195^a), or as a function of whether or
not their partner had been successful at a relevant similar
task (195^b). In the first study, results indicated that
Ss who had not been reinforced for their Judgments when
pretested alone were significantly more likely to shift
their estimates in the direction of their partner when
tested together. The second experiment (195**b) exposed
Ss to a partner who exhibited a great deal of skill or a
17
great lack of skill in a similar length estimate task.
Subjects who had a successful partner (confederate) were
significantly more likely to shift their estimates of line
length towards his. Mausner and Bloch (1957) essentially
replicated the (195^b) study Just discussed.
Rosenbaum and Tucker (1962) used hypothetical horse
races and found that when Ss are reinforced (by being
correct) whenever they copy their partner's prediction of
race outcome, the partner's level of success makes a sig
nificant difference in amount of imitating, such that as
the model's accuracy increases so does the number of
matching predictions by S.
These experiments clearly show distinct effects of
certain variables upon level of matching behavior; how
ever, it is not at all clear that these are mediated by
differential levels of attention. In no case were Ss ex
posed to multiple models to whom they could attend differ
ently, and in fact, none of these experiments involved the
learning of new behavioral responses. The nature of these
studies is such that differential attention may be called
in post-hoc to account for the findings; however, more
parsimonious explanations may attribute the results di
rectly to the variables which were manipulated. Addi
tionally, the data from these studies appears to be more
directly relevant to factors controlling performance, as
opposed to the factors relevant for learning.
Mausner (1953) and Lefkowitz, Blake, and Mouton
(1955) manipulated the status of the model, and found
that this led to differing amounts of imitation by their
Ss. However, again no new behaviors were learned—
Mausner's Ss were in a picture preference task, and Lef
kowitz et al. had their model and Ss crossing an intersec
tion on foot in a field experiment. Both of these studies
added an Interesting dimension in that Mausner's model was
always wrong even though he was a purported expert for one
group, and still more imitation occurred as his status
rose. Lefkowitz et al.'s model was breaking the law, but
when his attire indicated increased status he was more
frequently imitated. Nevertheless, attention seems a
less likely explanation for this effect than does the
expectation by the Ss that they would gain reward or avoid
punishment by imitation.
Other investigators have related other model char
acteristics to imitation. Bandura and Kupers (1964) found
that adult models were more effective in transmitting pat
terns of self-reinforcement than were peer models. How
ever, Bandura, Grusec, and Menlove (1967) found that when
presented with both an adult and a peer model who set
discrepant standards for self-reinforcement Ss imitated
the peer model. Jacubzak and Walters (1959) found that
adults were more effective than peers in leading Ss to
change their estimates of the autokinetic effect. Hicks
19
(1965) found that on novel aggressive behaviors a peer
model led to more imitation immediately after exposure,
whereas after 6 months Ss exposed to an adult model were
more aggressive. This study also included a learning
analysis on the delayed post-test, and this resulted in
the finding that the Ss had retained more aggressive be
havior than they had performed. Such a finding leads
logically to the conclusion that differential amounts of
imitation are not the result of differential attention,
but result instead from differing expectations of reward
and punishment, which are important controlling elements
in performance, but probably less important in acquisi
tion.
Many studies have used the sex of the model as a
variable, and found significant, although not consistent
results (Bandura, Ross, and Ross, 1963; Maccoby and Wilson,
1957; Rosenblith, 1959, 1961). On the other hand, at
least one experiment has found no effect of the model's
sex (Bandura and Kupers, 1964). Certainly the most in
teresting of these experiments with regard to learning new
material and the effect of attention is the one done by
Maccoby and Wilson. This was actually two experiments
which consisted of showing children a film and testing
for identification with film characters, and also looking
at the degree to which a character's behavior in the film
was learned by the Ss. The first experiment manipulated
the status of two models In the film. Results Indicated
that Ss did differentially identify with one model or the
other, and presumably therefore paid more attention to
his role. However, the Ss showed no differences on a
recognition test of which model had done what in the film.
The second film varied the sex of the models and found
that Ss did identify with the same sex model. Further,
using recognition, recall, and reconstruction tests, the
investigators found that on 12 out of 22 classes of the
models' behavior Ss retained the behavior of the same sex
model to a higher degree. In particular, boys remembered
aggressive content better, and girls remembered better
the scenes in which the two models interacted. Results
of these two studies are not consistent as regards atten
tion processes, since identification, which presumably
increases attention, led to increased learning in one
study, but not in the other.
Two characteristics of a model which would seem
likely to have an effect on the amount of attention paid
to the model are the variables of nurturance and power.
Bandura and Huston (1961) manipulated the nurturance of
the relationship between the model and the subject and
then placed them in a discrimination learning task such
as was introduced by Miller and Dollard (19^1). They
added to the experimental situation several incidental
behaviors demonstrated by the model. Their results
21
Indicated that Ss who had experienced a highly nurturant
relationship with the model were much more likely to
imitate the Incidental behaviors, even though there was
no difference in learning the discrimination task.
Grusec and Mischel (1966) and Mischel and Grusec
(1966) varied both rewardingness (nurturance) and power
(by means of future control of resources) of the model,
and had the model demonstrate two aversive and two neutral
behaviors while playing a game with the Ss. Mischel and
Grusec found that a nurturant relationship led to in
creased imitation of the neutral behaviors, and a power
ful model was imitated more often on both aversive and
neutral behaviors. Grusec and Mischel added a learning
analysis to their design and found that a model who was
high on both nurturance and power, as compared to a model
low on both, was more effective in total amount of learn
ing of imitative behaviors.
These experiments did employ models who were ex
hibiting novel behaviors, and at least one of them dis
tinguished between learning and performance. They furnish
evidence which suggests that certain variables may affect
learning by altering the amount of attention focused on
the model. However, they do not manipulate attention
per se, and they are made less definitive by two recent
experiments by Baron (1970a,b). These studies found that
attractiveness of the model, which presumably functions
22
similarly to his rewardingness to influence his affective
value for the observer, did not affect S performance on
aggression (1970a) or on gambling behavior (1970b). These
are not new behaviors, however, and the evidence provided
by Grusec and Mischel, and Bandura and Huston remains very
suggestive of the importance of attention in observational
learning.
Because it has commonly been observed that differ
ent people learn different amounts in identical observa
tional situations, many investigators have studied vari
ables relating to the observers themselves. The appli
cability of these experiments to attention and observa
tional learning varies a great deal. Several investiga
tors have related dependency to imitation. Jacubzak and
Walters (1959) found highly dependent Ss to be more easily
influenced by a partner on judgments of the autokinetic
effect. Kagan and Mussen (1956) assessed dependency by
TAT themes, and found high need for dependency to be sig
nificantly related to conformity in an Aschtype experiment.
These studies are, of course, open to criticism regarding
their applicability to the discussion of attention processes
in observational learning, because they do not involve the
acquisition of new responses, and they do not appear to
differentially manipulate attention.
One study by Ross (1966), however, does make a
strongly suggestive case for the role of dependence as
it affects attention and thereby alters observational
learning. An adult model taught 26 high-dependent and 26
low-dependent children how to run a post office, and
while doing so modeled some partially relevant and some
totally irrelevant behaviors. These behaviors were
labeled intentional and incidental learning respectively.
Results of the study were that the low-dependent Ss per
formed more imitation of intentional material, but less
of the incidental material than did the high-dependent
Ss. These findings suggest that low-dependent Ss may
attend only to those elements of the modeled behavior
relevant to the task at hand (learning how to run a post
office), whereas high-dependent Ss attend to all of the
model's behavior. Unfortunately, attention is again not
manipulated per se, and no learning analysis was included
in the experimental design.
Other studies have manipulated or used measures of
other variables and related these to imitation. Self
esteem (de Charms and Rosenbaum, I960; Gelfand, 1962), and
competence (Kanareff and Lanzetta, I960; Mausner and
Bloch, 1957) have been related to imitation, but these
studies did not involve learning new behavior, and did
not appear to affect attention levels.
One factor which does seem to potentially be very
powerful in affecting attentional processes is emotional
arousal. Easterbrook (1959) argued that an increase in
24
emotional arousal shortens the range of cues which a sub
ject utilizes, i.e. that attentional breadth is narrowed.
Whether or not this facilitates or inhibits learning is
largely dependent upon the complexity of the new behavior.
If it is relatively simple, then learning is likely to be
facilitated by emotional arousal because irrelevant cues
are not attended to. As the task becomes more complex
the number of relevant cues increases, and in these sit
uations learning may be hampered because attention becomes
too narrowly focused. Kausler and Trapp (I960) suggested
a revision of Easterbrooks' proposition by hypothesizing
about the importance of whether emotional arousal was
generalized or goal directed, and whether the irrelevant
cues were imbedded in the relevant cues, or were periph
eral.
Bandura and Rosenthal (1966) conducted an experiment
relating to vicarious emotional conditioning, and found
that some Ss reported attempting to ignore the situation
or imagine some other events when they became too uncom
fortable about the vicarious conditioning, which was
aversive in nature. This obviously was not an experiment
in which new overt behavioral responses were being learned;
however, it does strongly suggest that emotional arousal
may influence attentional processes as was proposed by
Easterbrook, and by Kausler and Trapp.
Several studies then (Bandura and Huston, 1961;
Grusec and Mischel, 1966; Mischel and Grusec, 1966; Ross,
1966) have provided suggestive evidence that variations
In attention may be caused by attributes of the model,
of the observer, and of their relationship, and these
variations may in turn result in different amounts of
observational learning. None of these experiments, how
ever, have manipulated attention directly. The experiment
being reported here did Just that, varying attention over
three levels, and testing for the effect of this manipu
lation upon the learning of novel behaviors. Further,
retention processes in observational learning were inves
tigated, and therefore it is now appropriate to focus
upon these processes.
CHAPTER IV
RETENTION PROCESSES
As has been previously pointed out a great many
new behaviors, especially many social behaviors, are
learned observationally, and performed later when the
modeling cues are not present. According to Bandura's
theory of modeling, such delayed performance necessarily
implies that the relevant behavioral cues are retained in
some symbolic form, which seems to be a logically neces
sary inference. Clearly, in order for such delayed repro
duction to take place some processes, other than merely
attending to cues which are present, must be occurring.
This is what is commonly referred to as memory, which is
itself the topic of a vast array to psychological liter
ature, and which involves many unresolved questions.
For the purposes of the discussion at hand it is
not fruitful to delve into these questions regarding the
nature of memory, but rather the present goal is to con
sider operations which enhance retention processes and
thereby allow for the delayed reproduction of observa
tionally learned behaviors. For example, rehearsal has
been shown to aid in the learning of new behaviors.
26
Margolius and Sheffield (1961) used training films in
combination with different patterns or schedules of live
practice to teach male undergraduates how to assemble an
automobile carburetor. Their findings were that when
practice was interspersed between demonstration units,
which were empirically found to be more or less natural
clusters, learning of the assembly process was better
than when larger demonstration units were used, or when
the whole film was shown without interruption for practice.
Unfortunately, the experimental design in this situation
did not include a control group which did not practice at
all, since it was assumed that practice did add to the
effects of mere demonstration alone.
With respect to the discussion of the role of
symbolic processes of retention, studies which compare
overt rehearsal with covert rehearsal are particularly
interesting. Perry (1939) compared overt and covert
rehearsal of five different tasks (tapping the bottom of
a hole, mirror tracing, card sorting, digit-symbol sub
stitution, and a pegboard task) and found that their
relative efficacy in enhancing learning varied from task
to task. His experimental procedure did not involve
learning these tasks by observational means. Twining
(193*0 found that in performance on a ring toss game real
practice as opposed to imaginal practice as opposed to
no practice at all resulted in performance increments of
28
137 percent, 36 percent, and no percent respectively.
Vandell, Davis, and Clugston (19^3) found that imaginary
practice did not differ from real practice on performance
tests of dart throwing or shooting basketballs. However,
these experiments also did not involve observational learn
ing, if, indeed, they involved learning at all.
Michael and Maccoby (1961) compared overt and covert
practice on the degree to which they aided learning from
a film. The film's topic was patterns of survival in case
of nuclear attack. Overt practice meant writing down
answers to factual questions which were sampled from a
final test on the subject matter of the film. Covert
practice meant merely thinking the answers to the same
questions. Results indicated that there was no signifi
cant difference between the efficacy of the practice
methods on a final written test. This study, however,
does not answer the question of whether or not people are
better able to take appropriate survival actions when they
practice them as opposed to when they think, write, or
talk about them.
McGuire (1961) and Kanner and Sulzer (1961) com
pared overt and covert practice in paired-associate type
experiments. Both studies found that covert rehearsal
succeeded at least as well as practice of writing down
the paired associate, and, indeed, McGuire found that at
29
increased presentation speeds covert clearly exceeded
overt practice as a means of improving paired associate
learning. Unfortunately, it is not clear to what degree
these verbal learning paradigms are analogous to social
learning.
Of particular interest to Bandura's theory of
observational learning and to the present proposal are
the symbolic coding operations which were discussed
earlier. It is Bandura's assertion that coding operations
do more to facilitate long-term retention of modeled
events than does rehearsal. Coding is an area of psycho
logical investigation which is related to the larger
phenomenon of memory, and which has received particular
attention from verbal learning investigators (Mandler,
1968; Tulving, 1968). Miller (1956) proposed a process of
coding to account for the discrepancy between the fact
that the span of immediate memory is 7+2 items, whereas
people are obviously able to store and retrieve a great
deal more information. Miller's unitization hypothesis
suggested that people categorize items into units which
he called "chunks." These "chunks" were informationally
rich categories which could themselves be categorized into
even richer chunks, making for an infinitely expanding
information storage potential. This coding system, Miller
suggested, was primarily verbal; however, he included the
possibility that imaginal codes might exist.
30
Numerous empirical investigations (Mandler, 1967;
Tulving, 1966) have shown that people do, indeed, organize
information into units which may be an aid in free recall.
Nevertheless, in spite of the fact that many authors have
spoken of coding operations, no really explicit definition
of the coding process has been proposed. Miller's orig
inal description suggested that:
The input is given in a code that contains many
chunks with a few bits per chunk. The operator
recodes the input into another code that contains
fewer chunks with more bits per chunk . . . there
are many ways to do this, but probably the simplest
is to group the input events, apply a new name to
the group, and then remember the new name rather
than the original input events. (1956, p. 93)
Two pieces of experimental research have been con
ducted to investigate the role of coding processes in ob
servational learning. Bandura, Grusec, and Menlove (1966)
exposed children to a film of a model performing several
complex sequences of behavior. All the children were
instructed to watch the film carefully. Additionally, one
group of Ss were told to verbalize each of the model's
behaviors. A second group was told to count rapidly while
they were watching the film in order to prevent them from
covertly coding the modeling cues. A third group served
as a control, and only watched the film attentively. The
results of this study showed that the Ss who verbalized
the model's behavior learned more of that behavior than
did those Ss who only watched the model, and these Ss in
31
turn learned more than those who were engaged in counting
during the viewing of the film. Unfortunately, in spite
of the fact that all the children were instructed to pay
close attention to the modeling film, equality of atten
tional processes could not be guaranteed. As a result the
experiment's findings are open to serious questioning re
garding the degree to which they are attributable to atten
tion rather than retention processes. Clearly, it was not
the intent of the Es to manipulate attention; however,
there is certainly the very distinct possibility that the
instructions to the groups were in fact differential in
structions to attend, especially since the Ss were young
children.
Gerst (1971) was satisfied that Bandura, Grusec,
and Menlove (1966) had demonstrated the importance of
coding in observational learning, although the possible
confounding variable of attention has been pointed out
above. Gerst proceeded to test the relative efficacy of
imaginal coding (imagining that one is performing the
modeled behavior), verbal coding (describing aloud the
modeling cues), summary labeling (subjectively developed
labels summarizing the constituent elements of the model
ing), and a counting control on their ability to facil
itate observational learning. The material to be learned
was ten sequences of motor responses from the manual
32
language of the deaf. Ss watched the model perform the
hand signals, then during a one-minute post-exposure
period they employed one of the four coding procedures
listed above. At the end of this one-minute coding period
the S was asked to reproduce the modeled behavior. He then
performed a fill-in task for 15 minutes, and at the end of
this period, he was asked to reproduce as many of the
modeled behaviors as he could in a free recall-type
situation.
Results of Gerst's study indicated that on the
immediate post-test all the coding groups were superior
to the controls, and the summary labeling and imaginal
coding groups were superior to the verbal description
group. On the delayed test, all groups showed a signif
icant decrease in number of reproduced responses, and only
the summary labeling group differed from the controls.
Problems arise with regard to the import of this
study as a demonstration of the role of coding processes
in observational learning. Procedurally, there is the
question of whether or not the fact that the coding took
place after the modeling cues were presented implies that
some preliminary coding occurred during observation. Also,
attention remains as a possible confounding variable since
coding instructions may have simply provided a guiding
structure for attention. The results of the study showed
that on immediate reproduction all groups demonstrated
over 55 percent of the modeled responses, but on the
delayed test no group exceeded 45 percent of the modeled
responses. If coding was blocked in the counting control,
then coding obviously is not a necessary process in obser
vational learning. Whether or not this is true is an
important theoretical question which remains to be an
swered. The experiment reported herein attempts to provide
additional information relevant to this issue.
CHAPTER V
INCENTIVE EFFECTS
Finally, the experiment reported here investigated
the role of incentives in observational learning. Con
sideration must immediately return to the distinction
that is made between learning and performance, if one is
to clearly conceptualize the role of incentive in observa
tional learning. As mentioned previously, a great deal
of evidence has shown that when Ss are rewarded for imi
tation of a model, the frequency of imitation increases.
In other words, Ss who have an Incentive or motivation
to copy a model are more likely to do so. However, this
finding has been subjected to experimental investigation
(Bandura, 1965b) which clearly suggests that the effect
of such incentives is seen in performance rather than
learning. In that experiment apparent differences in the
level of Imitation of aggression between groups disappeared
when the Ss were offered rewards for reproducing the be
havior they had seen. Moreover, Bandura (1965a) has also
discussed the phenomenon of modeling with particular
emphasis on the fact that observational learning can
occur without the observer being rewarded. This finding
3H
35
apparently was Important in his formulation of a conti-
guity-mediational theory of modeling, as opposed to rein
forcement theories.
Whereas the very Important role of incentives in
imitative performance has been very clearly established,
it is still a largely unanswered question as to how impor
tant incentives are in the process of observational learn
ing. Bandura (1969) asserts the following:
Incentive variables not only regulate the overt
expression of matching behavior, but they also
affect observational learning by exerting selective
control over the modeling cues to which a person
is most likely to be attentive. Further, they
facilitate selective retention by activating
• deliberate coding and rehearsal of modeled responses
that have high utilitarian value, (p. 142)
This statement not only makes a claim for the role
of incentives in observational learning, but also points
out the difficulty of assessing the importance of that
role. Specifically, it is extremely difficult to separate
incentive processes from either attention processes or
retention processes, as Bandura has pointed out above.
It may well be that covert processes such as attention
and retention are dependent on a more fundamental process
of Incentive or motivation, in much the same way that overt
performance depends upon reinforcement. However, incentive
may be a cognitive variable as regards covert processes,
rather than the extrinsic reinforcer usually associated
with overt behavior.
Perhaps because of the aforementioned difficulty
of separating incentive, attention, and retention pro
cesses little effort has been made to test the signifi
cance of having an incentive to learn from a model.
Bandura, Grusec, and Menlove (1966) did conduct a modeling
experiment using children as Ss, in which half of the
Ss were told before watching a film that they "would be
asked to demonstrate what they had learned, and that they
would receive candy treats for each item they reproduced
correctly." These Ss were considered to have been given
an incentive set, as opposed to the other half of the Ss
who were provided no incentive. Contrary to the Es'
predictions, the incentive manipulation produced no sig
nificant difference. In fact, the incentive group did
slightly less well than their counterparts who lacked
incentive. Questionnaire data indicated that the incen
tive manipulation may even have interfered with learning
by arousing achievement anxieties in some of the children.
Therefore, it remains to be demonstrated that an
introduction of an incentive set prior to observation of
the model influences the level of observational learning.
The present research sought to address this issue along
with the role of attention and retention processes.
CHAPTER VI
STATEMENT OF THE PROBLEM
The major purpose of the present study was to
investigate the relative contribution to observational
learning made by the sub-processes of attention, reten
tion, and incentive. As has been noted, several studies
(Grusec and Mischel, 1966; Mischel and Grusec, 1966;
Ross, 1966) have used observational learning tasks to
demonstrate that various model and observer characteris
tics are influential in the amount of learning that takes
place, and these studies have offered attentional differ
ences as a post-hoc explanation for their findings. No
experiment, however, has systematically and intentionally
manipulated attention. Further, two experiments (Bandura,
Grusec, and Menlove, 1966; Gerst, 1971) have investigated
retention processes. Both of these studies have failed
to equate attention in the coding and non-coding groups.
Similarly they have failed to establish that retention
(in the form of some type of coding operation) is a neces
sary sub-process in observational learning, since even
those Ss who were supposedly prevented from coding did
learn some of the modeled behaviors. Lastly, only one
37
study from the modeling literature has considered the
effect of incentive on observational learning in contrast
to imitative performance (Bandura, Grusec, and Menlove,
1966). The results of this study indicate that an incen
tive to learn produced no significant effect on the amount
of imitative responding.
The present study employed a 3 x 3 x 2 factorial
design in order to test hypotheses about the effects of
attention, retention, and incentive variables on observa
tional learning. Three levels of attention were induced
using a set of instructions designed to create discomfort
by means of arousing cognitive dissonance in Ss who did
not pay close attention on the one hand, and in Ss who
did pay close attention on the other hand. The Ss exposed
to the former instructional set were expected to have
their level of attention increased in order to avoid the
dissonance, and the Ss given the latter instructions were
expected to pay less attention for the same reason. A
third group of Ss served as a control, and was asked only
to "pay close attention."
Three levels of retention processes were induced
by instructing Ss to use different coding procedures.
In order to demonstrate that coding is necessary for
observational learning it is necessary to show that learn
ing is prevented by blocking the coding operation.
Therefore, one group was instructed to observe the model,
and accurately code the model's behavior, using a verbal
coding system. A second group was instructed to observe
the model, but to inaccurately code the behavior by ver
bally describing some behavior that was essentially simi
lar to but not exactly like that of the model. This
group was thus forced to attend to the modeling cues in
order to see what the model did, but was prevented from
accurately coding the modeled behavior, and, therefore,
should have been seriously impaired in efforts to repro
duce the behavior. A third group was simply instructed
to attend to the behavior without being told to apply any
particular coding system.
The third factor in the design was the incentive
set under which the S was viewing the modeling tape. One-
half of the Ss were told prior to watching the model that
E was interested in how well they could learn and remember
what they saw the model do. This was the incentive group.
The other half of the Ss were not given any information
at all regarding the issue of learning and remembering
what the model did. Neither group was told specifically
that they would be asked to imitate the model.
The specific hypotheses tested by this study per
tained to the three factors as follows:
40
Attention— If attention processes are determinants
of the amount of observational learning which occurs, the
group whose attention level was enhanced would: (a) be
able to produce more of the modeled behavior than would
the control group, and the controls would in turn learn
more than those Ss whose attention was distracted; (b)
make fewer errors in reproducing the model's behavior
than would the control group, who would in turn make
fewer errors than those Ss whose attention was distracted.
Retention— In accord with the premise that coding
is essential for the retention of modeled behavior the
accurate coders would: (a) be able to reproduce more of
that behavior than the controls, who would in turn be able
to reproduce more behavior than the inaccurate coders;
(b) make fewer errors in reproducing the model's behavior
than would the control group Ss, who would in turn have
made fewer errors than the Ss who coded the modeled behav
ior inaccurately.
Incentive— If an incentive set does influence how
well Ss attend to the model, or how much effort they put
into remembering what the model does, those Ss who were
told that they would be asked how well they could remember
what the model did would: (a) be able to reproduce sig
nificantly more of the modeled behavior than the uninformed
Ss; (b) make fewer errors during their imitative perform
ance than the Ss who were not given incentive-inducing
instructions.
CHAPTER VII
METHOD
Subjects
Undergraduates enrolled in introductory psychology
courses at the University of Southern California partici
pated in the present study in partial fulfillment of a
course requirement. A total of 90 Ss were randomly
assigned to 18 groups, for a total of five Ss per cell.
Each cell contained three males and two females, so that
a check could be run for sex differences.
Apparatus
Videotape equipment: A videotape camera, recorder,
and monitor were used to present the modeling cues to the
Ss, and also to record the Ss1 attempted imitation of the
modeled behavior. Use of this equipment provided for the
standardized presentation of the modeling sequence to each
S. It also allowed for accurate and independent scoring
of the imitative efforts of each S by virtue of the fact
that these efforts were recorded, and therefore could be
viewed at a later time by individual scorers as often as
was necessary to ensure accuracy.
42
1*3
Modeling tape: The modeling cues were presented
to the Ss on the videotape monitor. The modeling tape
consisted of eight segments or sequences, with ten seconds
of blank tape between segments. The entire modeling
presentation lasted slightly over six minutes. Each of
the eight individual segments lasted 35-^0 seconds, and
showed a female model (the same model for each segment)
performing novel behaviors with one or two objects. Dif
ferent objects were used for each segment, and different
behaviors were performed with each object. Thus, segments
were unique in terms of the behaviors performed and the
objects used. The purpose of this use of "props,'1 as the
objects may be called, was based on inferences drawn from
the contiguity-mediational theory of modeling. Since
external or internal cues elicit perceptual responses,
and since this series of symbolic responses guides overt
behavior, these props served as external cues which facil
itated recall and performance by triggering the chain of
perceptual responses. The props and the modeled behavior
associated with each of them are specified in Appendix I,
and are presented in the order that they appeared in the
modeling tape.
Procedure
The study employed three levels of attention, three
coding processes, and two levels of incentive. Subjects
entered the experimental room, and were seated in front
of and directly facing a television screen on which the
videotape was to be shown. It was explained to them
that the E was interested in studying people's reactions
to what they see, and what things determine how people
react to what they see. The Ss were told that E was
particularly interested in how the content of what they
saw affected their reactions, and how what they were doing
as they watched the film affected their reactions. They
were further told that by their "reactions" the E was
referring to their impressions or evaluations of the
film along various dimensions. After having been told
this alleged rationale for the study (see Appendix II),
the Ss were exposed to the following experimental manipu
lations .
Retention processes— coding systems; Coding sys
tems were manipulated according to the degree of accuracy
embodied in the code. One-third of the Ss were instructed
to use an accurate code, which simply means that they were
told to accurately describe the model's behavior as it was
going on. Such coding should result in a high level of
retention according to contigulty-mediational theory.
Another one-third of the Ss were instructed to use an
inaccurate code, i.e. they were asked to describe what
they saw the model do with the different objects, but to
H 5
change some of the details of what she was doing so that
the description was slightly inaccurate. The theory would
predict that such behavior on the part of the observers
seriously interferes with retention and subsequent accurate
reproduction of the modeled behavior. The final one-third
of the Ss were not instructed to use any coding mechanism
in particular. Prom contiguity-mediational theory one
might hypothesize that these Ss will do some coding,
probably imaginal coding, but presumably they will be less
efficient at coding than those Ss who were instructed to
use a specific mechanism. (The details of the instruc
tions to Ss in each of these groups may be found in
Appendix III.)
Incentive effects: After hearing the coding in
structions the Ss were exposed to the incentive manipula
tion. One-half of the Ss were given no further informa
tion which might indicate that E was interested in how
well they remembered the modeling tape. This "no incen
tive" group was in contrast to the other half of the Ss,
who were told at this time that E was also interested in
how well people could learn and remember what they saw
the model do, and that E would ask them about this at a
later time. It should be noted that this "incentive"
group was not specifically told that they would be asked
to imitate the model (see Appendix IV, for text of
46
instructions to the incentive group). This was in con
trast to Ss in a previous study who were told that they
would be asked to imitate the model. This fact was in
tentionally not specified in the hope that the Ss would
not become anxious at the prospect of performing the novel
and unusual behaviors which they observed.
Attention processes: Finally, after being exposed
to the incentive manipulation, and having heard the in
structions as to how they should code what they are watch
ing, the Ss were exposed to informative statements de
signed to manipulate their level of attention. In order
to increase attention one-third of the Ss were told that
E had found in the past that college students were able
to pay close attention to the film because they were young
and intelligent, and therefore able to concentrate well
and focus their attention closely. As a result of this
piece of information the Ss were placed in a situation in
which the failure to attend closely would indicate that
they were not as intelligent as their peers. Conversely,
another one-third of the Ss were told that, because they
were young and intelligent, college students often found
it unnecessary and uninteresting to pay close attention.
This information meant that for them to attend closely to
the tape was indicative of lower intelligence than their
peers possessed. This information was designed to decrease
47
their attention level. The final third of the Ss were
simply requested to pay close attention, and served as a
control group for the two preceding conditions. A post-
experimental questionnaire was used to attempt to deter
mine the impact of the attention manipulation. (The text
of the attention manipulation statements is contained in
Appendix V.)
After each S had been given the instructions appro
priate to his group, he was shown the modeling tape. The
entire modeling tape lasted slightly over six minutes,
and was shown to the S one time while he attended and
coded appropriately. Between each modeling sequence
there were ten seconds of blank screen to allow the S to
discriminate the sequences, without simultaneously giving
him a great deal of time for covert rehearsal.
At the end of the exposure to the videotape Ss were
led to another room close to the experimental room and
asked to fill out the Orientation Inventory (Bass, 1962).
This short questionnaire not only provided desirable in
formation about certain characteristics of the Ss, but
also served to allow for a short period of time between
observation and attempted reproduction of the modeling
cues, thus insuring that a reproduction of those cues
involved some retention processes. Additionally, previous
studies (Bandura et al., 1966; Qerst, 1971) have deemed
it desirable to prevent covert rehearsal. Although such
48
rehearsal Implies the presence of symbolic mechanisms
from which the modeling is rehearsed, it remains desirable
to eliminate such practice in order to investigate coding
processes per se. and the Orientation Inventory served
this function by requiring enough attention to prevent
practice without simultaneously adding new input that
required remembering.
Having completed the Orientation Inventory, the Ss
then returned to the experimental room, in which they
found the props arrayed unsystematically about the room.
The E explained to them that he was now interested in
seeing how much of the modeled behavior they were able
to reproduce. The Ss were either given ten minutes to
reproduce as many of the modeling cues as they were able
to, or as long as it took them to reproduce all eight
sequences or decide that they could recall no more. In
fact, no S required the full ten minutes. Their behavior
was recorded on videotape so that scorers could indepen
dently observe and rate their performances. The Ss were
given no feedback during or after this behavioral test
as to the accuracy of their reproductions. (The direc
tions to the Ss during this test period are contained in
Appendix VI.)
Finally, after they had finished the recall test
the Ss were asked to fill out two short questionnaires,
49
which inquired into their expectations before and during
the experiment, and into several aspects of their behavior
and intentions during the time that they were observing
the modeling cues. (The questionnaires themselves are
found in Appendix VII.) Following completion of these
questionnaires the Ss were debriefed. They were told that
E was interested in two things: how much of the tape they
could imitate, and what their impressions were as revealed
by the questionnaire. The function of ORI was described,
as was the general experimental design, and all questions
were answered truthfully. On leaving each S was asked to
refrain from discussing his experience with friends and
classmates.
Dependent Measures
Of primary interest in this experiment was the
imitative behavior of the Ss, and the degree to which
they were able to accurately reproduce the model's behav
ior as a function of different treatment conditions.
Several dependent measures were used singly and in combi
nation to assess this topic of primary interest.
The Ss were asked to imitate what they saw the
model do with the various props which she used. Her
behavior was broken down into eight sequences (each with
a different prop), which contained five essential elements
each. In order to score the Ss1 attempted behavioral
50
reproduction two scorers were used, each scoring 60 of
the 90 Ss, and thus providing an overlap on 30 Ss so that
interrater reliability could be tested. The scorers were
given written descriptions of the five essential elements
of each segment, and they also viewed the modeling tape
until they felt completely familiar with it. When scoring
the tapes containing the Ss* imitative behavior the scorers
used a checklist containing a notation for each element in
each segment. (The checklist is seen in Appendix VIII.)
As any individual S moved through his imitative
responses the scorers observed which prop he was using,
and thus identified which modeling segment he was supposed
to be imitating. For each element of the sequence they
scored his performance according to the code depicted in
the following table:
Table I
Notation for
scoring of an
element Meaning of notation
P
(left blank)
W Element wholly and accur
ately reproduced
Element partially reproduced
S did not attempt imitation
of this element
51
Appendix I specifies the criteria which were used to de
termine whether or not any particular element was wholly
(W), partially (P), or not at all (left blank) reproduced.
Based on this scoring system, the first dependent
variable score for an S was the number of "Ws," or whole
reproductions, he attained over all segments. The second
dependent variable was the number of "Ps," partial repro
ductions made by S in all segments. Summing the number
of "Ws" and "Ps" gave an indication of S's overall imita
tive efforts, and this total over all segments was used
as the third dependent measure.
The scoring sheet also contained a space in which
notation could be made of whether or not a sequence, which
was reproduced with all five elements, had the elements
arranged in the same relative order as did the modeling
tape. If this was the case, the scorers simply put a
check mark in that space. The total number of complete
and properly ordered segments which S reproduced was a
fourth dependent measure of his observational learning.
Additionally, one piece of previous research
(Bandura, Grusec, and Menlove, 1966) made anecdotal refer
ence to the fact that some of the Ss interchanged elements
from different behaviors. The present experiment sought
to use such "errors" as a measure of observational learn
ing. To this end the scoring sheets provided two spaces
at the end of each sequence. The first of these was used
52
for a simple frequency count by the scorers of the number
of novel behaviors (i.e., behaviors unlike anything the
model had done with any prop) produced by S with the props
in that segment. The second space was provided for a
frequency count by the scorers of the number of borrowed
behaviors (i.e., behaviors done by the model, but with a
prop other than the one used by S) produced by S with the
props in that segment. Thus, the total number of novel
behaviors produced by S in all sequences provided a fifth
dependent variable, and the total of borrowed behaviors
comprised a sixth. The sum of novel and borrowed behav
iors gave an indication of the number of "errors" by an
S overall, and this sum was labeled "extraneous" behaviors
and used as a seventh dependent measure.
Additional data were collected. The time spent
between the end of the observation period and the begin
ning of the test period was recorded to investigate the
possibility that accuracy of imitation varies with the
amount of time that reproduction is delayed. The Orienta
tion Inventory, which was designed to indicate whether
the respondent is more concerned with social interactions,
working on the task at hand, or personal achievement, was
administered. The effect of the independent variables on
the three scales of this inventory was tested, and the
scales were correlated with other dependent variables.
Further, the questionnaire data provided information re
garding expectations of the Ss, intentions to learn,
ratings of the film, ratings of various aspects of the
experiment, and subjective reactions to the effect of the
attention, retention, and incentive manipulations. This
information was tested regarding the effects of these
manipulations on the questionnaire variables, and was
used to gain correlational data regarding these variables
and the degree of observational learning.
CHAPTER VIII
RESULTS
A3x3x2x2 factorial design was used in this
experiment, and the data were treated statistically by a
multivariate analysis of variance. This included three
levels of attention, three means of coding for retention
purposes, two incentive levels, and two levels for test
ing sex differences.
Several dependent measures were used to assess
observational learning, and the scores for individual
subjects on each of the seven dependent variables of
primary interest are given in Appendix IX. As was de
scribed above, the scoring was done by two independent
observers, who were both highly familiar with the modeling
cues, and who rated the S's imitative responses from video
tapes of those efforts. On 30 of the 90 Ss who partici
pated in the experiment the two scorers both rated the
video tapes, but of course did so independently of one
another. Analysis of their scoring on these overlapping
Ss showed that the scorers were in perfect agreement on
97 percent of the items scored.
54
55
Analysis of the data relating to the amount of
observational learning which took place is not simple
to report since several dependent measures were used to
assess this issue. In general, it may be said that the
attempted attention manipulation produced no significant
effect between groups. The retention factor, on the
other hand, was a very valuable source of variance on
most of the primary dependent variables. Further, the
incentive manipulation also proved to be a significant
source of variance on several of the dependent measures
used. Sex differences were not typically found, but did
appear in relation to one of the dependent variables of
primary interest. Interaction effects were found only
infrequently, and were confined entirely to the dependent
measures describing the number of errors made during imi
tation, i.e. the number of novel behaviors, the number of
borrowed behavioral elements from other sequences, and the
number of extraneous behaviors in total (novel plus bor
rowed). There was a two-way interaction (involving the
retention and sex factors) on two of the dependent mea
sures, a three-way interaction (involving the attention,
incentive, and sex factors) on two variables, and there
was a four-way interaction on one variable.
Attention Effects
The attempted attentlonal manipulation produced no
56
significant differences between groups on the number of
whole reproductions, partial reproductions, wholes plus
partials, or completely reproduced sequences (see
Tables 2-5). Similarly, no differences appeared in the
number of novel or borrowed behaviors in an imitative
sequence, nor in the total of novel plus borrowed, i.e.
extraneous behaviors (Tables 6-8). Additionally, the
informative statements designed to induce differing atten-
tional levels did not show significantly different effects
on scales of the Orientation Inventory, time spent work
ing on the inventory, or on any of the questionnaire data
regarding subjective evaluations of the film and of S's
reaction to the experiment as a whole.
Retention Effects
The manipulation of the retention factor by means
of different coding systems produced highly significant
differences between groups as was indicated by several
dependent variables. Tables 2 through 8 show the analysis
of variance results for the seven dependent variables of
primary interest. Perusal of these tables shows that on
all of these variables except one, i.e. the number of
partial reproductions, the retention manipulation produced
significant differences at the level of p <.001. That is
to say, the retention factor significantly affected the
number of whole reproductions, wholes plus partials,
57
Table 2
Analysis of Variance for the Number
of Whole Reproductions of Elements
Source Sum of
Squares
df F
P<
A 32.066 2
.759 .473
(Attention)
R 981.064 2 23.232 .001
(Retention)
I 173-610 1 8.222 .006
(Incentive)
S 9.869 1 .467 .497
(Sex)
AR 33-268 4 .394 .812
AI 61.088 2 1.447 .244
AS 56.470 2
1.337 .271
RI 15.022 2 .356 .702
RS 7.692 2 .182 .834
IS 7.824 1 .371 .545
ARI 21.576 4
.255 .905
ARS 74.916 4
.887 .478
AIS 22.004 2 .521 .597
RIS 85.068 2 2.014 .143
ARIS 62.384 4
.739
.570
Within 1140.534 54
58
Table 3
Analysis of Variance for the Number
of Partial Reproductions of Elements
Source Sum of
Squares
df F
P<
A 9.588 2 1.278 .287
(Attention)
R 1.688 2
.227 .797
(Retention)
I
6.945
1 1.870 .177
(Incentive)
S
.017 1 .004
.947
(Sex)
AR 2.312 4 .156 .960
AI 61.088 2 1.447 .639
AS 56.470 2
1.337 .413
RI 15.022 2 • 356 .702
RS 4.144 2 .558 .576
IS .046 1 .012
.911
ARI 21.576 4
.255 .932
ARS 6.412 4 1.240
.305
AIS 22.004 2 .521 .428
RIS 85.068 2 2.014 .561
ARIS 62.384 4
.739
.548
Within 200.448 54
59
Table 4
Analysis of Variance for the Number
of Whole Plus Partial Reproductions
Source Sum of
Squares
df P
P<
A 12.066 2
.289 .751
(Attention)
R 1113.796 2 26.632 .001
(Retention)
I 127.210 1 6.084
.017
(Incentive)
S 18.891
1
.903 .346
(Sex)
AR 35.332 4 .410 .800
AI 96.288 2 2.302 .110
AS
71.914
2 1.720
.189
RI 23.622 2
.565
.572
RS 12.904 2
.309 .736
IS 4.817 1 .230
.633
ARI 17.976 4
.215 .929
ARS 128.072 4
1.531
.206
AIS 11.100 2
.265
.768
RIS 87.598 2
2.095 .133
ARIS 101.132 4
1.209 .318
Within 1125.360 54
60
Table 5
Analysis of Variance for the Number
of Completely Reproduced Sequences
Source Sum of
Squares
df P
P<
A .066 2 .096
.909
(Attention)
R 7.200 2 8.888 .001
(Retention)
I 3.600 1 10.322 .002
(Incentive)
S
.417 1
1.195 .279
(Sex)
AR 1.332 4 .956 .439
AI .466 2
.669
.516
AS .212 2
.303
.740
RI 2.466 2 3.536 .036
RS .634 2 .908
.409
IS .669 1
1.917
.172
ARI 1.068 4
.765 .553
ARS 1.388 4 .996 .418
AIS 1.292 2
1.853 .167
RIS .070 2 .101 .904
ARIS 1.284 4 .921 .458
Within 18.522 54
61
Table 6
Analysis of Variance for the Number
of Novel Behaviors Produced
Source Sum of
Squares
df F
P<
A 5.088 2
1.913
.158
(Attention)
R 30.422 2
11.435
.001
(Retention)
I .044 1 .033 .856
(Incentive)
S 10.417 1
7.831 .007
(Sex)
AR 27.576 4
5.183
.001
AI 1.488 2 .364
.291
AS 3.634 2 1.366 .264
RI 2.688 2 1.011 .371
RS 10.744 2 4.038 .023
IS 1.780 1 1.338 .252
ARI 10.512 4
1.975
.111
ARS 8.588 4 1.614 .184
AIS 5.514 2
2.073
.136
RIS 12.736 2 4.787
.012
ARIS 9-952 4 1.870 .129
Within 71.743 54
62
Table 7
Analysis of Variance for the Number
of Borrowed Behaviors Produced
Source Sum of
Squares
df F
P<
A
(Attention)
4.422 2 1.082 .346
R
(Retention)
63.356 2 15.504 .001
I
(Incentive)
.136 1 .278 .714
S
(Sex)
.000 1 .000 1.000
AR 10.776 4
1.319 .275
AI 1.488 2 .364 .696
AS 3.634 2
.889 .417
RI .556 2 .136 .873
RS 9.644 2 2.360 .104
IS 2.963 1 1.450 .234
ARI 3.976 4
.487 .745
ARS 1.516 4 .190 .942
AIS 17.382 2
4.253 .019
RIS 2.592 2 .634 .534
ARIS
Within
65.096
110.322
4
54
7.965
.001
63
Table 8
Analysis of Variance for the
Total Number of Extraneous Behaviors
Source Sum of
Squares
df F
P<
A 10.556 2 1.776
.179
(Attention)
R 179.288 2 30.161 .001
(Retention)
I .100 1 .034 .855
(Incentive)
S 10.417 1 3.505 .067
(Sex)
AR 24.444 4 2.056 .099
AI 8.600 2 1.447
.244
AS 2.500 2 .421
.659
RI .800 2
.135
.874
RS 39.878 2 6.708 .002
IS .150 1 .050 .823
ARI 21.200 4 1.783 .146
ARS 3.724 4
.313
.868
AIS 40.010 2
6.731
.002
RIS 4.034 2 .679
.512
ARIS 123.188 4 10.362 .001
Within 160.434 54
6U
complete sequences, novel, borrowed, and extraneous be
haviors. Tables 9 and 10 show the mean number of wholes
plus partials and extraneous behaviors respectively In
each of the three levels of the retention factor. These
means were selected because of their summary nature re
garding the questions of (1) correct Imitation, and (2)
error. Tables 11 and 12 show this same data as a function
of all four factors, including sex.
More detailed analysis of this data indicates that
the pattern of significant effects was not exactly as
predicted. For example, accurate coders did not differ
from those Ss who employed no code at all on most of the
dependent measures. In fact, it was only on the variable
representing the number of completely reproduced sequences
that a significant difference appeared. In this case the
data showed that Ss using no code at all reproduced more
complete sequences (F]. 54 * 10.752, p. 002) than did the
accurate coders, which was contrary to prediction. The
accurate coders did not differ from the inaccurate coders
with regard to the number of complete sequences.
On other dependent variables the accurate coders
and the non-coders show an identical relationship to the
Ss who employed an inaccurate code. Both reproduced sig
nificantly more whole elements, more wholes plus partials,
fewer novels, fewer borroweds, and fewer extraneous
65
Table 9
Mean Number of Whole Plus Partial Imitations
as a Function of Type of Code Used
Type of Code Number of
Imitations
As
Percent
Accurate
26.033 65%
Inaccurate
18.333 49
No Code
25.533 63
Mean Number of
a Function
Table 10
Extraneous Behaviors as
of Type of Code Used
Type of Code Number of Extraneous
Behaviors
Accurate 1.633
Inaccurate 4.367
No Code 1.167
Table 11
Mean Number and Mean Percentage of Whole Plus Partial Imitations as
a Function of Attention, Retention, Incentive, and Sex
Increased Attention Normal Attention Decreased Attention
Accurate Inacc. Nc Accurate Inacc. No Accurate Inacc. FJo _
_________ Code______ Code____Code____Code______ Code____Code____Code Code Code_____X____
Incen
tive-
Male 26.667 24.333 27.667 26.333 18.000 26.000 26.000 15.000 23.333 23.506
.716 .608 .696 .653 .450 .650 .650 .375 .583 .597
Female 29.500 17.500 28.000 27.500 21.000 30.500 28.000 16.500 29.500 25.333
.737 .437 .700 .687 .525 -762 .700 .412 .737 .633
No In
centive
R a l e 22.333 15.333 27.333 24.000 16.667 26.000 24.000 19.667 22.000 21.926
.583 .383 .683 .600 .416 .650 .600 .491 .550 .548
Female 27.000 ' 19.000 15.500 23.000 20.000 24.000 28.500 17.500 27.000 22.380
.675 .475 .387 .575 .500 .600 .712 .437 .675 .559
26.883 19.041 24.625 25.200 18.916 26.625 26.625 17.166 25.457
.677 .475 .616 .630 .472 .665 .665 .429 .636
23.518 23.583 23.083
.589 .589 .577
o\
O N
Table 12
Mean Number of Extraneous Behaviors as a Function
of Attention, Retention, Incentive, and Sex
Increased Attention Normal Attention Decreased Attention
Accurate Inacc. No
Code Code Code
Accurate Inacc No
Code Code Code
Accurate Inacc. No
Code Code Code
Incentive
Male 1.667 1.000 1.000 1.333 6.667 1.000 2.000 2.000 2.333
Female 2.000 3.500 .500 1.500 3-500 .000 .500 12.000 1.000
No Incentive
Male 1.000 3.000 1.667 2.333 1.000 1.000 1.333 6.333 1.333
Female 1.500 6.000 2.500 2.500 7.500 .000 2.000 3.000 1.000
ON
-1
68
behaviors than did the inaccurate coders. All of these
differences occurred with p <.001. Again, there was no
difference between the levels of the retention factor on
the measure of partially reproduced elements.
In general, then, the data indicate that the ac
curate coders did not differ from the non-coders, but that
both of these groups learned more of the model's behavior
than did the inaccurate coders. This superior level of
learning manifested itself both in the number of correct
imitative responses made by the Ss, and in the number of
imitative errors made by the Ss.
Incentive Effects
When Ss were told before watching the modeling film
that they would be asked about how much of it they could
remember they did significantly better on the imitative
responses than did Ss who were not so forewarned. This
improvement of observational learning manifested itself
on three important dependent measures. Subjects given an
incentive set produced significantly more whole reproduc
tions of modeled elements. (F^ 54 ■ 8.222, p <.006)
(Table 2), more wholes plus partials (F1 ^ * 6.084,
p <.017) (Table 4), and more completely reproduced se
quences of modeling cues (F1 ^ * 10.322, p <.002)
(Table 5). Group means for the incentive factor
on these dependent measures are shown in Tables 13 through
69
Table 13
Mean Number of Whole Elements Imitated as
a Function of Incentive
Incentive Set Number of
Whole Elements
As
Percent
Incentive 20.956 52
No Incentive 18.178
45
Table 14
Mean Number of
Imitated
Whole Plus Partial Elements
as a Function of Incentive
Incentive Set Number of
Elements Imitated
As
Percent
Incentive 24.489 61
No Incentive 22.111
55
Table 15
Mean Number of Completely Reproduced
Sequences as a Function of Incentive
Incentive Set Number of As
Complete Sequences Percent
Incentive .533 7
No Incentive .133 2
70
15, and are presented both as raw numbers and as percent
ages. Results of the analysis of variance conducted on
this data are shown In Tables 2, 4, and 5 respectively.
Other dependent measures designed to test for the
degree of observational learning did not show any effect
due to the Incentive manipulation. These variables in
cluded the number of partially reproduced elements (a
measure which was not significantly affected by any
manipulation), and the three measures of imitative errors,
i.e. novel, borrowed, and extraneous behaviors.
In general, the incentive manipulation was an impor
tant factor in observational learning as viewed in terms
of the number of imitative responses correctly produced.
However, this factor did not significantly affect the
number of errors produced by Ss while attempting to imi
tate the modeling cues.
Sex Differences
For the most part males and females did not differ
on the amount of observational learning. The variables
which tapped the number of correct imitative responses
(number of whole reproductions, number of partials, wholes
plus partials, and complete sequences) did not show any
effect due to sex differences. Similarly, the number of
borrowed behavioral elements imitated in a sequence, and
the number of extraneous behaviors in total did not differ
71
between sexes. However, on the measure relating the number
of novel behaviors produced during Imitation, the female
Ss produced a significantly larger number of such behav
iors (Px 51, ■ 7.831, p <.007) (Table 6).
Interaction Effects
In all, only seven significant Interactions were
revealed In the analyses of the seven dependent measures
of primary Interest. These Interaction effects were con
fined to the dependent variables which relate the number
of errors made during the attempted imitation, i.e. the
number of novel behaviors, the number of borrowed behav
iors, and the number of extraneous behaviors (novel plus
borrowed). On the variable of novel behaviors, retention
interacted with attention (F4 54 * 5.183, p <.001) and
with sex (^2,54 “ ***038, p <.023), and there was a 3-way
interaction of retention, incentive, and sex (F2,5^ “
4.787, p <.012) as seen in Table 6. The measure of bor
rowed behaviors displayed a significant interaction between
the factors of attention, incentive, and sex (F2 ■
4.253, p <.019), and also showed a significant 4-way
interaction involving all the experimental factors (atten
tion, retention, incentive, and sex) as revealed in
Table 7 (F^ ^ * 7*965, p <.001). Finally, the dependent
variable "extraneous behaviors" revealed interaction
effects of retention and sex, attention and incentive
72
and sex, and the 4-way interaction of all factors (see
Table 8).
Various Effects on Other
Dependent Measures
In addition to the seven dependent variables which
were of particular interest because of their relevance
to the question of observational learning, numerous other
dependent measures were investigated. The Orientation
Inventory, and the two questionnaires which were adminis
tered furnished this data. Significant effects were,
again, infrequent. The retention factor produced a sig
nificant difference in Ss' ratings of their nervousness
while viewing the modeling film (F^ ^ * 5.296, p <.025)
between the group of accurate coders and the group of non
coders, the non-coders being less nervous. Further, the
retention factor resulted in different ratings of the
amount of confusion experienced during the imitation
period (F^ ^ ■ 7.773, p <.007) between these same two
groups. On this scale the non-coders rated themselves
as being less confused.
Sex differences were found on two questionnaire
items. Females rated themselves as more interested during
the imitation period (F^ ^ * 9.135, p <.004), and as more
confused at the same time (F1 ^ » 14.912, p <.001).
73
The incentive factor resulted in differential ap
praisal by the Ss of the degree to which the film made
sense. Subjects who did not receive the incentive set
instructions rated the film as being more sensible than
did the Ss who were given an incentive set ^ ■ 9.348,
p <.003).
Attention level and sex interacted on the question
naire items of how much sense the film made, and how un
usual it was (^2,514 * 3.550, p <.036; F2 ^ - 6.030,
p <.004, respectively). Attention level, incentive set,
and sex interacted on items relating to how much sense
the film made, and how much sense individual sequences
made (p2,54 * 3.097, p <.053; p2,54 * 5.321, p <.008,
respectively).
Data derived from the Orientation Inventory, which
included the time spent on the inventory and the three
scales of the inventory, did not indicate main or inter
action effects of any kind. Neither did the ORI scores
show any significant correlations with imitative perform
ance nor with questionnaire responses.
CHAPTER IX
DISCUSSION
The results of this experiment shed some additional
light of the process of observational learning, and par
ticularly on the sub-processes of attention, retention,
and incentive. The contiguity-mediational theory of
observational learning, which Bandura proposed (1969),
has considered these sub-processes to be essential ele
ments in the overall learning process. The research re
ported, herein, has explored the effects of attention,
retention, and incentive on several measures of observa
tional learning, and has revealed some very interesting
patterns of effects.
However, before looking at these patterns more
closely it is necessary to make mention of a major concep
tual difficulty regarding the various experimental factors
used herein. The introductory sections of this experi
ment specified some of the possibilities whereby attention
and retention processes might be confounded. A primary
goal of this research was to use a form of retention manip
ulation which minimized the likelihood that the actual
manipulated factor was attention. Standing in the way
74
75
of this goal was the fact that the terms "attention" and
"retention" are not so distinct conceptually that one may
clearly operationalize them in a distinct manner. Simi
larly, the operationalization of the concepts of "atten
tion" and "incentive" is easily challenged with regard to
how well the label actually fits the operation. This
experiment is the first in the field of observational
learning to attempt to directly manipulate attention, and
certainly other manipulative operationalizations are
necessary and desirable. Similarly, observational learn
ing paradigms have used only a few different techniques
to manipulate retention processes, and only one operation
alization of incentive. In order to clarify the confusion
surrounding these variables it is necessary to both util
ize a variety of operationalizations, and to pursue a
clearer theoretical conceptualization of these processes.
Attention
The subprocess of attention was of interest in this
research since no previous modeling experiment has at
tempted to manipulate this factor directly, although
several experiments have been able to use attention as a
reasonable post-hoc explanation of their findings (Grusec
and Mischel, 1966; Mischel and Grusec, 1966; Ross, 1966).
Additionally, two previous experiments which were designed
to investigate the retention factor seemed open to
76
possible confounding by variation in attention levels
(Bandura, Grusec, and Menlove, 1966; Gerst, 1969). How
ever, in the present research the attempted manipulation
of attention produced no significant differences in the
amount of observational learning which took place. This
result was consistently found through all of the primary
dependent variables. The hypotheses that Ss with in
creased attentional levels would imitate better, and make
fewer errors, were not supported. Such findings suggest
two possibilities: (1) that attentional variations do
not influence the level of observational learning; or (2)
the manner in which this experiment attempted to manipulate
attention did not actually affect attention levels.
In line with the first of these possibilities it
might be suggested that attention is not a variable which
functions over a wide range of different levels. Rather,
it appears to be more of an all-or-none phenomenon.
Either the individual is attending to the behavior closely
enough to learn it, or he is not attending closely enough.
Following this interpretation of the results of this ex
periment, it is necessary to conclude that all Ss were
attending adequately, since in fact all Ss were able to
imitate to some degree. Any variations in imitative
performance are attributable to other experimental
factors.
An alternative explanation, which is based on the
77
second possible interpretation of the data, must be con
sidered, however, in view of the aforementioned research
pointing to the relevance of attention as an important
factor. Also, the logic of the idea that paying closer
attention will lead to more learning is so trivial, yet
persuasive, that it is difficult to accept negative find
ings. The previously suggested alternative explanation
of these findings is that the manipulation of attention
was not powerful enough to, in fact, change attentional
levels. Such an explanation is supported by post-
experimental questionnaire data which revealed no inter
group differences in subjective ratings of attentiveness,
or in subjective awareness of situational factors enhanc
ing or depleting attention.
Moreover, this does not seem to be an unreasonable
conclusion, since the attempted manipulation may have been
negated through at least two means. First, the manipula
tion was carried out entirely on a cognitive level, and
was probably less forceful than some more overt distrac
tion such as a competing stimulus or task. In fact, the
cognitive nature of the retention and incentive manipula
tions may have diluted the power of the attention manipu
lation. Secondly, the situation in which the experiment
was conducted did not allow the S to do much else but
pay attention to the modeling cues. (Ss were seated only
k to 5 feet from the television screen, in an otherwise
78
quite barren room, and in two conditions of retention were
required to attend closely to the video tape so they might
describe and code it.)
The manipulation of attention in this experiment
was quite subtle, as may be appropriate when first begin
ning to systematically investigate the effects of a factor.
Unfortunately, no prior research has attempted to directly
manipulate attention in a modeling context. The negative
findings, however, should not be left unquestioned.
Clearly, more research, using manipulations which have
greater impact, is called for. Such manipulations should
seek to vary the amount of attention which an S is able
to pay to the modeling cues, without simultaneously pro
viding additional material to be remembered, and thus
confounding attention by retention.
The present results do, nevertheless, clearly show
that attempting to manipulate attention by the induction
of cognitive sets does not produce differences in the
amount of observational learning which takes place, at
least in simple observational settings. The central
question is whether or not these negative findings indi
cate a lack of importance for the attention factor, or a
lack of impact on the part of the attention manipulation.
79
Retention
Previous research has indicated the importance of
the retention factor, in the form of coding, as it re
lates to observational learning, and this study shows
clear effects of the accuracy of the code used. The
pattern of differences predicted in the hypotheses was
not produced exactly, but there is clear evidence that
coding differences produce learning differences, when
attention is held constant.
Those Ss who used an accurate code or no code at
all both learned and reproduced more of the modeling
cues, and made fewer imitative errors than did the inac
curate coders. These results were as predicted, but the
general lack of distinction between the accurate and non
coders (with the exception of the variable relating to
the number of complete sequences) was contrary to predic
tion. Perhaps this is evidence for a ceiling effect,
resulting from the fact that the modeled material was
limited in both complexity and the length of presentation,
and therefore differential imitative performance could not
result from disrupting factors. However, other possi
bilities exist. For example, this finding can perhaps be
partially explained by noting that the "no code" group
may in fact be a misnomer. This group was given no
instructions to code, and therefore should have been
less efficient at any coding it did do, perhaps not
80
coding at all for certain segments or in the case of
certain Ss. For these reasons, this group was expected
to learn less than the accurate coding group. Neverthe
less, contiguity-mediational theory, as stated by Bandura
(1969) would predict some imaginal coding during observa
tion.
So, the present study predicted that these Ss who
were not instructed to use any code at all would learn
less than the accurate coders, and more than the inaccu
rate coders. The finding that non-coders and accurate
coders did not differ in the amount learned was contrary
to predictions, but consonant with some previous research
(Gerst, 1969). In contrast to the study reported here,
that previous research used active imagery instructions,
i.e. Ss were told to imagine performing the modeled
behavior while they were watching it. The predictions
for the present experiment were based on the assumption
that any imaginal coding which took place would be passive
and rather inefficient.
Contrary to these experimental predictions, the
results obtained in this study indicate that it is not
necessary for an individual to actively employ an imaginal
code. Rather, simple observation of the modeling cues
can produce learning that equals or exceeds (on the mea
sure of complete sequences) that which is produced by
verbal description. In view of the fact that the non-
81
coders presumably took no active role as coders, and that
the imitation period was delayed an average of a little
more than ten minutes, these results indicate that imagi
nal codes may be very powerful in aiding retention.
The power of imaginal codes as retention aids may
be a result of several different factors. First of all,
visualization of an act takes less time than verbaliza
tion of the same act. So, any attempt to learn an act is
less likely to be hindered by visualization of an imme
diately preceding action than by verbalization of the same
preceding action. This would most likely be especially
true for behaviors that were difficult to verbalize, since
verbalization of these would take longer. Subjects in
the coding groups rated the behaviors in this experiment
as moderately difficult to verbalize. (Using a 13-point
scale, on which 7 was "moderately difficult," the Ss gave
a mean rating of 6.5 to the rating task.) Perhaps, then,
an imaginal code worked well in the present study because
the modeling cues were moderately difficult to describe
verbally.
In addition to the difficulty of verbalizing what
one sees, the fact that it takes longer to verbalize than
it does to visualize may make it a less effective guide
during covert rehearsal. In the present study the Orienta
tion Inventory most likely prevented a great deal of covert
82
rehearsal. But in that time between being asked to imi
tate the model, and actually beginning to imitate, it
would have been much easier to visualize the modeling
cues than to verbalize them. Such an advantage may have
been an aid to the non-coders in this study, if they did,
indeed, use an imaginal code.
Finally, imaginal codes may provide a more complete
and integrated guide during imitation than do verbal de
scriptive codes. Perhaps this is a "one picture is worth
a thousand words" effect, and may account for the finding
that Ss who were not instructed to use any particular
code produced the greatest number of complete sequences.
This finding was directly contrary to prediction, and
speaks strongly for imaginal codes as facilitating inte
gration of elements in a sequence. The process of having
to verbalize each element of behavior may, in fact, be a
hindrance to such integration, a hindrance with which the
non-coders did not have to contend.
While the predicted differences between the accu
rate and non-coding groups did not materialize, the hy
pothesized inferior learning of the inaccurate coders
manifested itself quite clearly. Of particular interest
was the finding that inaccurate coders not only remembered
the modeling cues less well, but also made more errors
during imitation. Unfortunately there is no systematic
data to indicate that Ss in this group made errors which
83
were in line with their inaccurate codes, but some re
ported that during imitation they were unable to distin
guish what they saw from what they said.
In general, the poor performance of the Ss using
an inaccurate code is in agreement with previous experi
mental findings which show the importance of the coding
function to the correct reproduction of the modeling cues.
However, this study, and all previous studies, have failed
to create a condition which effectively prevented observa
tional learning of any degree.
It has been stated by other researchers in the area
of observational learning that the true test of contiguity-
mediational theory is whether or not any observational
learning will occur when imaginal coding, as well as
verbal coding, is effectively blocked. This statement
is quite true, but it does not adequately point out some
of the difficult logical and theoretical issues which
stand in the way of such a test. As the theory is stated
by Bandura (1969), it implies that any observing response
which elicits a perceptual response will automatically
lead to an imaginal code. Such a code will aid learning
and recall, as apparently happened among the non-coders
in the present experiment. Conversely, however, to block
all such passively acquired imaginal codes, It would be
necessary to prevent perceptual responses. If this is
accomplished, the question immediately arises as to
84
whether or not an observing response has actually occurred.
That is, does sensation alone constitute observation, or
does observing imply a perceptual response. Obviously,
observation is an essential element of observational learn
ing, and the question of what constitutes an observational
response becomes paramount, although not necessarily an
swerable. It would appear, then, that in order to have
a crucial test of contiguity-mediational theory, it is
necessary to devise a means of blocking perception without
hindering observation. Until this is done it will not be
possible to ascertain whether some form of coding is essen
tial to observational learning, or whether simply observ
ing the modeling cues is adequate.
More practically, one could approach this problem
by attempting to minimize coding through the use of input
on another channel. Specifically, one could first try
using auditory input of an inherently interesting nature,
e.g. some sexually arousing passage from a book. By pre
senting this input simultaneously with modeling cues for
some novel and unrelated behavior one might block visual
perception. Secondly, again using the auditory channel,
Ss could be presented with material which they are re
quired to analyze, interpret, or code, while at the same
time watching the modeling cues. In both of these cases,
the goal is to provide stimuli which eliminate any active
effort at coding the visual input, without also blocking
85
the observation of those cues. Unfortunately, these
methods once again pose the problem of possibly confound
ing retention by attention, since there is no simple way
to guarantee that the visual stimuli are being attended
to.
The lingering question is whether or not coding in
any form is a necessary condition for observational learn
ing. As contiguity-mediational theory is presently
stated it is difficult to determine the distinction be
tween attention processes on the one hand, and retention
processes, as represented by coding, on the other. (For
instance, how can one distinguish between simply paying
attention and passively applying an imaginal code.) If
the inhibition of coding by use of an inaccurate code,
as in this study, or by use of stimuli presented on the
auditory channel, as suggested above, does not prevent
observational learning then one must consider the possi
bility that symbolic retention of the modeling cues can
occur without coding.
Incentive
Hypotheses concerning the incentive manipulation
were partially confirmed, with results which indicated
an interesting pattern of effects. Those dependent vari
ables which were indicative of how much of the modeling
cues the S could reproduce correctly showed a clear
86
advantage for Ss receiving the incentive instructions.
On the other hand, variables relating to errors made
during imitation showed no differences as a function of
incentive.
Perhaps imitative errors are a function of con
fusion experienced during coding and recall of the model
ing cues. Whereas incentive provides the Ss with a reason
for attempting to learn and remember what they observe,
thereby influencing the number of correct imitations, it
does not have any effect on the amount of confusion Ss
experience, and therefore does not influence the number
of initative errors. This interpretation is consistent
with the finding that an inaccurate code did influence
the number of imitative errors, and the notion that this
finding resulted from confusing what was observed with
what was inaccurately described. Nevertheless, the valid
ity of this interpretation is open to question on the
basis of the finding that the retention manipulation did
not singly, or in combination with any other factor, sig
nificantly affect Ss’ ratings of their level of confusion
during imitation, as confusion relates to imitative errors.
Indirect support for the notion that confusion is
the important factor with regard to the number of errors
made comes from the findings relating to sex differences.
No sex differences were predicted in the experiment, but
the results reveal that female Ss produced more novel
87
behaviors than did males. Moreover, females reported
themselves as being more confused, as compared to male
Ss* self-ratings. It would appear, then, that increased
confusion may account for an increased number of imita
tive errors.
Interaction effects, like sex differences, were not
predicted by the experimental hypotheses, and by and
large, they did not appear. However, the retention and
sex factors, and the retention and attention factors did
interact on the variable of novel behaviors, and these
effects may bear some relevance to the topic of confusion
during imitation.
For male Ss the differing coding activities pro
duced a constant gradual increase in the number of novel
behaviors, moving from accurate coders as the lowest,
through non-coders, to inaccurate coders as the highest
producers of novel behaviors. Female Ss, on the other
hand, were least likely to perform novel behaviors when
they were non-coders, slightly more likely as accurate
coders, and far more likely as inaccurate coders. In
fact, the retention and sex factors, which each singly
affected this variable, combined to produce more than
twice as many novel behaviors among inaccurately coding
females than were produced by any other group. This is
consistent with the idea that confusion, whether caused
by an inaccurate code or by some type of sex differential
88
response to the experimental situation, leads to an in
crease in the number of errors made, especially in the
form of novel behaviors. Similarly, when an inaccurate
code was combined with instructions designed to decrease
the level of attention, the number of novel behaviors
produced was twice that produced by any other group
formed by these two factors.
By and large the other interaction effects which
were found are difficult to interpret, and offer no read
ily understandable information regarding the observa
tional learning process. This is particularly true of
the 4-way interaction which resulted, but which the
present experimental design does not make interpretable.
It is interesting to note that all of the interactions
occurred with relevance to dependent measures involving
imitative errors, but this observation can only be noted
and not explained on the basis of the present data.
The results of this study also have broad implica
tions for various theories of observational learning. For
example, those theories which place emphasis on rein
forcement do not seem to explain the findings of this
experiment, in which no identifiable reinforcement ever
occurred. Similarly, they are ill-equipped to explain
the finding that a delay of ten minutes on the average
between observation and imitation did not prevent learn
ing. Apparently, learning was accomplished without an
89
imitative response (R) or a reinforcer (Sr), which are
two of the essential elements in operant analyses of
modeling. As was stated in the introductory sections of
this paper, reinforcement seems more important as a fac
tor in imitative performance than in observational learn
ing.
Contiguity-mediational theories of observational
learning have not been really explicitly spelled out,
but they have basically laid particular emphasis on
symbolic processes, which serve as guidelines for imita
tive performance. The present results have supported the
emphasis on symbolic processes, and have demonstrated
that interference with such processes inhibits observa
tional learning and disturbs accurate imitation of the
observed behavior. However, the necessary and sufficient
conditions for such learning remain elusive.
It would appear that a certain level of attention
is required, but that variations above that level do not
cause concomitant variations in the amount of learning
which takes place. Further, although retention of model
ing cues in some symbolic form seems to be the most
feasible and parsimonious way to explain delayed repro
duction of those behaviors, the results of this study
and previous studies point out that such symbolic medi
ators may be acquired unintentionally, passively, and
even in spite of competing symbolic activity. This
90
robustness of retention, even when coding is not actively
pursued, makes it questionable whether or not coding, in
the sense of combining elements into groups and remember
ing the group by some code, is a necessity for observa
tional learning. Other research discussed previously has
shown that coding can enhance such learning, but the evi
dence to date suggests that the learning and retention
processes can take place without coding. Similarly,
the results of this study show that an incentive to learn
can facilitate and promote the learning and recall pro
cesses, but that such an incentive is not absolutely
necessary for observational learning to take place.
In general, the role of symbolic processes and
symbolic mediators in observational learning appears to
be well corroborated and established. The establishing
of these mediators and the functioning of these processes
require an adequate level of attention, and a mechanism
for retention. The necessity of coding, however, is not
established, and the function of incentive is unclear.
Incentive to engage in a process of learning per se does
not appear to be required, but the interplay of attention
and incentive, which seems to be so logical, makes it
difficult to determine the exact role of incentive.
New theorizing and research should be directed
toward the issue of whether or not some degree of obser
vational learning necessarily follows from any visual
perceptual response. The question of whether or not, and
to what degree, varied attentlonal levels affect observa
tional learning also deserves more Investigation. Those
factors which control the specificity of such learning,
i.e. why one observed behavior is learned while another
is not, are yet another worthy topic. In fact, all these
questions, and doubtless many others, must be answered
in order to genuinely understand the process of observa
tional learning.
APPENDICES
APPENDIX I
Listed below are the props used by the model, and a
description of the modeled behaviors associated with each
prop. The modeling stimuli consisted of eight sequences,
each using a distinct prop and its associated behaviors.
The props are listed in the order in which they appeared
in the modeling tape. In addition, portions of the modeled
behavior described here are underlined. These portions
comprise those behaviors which are considered essential for
any imitative effort to be scored as a "whole" reproduc
tion. Any imitative effort which contained part of the
underlined portion of an element, but not all of that
portion, was scored as a "partial" reproduction.
1. A pair of work gloves:
a. put the gloves on and hold hands up in front of
you with open palms towards the camera.
b. bend over at the waist and cross your hands on
your knees or lower thighs.
c. stand up straight and fold your arms across your
chest with your hands tucked between your arms
and the sides of your body.
d. remove your hands from the gloves and put your
hands at your side while holding the gloves be
tween your arms and your sides.
e. still holding the gloves under your arms, kneel
down.
2. An apron:
a. put the apron around your neck as if it were a bib.
b. pull the apron off with one hand and wave it
through the air with that hand.
c. drape the apron over one hand, and hold the hand
up at your side, and turn a circle around the
covered hand.
d. put the apron oyer the other hand, and turn a
circle around that hand.
93
94
e. wad the apron up in your hands, and throw the apron
towards the camera.
3. Two twirling batons:
a. cross the batons in an "X" in front of your face
or upper body.
b. uncross the batons and put them behind your back.
c. remove the batons from behind your back and lean
the batons on your shoulders, one on each shoulder.
d. lay the batons on the floor directly in front of
your feet (the manner in which you lay the batons
down, i.e. crossed or uncrossed,makes no differ
ence) .
e. step over the batons.
4. A football helmet:
a. put the helmet on your head.
b. take the helmet off and hold it under your arm and
against your side.
c. place the helmet between your knees and hold it
there.
d. put the helmet on the floor beside you.
e. kick the helmet.
5. A vest:
a. put the vest on correctly.
b. with the vest still on, sit down on the floor and
cross your arms and your legs^
c. still seated, remove the vest.
d. still seated, put the vest over your head and then
stand up.
e. remove the vest from over your head, and put it on
backwards.
95
6. A mask which covered upper half of the face:
a. put the mask on correctly.
b. with the mask still on, use one arm to point to the
side, and turn your head as If to look In that
direction.
c. put your arm down and remove the mask, and then
put the mask on Incorrectly (e.g., around your
neck, on top of your head,backwards).
d. with the mask still on Incorrectly use your arm
(the opposite arm from the one used before) to
point to the side (also the opposite side from
before) and look In that direction.
e. put your arm down, face the camera, remove the mask.
and hold It In front of you between your hands as
If praying.
7. A skillet, and a plastic ball 8" in diameter.
a. set the ball on the floor, and place the skillet
upside down upon the ball.
b. remove the skillet from the ball and pass the
skillet behind your back from one hand to the
other.
c. sit the skillet on the floor, and place the ball
In It.
d. put one hand on top of the ball (which is still in
the skillet), and the other hand on top of your
head.
e. remove your hands from previous positions, and
use one hand to hit the handle of the skillet,
making the ball bounced
8. A baseball bat and a Raggedy-Ann doll:
a. with bat in one hand and the doll in the other,
turn to the side, bend forward at the waist, and
lean on the bat.
96
b. stand up straight (facing any direction is O.K.)
and kiss the doll.
c. spread your arms, holding the doll in one hand and
tne bat in the other.
d. lower arms, place doll in your mouth and shake the
doll back and forth.
e. remove doll from your mouth, and hit the doll with
the bat.
APPENDIX II
Alleged Rationale for the Experiment
Please sit here, facing the TV screen. The purpose
of this experiment is to determine how people react to what
they see. We are especially interested in two things:
First, we want to know how what you see affects your reac
tion. As you watch the TV you will see a videotape of a
girl doing different things with some different objects.
Some of the things she does will be what you would expect
her to do with that object, and some of them will be kind
of unusual. Now, the second thing we are interested in is
how what you are doing as you watch the tape affects your
reaction to it. So, we ask different people to do differ
ent things as they watch the tape. Then, later on, I'll
give you some questionnaires to fill out, so that we can
find out your reactions to the tape, and see what kind
of attitudes and opinions you have about what you saw.
97
APPENDIX III
Instructions to the Ss regarding the manner in
which they are to cocTe the modeling stimuli.
Accurate coders: Now, as I said, I'm also interested
in how what you do as you watch the tape affects your reac
tions to it. What I'd like you to do is watch the tape,
and describe aloud what you see the girl do, as accurately
as possible. For instance, if she walks into a room, picks
up a football, throws it into the air, catches it, and then
she sticks it in her mouth, you would say something like
this: "Well, the girl walks into the room, she picks up
a football, throws it in the air and catches it as it comes
down. Then she sticks it in her mouth." So, all you do
is accurately describe what the girl does, as you are watch
ing her do it.
Inaccurate coders: Now, as I said, I'm also inter
ested in how what you do as you watch the tape affects your
reactions to it. What I'd like you to do is watch the tape,
and describe what you see the girl do, only I'd like you to
change some of the details so that your description is
slightly inaccurate. For instance, if she walks into the
room, picks up a football, throws it into the air, catches
it, and then sticks it in her mouth, you would say something
98
99
like this: "Well, the girl walks into the room, picks up
a football, throws it down on the ground, picks it up
again, and then sticks it in her ear." So, what you do
as you watch the film, is describe it inaccurately, by
changing some of the details.
Non-coders: Now, as I said, I'm also interested
in how what you do as you watch the tape affects your
reactions to it. I ask different people to do different
things. All I want you to do is simply sit and watch
the tape.
APPENDIX IV
Those Ss who made up the incentive group heard
the following statement immediately after
receiving their coding instructions. The
"no-incentive" simply heard no additional
instruction.
In addition, after you have seen the tape I’ll be
interested in how much of it you can remember, so I'll be
asking you about that later, too.
100
APPENDIX V
Statements to the Ss designed to manipulate
their attentional level.
Attention increasing statement: We have found that
college students, because they are younger and more intel
ligent than the general population, are able to concen
trate very well and focus their attention very specifi
cally on what they are watching. In fact, you can really
say that because they are intelligent they are able to
"drink in" every detail of what they see because they can
pay such close attention to it.
Attention decreasing statement: We have found that
college students, because they are younger and more intel
ligent than the general population, often find it unneces
sary and uninteresting to focus their attention very speci
fically on what they are watching. When the subject matter
is simple, they find that it isn't necessary for them to
concentrate on Just one thing at a time because they can
handle a lot of information all at once. In fact, you can
really say that because they are intelligent they are able
to pay close attention to many things going on around them.
Neutral attention level statement: So, I Just want
to remind you again to pay close attention to the TV.
101
APPENDIX VI
Instruction to the Ss prior to their effort
to imitate the modeling stimuli.
Now, I'm also interested in how much you can remem
ber of what you saw the girl do with these objects. The
easiest way to find out about that is to ask you to imitate
what you saw her do with each of these things. As you
imitate her, I'll record what you do on this videotape
machine, so that I can go back over it at another time,
more slowly, and score what you do. It doesn't make any
difference what order you do things in, Just start with
the object which is easiest for you to remember, and do it
in whatever way is easiest for you. I would like for you
to try to do the things she did with a particular object
in the same order that she did them, but, as I said before,
you don't need to use the objects in the same order that
she did. I won't be able to tell you whether or not you
are getting things right as you go along, so Just do your
best. You have ten minutes to do this, but you don't have
to take that long unless you want to.
102
APPENDIX VII
Post-Experimental Questionnaires Completed
by Each S
Name
1. How Interesting would you say the film was?
■ * ■ ___ i. < ____i ___ « » « ___ ■ ____■ _i____i. t
Not at all Average Very
interesting interesting
2. How interested would you say the girl in the film was
in what she was doing?
l____i ____i ____i ___ ■ i j » ____i ___ i ___ i ___i ____i ___ i .
Not at all Somewhat Very
interested interested interested
3. Would you say that the film made any sense?
i _____■ ‘__i i , > ___ i ____i ___i -i__». « i--
No sense Made some Very
at all sense sensible
4. Would you say that individual segments of the film
made sense?
L ------------I ---------------------- 1 ------- i-------- ■ ---------------- I - - - t J ---------1 . - I 4 , -------------- 1 ------------- 1 ---------
No sense Made some Very
at all sense sensible
5. How complicated would you say that the film was?
kot at all ' ' ' Somewhat 1 1 Very*
complicated complicated complicated
6. Would you say that the film seemed unusual to you?
L , _J_
Quite Somewhat ' Very *
unusual unusual ordinary
103
From your observations of the film, how would you rate
the girl you saw on these dimensions?
1 1 »
Intelligent
j____i — - * *
Average
• * ___l____i___ j___i — i
Unintelli
gent
■ ■ i i i i ■ < i i
Attractive Average Unattractive
i i i i i i i i i J i i t i
Happy Average Unhappy
i i i i i i i i i l • i t i
Friendly Average Unfriendly
i t i t i i i i i iifi
Strong Average Weak
How would you
the film?
describe yourself while you were watching
i i i i * i i i ■ i i i i ■
Interested Average Disinteres
ted
i i i i i i i i i i 1 i i i
Calm Average Nervous
i l i t I I I < l I i 1 i I
Attentive Average Unattentive
L 1 * 1 i i i I i l 1 I 1 . 1
Intrigued Average Bored
How would you
ting the film?
describe yourself while you were imita-
i i i i i i i ( i i i i | I
Interested Average Disinter
ested
i i i i i i t i ! i i i t 1
Calm Average Nervous
i i i i i i l i 1 i i i i i
Relaxed Average Uneasy
i i > i
* __i ____l ____i 1 1 , .
i i i i
Serious Average Frivolous
105
i i ----1 ----1 ----1 ----1 ----1 ----1 ----1 ____i ____i ____i ____i ____i
Confused Average Not at all
confused
106
Name______________________
Cond.__________________________
1. What were your expectations about the experiment before
you came here?
2. After hearing your instructions, did you think that
you'd be asked to imitate the film? Yes______ No_____
Did you ever begin to expect that you would be asked to
imitate the film before the experimenter asked you to
do so? Yes No . If so, when?
3. Do you feel that anything that happened distracted you
from the film? Yes No If so, what?
*4. Do you feel that anything that happened helped you to
pay attention to the film? Yes No If so,
what?
5. Did you attempt to learn and remember what you were
watching? Yes No If you did, how did you try
to help yourself do this?
6. Did you try to do anything to make remembering easier?
Yes No_____ If so, what?
7. Do you feel that anything that happened during the
experiment helped you to remember the film? Yes__
No If so, what?
8. How did you try to recall what you saw on the film:
Did you try to describe to yourself? Yes No_____
Did you try to visualize it? Yes______ No
Did you use some other method? Yes____ No if so, what?
107
This page was given only to the "non-coding"
Ss.
9. During the imitation period, when you were asked to
imitate the film, did you suspect that there was a
hidden reason behind asking you to imitate? Yes____
No______ If so, what?
10. Did you imitate every behavior that you could remember?
Yes No If no, what not?
108
This page was given only to the accurately
coding Ss.
Name
9. While you were watching the film and trying to describe
what you were seeing, how hard would you say that It
was to describe the different sequences?
i____1 , I ____I ____i i i ____i ____I ____I ____L____I ----1---
Not difficultModeratelyVery
at all difficult difficult
10. During the imitation period, when you were asked to
imitate the film, did you suspect that there was a
hidden reason behind asking you to imitate? Yes___
No If so, what?
11. Did you imitate every behavior that you could remember?
Yes No If no, why not?
109
This page was given to only the inaccurately
coding Ss.
Name
9. While you were watching the film and trying to describe
something that was similar to what you were watching
but not exactly like it, how hard would you say that
it was to think of these things and describe them?
i ---1 — __i----1 ----1 ____i ____i ____i ____i ____i ____i i i
Not difficultModeratelyVery
at all difficult difficult
10. During the imitation period, when you were asked to
imitate the film, did you suspect that there was a
hidden reason behind asking you to imitate? Yes_____
No______ If so, what?
11. Did you imitate every behavior that you could remember?
Yes No If no, why not?
APPENDIX VIII
Scoring Sheet Used by Raters to Score the Imitative
Performance of Each S.
S0
Name
Cond. 0 Cond. des.
Score Extraneous Beh.
Pro£
1. gloves
Description
gloves on
show palms
crossed
on legs
crossed
on chest
gloves
under
arms
Order
kneel
down
apron put on
as bib
wave as
flag
over hand
turn circle
other hand
turn circle
throw
at
camera
batons cross
batons
behind rest on lay batons step over
back shoulders on floor batons
Extraneous Behav.
Novel Borrowed
4. helmet put hold hold helmet by kick
helmet under between side on helmet
on arm knees floor
o
Cond. §
APPENDIX VIII (cont.)
Score for this page_______ Extran. beh.
Prop
5. vest
6. mask
7. skillet
and ball
8. bat and
doll
Description Order Extraneous Behav.
Novel Borrowed
put vest
on
sit down—
arms and
legs crossed
remove vest over vest
vest head— on
stand up back
wards
mask on
skillet
on ball
turn head
right—
point right
skillet
behind
back
mask on
wrong
ball in
skillet
turn head hold
left— mask
point left hands
praying
hand on hit
ball— hand skillet
on head handle
turn to kiss
side— lean doll
on bat
spread arms— doll in hit doll
doll in one mouth and with bat
hand, bat in shake
other
111
APPENDIX IX
The following is a list of all Ss raw scores on
each of the seven primary dependent variables, and these
scores are shown for each subject in the appropriately
labeled columns. The Ss are presented in the order in
which they participated in the experiment, and the treat
ment condition to which they belong is described by the
four numbers which follow the S number.
The first of these digits refers to the attention
variable according to the following code: a "1" means
that the S was operating under attention increasing in
structions; a "2" means that the S heard attention de
creasing instructions; and, a "3" means that the S was
in the attentional control group.
The second digit refers to the retention factor,
according to the following code: a "1" means that the
S used an accurate code; a ”2" means the S used an inac
curate code; and, a "3” refers to the non-coding Ss.
The third digit uses either a ”1," which identifies
an S who was given an incentive set, or a "2," for the Ss
with no incentive.
Finally, the fourth digit uses a "1” to indicate
that the S was male, and a ”2" to indicate female Ss.
112
c o c o o o o o o o o o r o r o r o r o r o r o i o r o i o r o P J P-, P-, P-'i-, H P j P-, P-, P - ' o o o o o o o o o
CD -trCO l\) H O V O O O - J C T N C T I XrCO M H O VO 0 0 -0 CT\U1 - t U ) M H O V D OO—J C T\V_n J r W t U H
pJ roP-,rocorocop-'roojMtouoroPJu>fo
P-,P-,cococot-iroojooroP-,foooroPj roro
t o P j P-‘ Pj H P -1t o P j r o r o t o r o ! - ‘ t o P J P-‘ to
W H M H I U M W H H H P H M M I O H W
M ro M CO CO CO ro ro CO M ro CO ro CO H CO r —1to
ro CO CO CO M M ro ro M M 1 —*M to p j ro CO m i-*
m ro M ro m ro to M M M to to m ro to ro to ro
H ro ro ro ro M M m M M M ro ro M ro M to m
oorotoroooropj p-,roi-,iop-,rororoi-,i-,rorooorooofop-,p-,tototororororoi-,roro
MCTl - t i r C - ^ N - ^ V O V O -CrVOVOOOVOCO-'lcn O to CO O -ten P O O S O S - J O J O ctncoco
r o r o r o r o r o r o H H r o H r o M r o H r o M H t o M r o M r o M M O r o M M t o r o H M M M r o
O O P O O - J U l M U ) -Crvjn P-> -t=- ON OOCO ororoooooco-fc-vo-ervocovoCTN-fc-covo—o-r=-cr\ro
o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o
U) -fr J=-XTUJ rOVJI Jrvjl -CrCOCnCOCn ONCO VJI CO ro -fUl CTNVJl P W U 1 O O t U ) Xr t W tO-O P*
o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o
O O O M H O O O O O O O H O M O O O O I O O O M O O M O O O H O O O O O
o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o
P O O O P M U W O W O H O O P P M O O P H H P O I O O W t O P O W M H O
O O O O O O O O O O O O O O O O O O O O O O O O M O O O O O O O O O O
O H O H O H O O O O \ H O H O O H O V J 1H O O O O W O O O H O O ) U ) H U 1H H H
O O O O O O O O O O O O O O O O O O O O O O O O H O O O O O O O O O O
H H O H H U I U M O C O P H P O O M P S H O H P H W O M O U ) -fc-co t P S W M P
= * = Whole + partial
=* Whole reproductions
^ ^artial reproductions
=» Complete sequences
= * = Novel behaviors
= * : Borrowed behaviors
=* Extraneous behaviors
(novel + borrowed)
OVCTvOVOVOVOVOVOVVnvnVJIvnUlvnvnvnvjIVn t i r t jr -to o OO OOOO m
— J OVVJ1 -to o W H O V O OO —J C TW J1 XroO ro I—1 O VO 00-0 CTVVJ1 Xr OO IO M O V£> 00-0 CT\ %
r o r o r o r o o o t - ‘ i - , r o r v ) t - J t u o o t - ‘ iv > i-Jo o r o t - ‘ i - ‘ o o o o i - ' H r o o o o o o o r \ j r o r o o o u o
r o r o o o H H o o i - ' i - ' H r o o o o o M i - ' O J i - ' H o o o o i - ' H ' f o r u r o r o o o f o H ' O J O o o o r o
H H W W H W W W H W H W r O W H M H H W W H I U H r O W H W H W H W H
i - , H i - , t - , i - , h-, i - ‘ r v ) r v ) l- , i - , t - , t - , r v ) i - ‘ r v ) i - , i - , i - , i - , i - ‘ h-, r v ) i - ' i - , i - l i - , H , r o i v ) i - ‘ h J
o t - 3
o ^
3 0 >
(X p
H- c+
c+ 3
P- ID
O 3
3 c t
r o r o r \ j r o i v j u o r o r o r o * - , r v > r v > r o r o o o o o r o o o r o r o r u » - , r o t - , i - l o o r \ j r u r o o o r o r u
— - t o o v o x = - o r o o o o o o o - o o o H o o v j i o r o H H o i o u i o ^ o o i -1 w o v o o o w o
=* Whole + partial
rofo^-‘ ^Jrofu^-‘ rorooro^o^-, f\Jooruroro^-‘ ro^o^-■^-J^-'^-‘ ro^-‘ ro^oro^-‘ ^-,
- t ro v o v n ro - t oo o oo v o rv> - t o o H o o i - ' o o o v - o r o o o o o c o . t . t o v o n moo o o - o ov
= * = Whole reproductions
o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o
o o ro - t . t ro o\ - t o o on - t v n - t o o ro ro v o u i on - t o o oo ro rv> ro - t v j i - o t - o n ro v n - t
=* Partial reproductions
o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o
O O O O O H O O O O H O M O O O O O O O O O O H O O IO O O O roO M
=* Complete sequences
o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o = * = Novel behaviors
( - * O O O O O O f O r o M O O O H J M I - , O O O O O X r h - i O r O O O O I - , O O O f O
o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o Borrowed behaviors
H H O M O O O O H O O H O roO O O H O W O U lO W M VM H O O O H H
o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o Extraneous behaviors
wHowoooouioiuwoMOHHMoyiooo txrwswx-pooHoo (novel + borrowed)
vo o o o o c o c o c o o o o o o o o o o o - J — j — a — j ^ - o - o - o o n o n | W
o v o o d ~ j o v u i - t - w ro h o v o o o - ^ o v u i - t u j m h o v o oo =»=
H W H H H H
M ro iv) ui m ro
ro ro ro ro H h
ro oo »-* i-* oo v -1 to
w ro h i\)w w ru
i— ■ i— * • — 11 -11 —1 ro ■
W W H H H H H
I —1 I —1 OO h- 1 OO ro
ro oo ro ro r o o o
ro ro h* to ro M
ro ro H M ro ro
o H
o
ro r o o o o o 3 (D
L O U 1 W H
0, fB
ro i-1 ro i-1 H- ct
H* M ro ro ct 3
m - a
O 3
3 ct
U H H H I o r o H H W W H U J H H H H r O H U I o r o W W =»fc
o o o v j i -Cr— J o v n v n r o o o r o i - ' v o o o - o r o r o C D M C T v - o i - ' O N
l O H H H I O H H H W M H r O H H H O H H I O M W r O l O =»fc
s u i o H - t s w H o u i o o v - J - t u i o v o u i o o o o -trvji ro
O O O O O O O O O O O O O O O O O O O O O O O =*:
r o o o vji oo oo oo ro -t ro o o ro on ro ro r o o o o o o o o o o o o o crv -t
O O O O O O O O O O O O O O O O O O O O O O O "Us
t - ' O O O O O O O H O O O O O O O O O O J I - ' l - ' r o O
O O O O O O O O O O O O O O O O O O O O O O O =*:
H V O W O O O W S O O M O U I W t H O M O H O H O
O O O O O O O O O O O O O O O O O O O O O O O =*:
H t O H O H O ^ M H H O - t o o O H W H O M O O O
O M O O O O O M O O O O O O O O O O O O O O O =*:
W U ) U l H O H W H W H W O - ^ ) J I £ - | \ ) U ) W O U ) O H O
Whole + partial
Whole reproductions
Partial reproductions
Complete sequences
Novel behaviors
Borrowed behaviors
Extraneous behaviors
(novel + borrowed)
REFERENCES
116
REFERENCES
Allport, F. H. Social Psychology. Cambridge, Mass.:
Riverside Press, 1924.
Aronfreed, J. Conduct and Conscience. New York: Academic
Press, 1968.
Baer, D. M., Peterson, R. F., and Sherman, J. A. The de
velopment of imitation by reinforcing behavioral
similarity to a model. Journ.of the Exptal. Analy
sis of Behavior. 1967, 10, 405-416.
Baer, D. M., and Sherman, J. A. Reinforcement control of
generalized imitation in young children. Journ.
of Exptal. Child Psych., 1964, 1, 37-49.
Bandura, A. Vicarious processes: A case of no-trial
learning. In L. Berkowitz (ed.), Advances in Ex
ptal. Social Psychology. Vol. II. New York:
Academic Press, 1965, pp. 1-55* (a).
Bandura, A. Influence of models' reinforcement contin
gencies on the acquisition of imitative responses.
J. Pers. Soc. Psych. ,1965i 1, 589-595.
Bandura, A. Social-learning theory of identificatory pro
cesses. In D. A. Qoslin (ed.), Handbook of Social
ization Theory and Research. Chicago: Rand McNally,
1968.
Bandura, A. Principles of Behavior Modification. New
York: Holt, Rinehart, and Winston, 19^9•
Bandura, A., Grusec, J., and Menlove, F. Observational
learning as a function of symbolization and incentive
set. Child Development, 1966, 21, 499-506.
Bandura, A., Grusec, J., and Menlove, F. Vicarious extinc
tion of avoidance behavior. J. Pers. Soc. Psych.,
1967, 1, 16-23. (a)
Bandura, A., Grusec, J., and Menlove, F. Some social de
terminants of self-monitoring reinforcement systems.
J. Pers. Soc. Psych., 1967, 5» 449-455. (b)
117
118
Bandura, A., and Harris, M. B. Modification of syntactic
style. J. of Exptal. Child Psych.. 1966, 4, 341-
352.
Bandura, A., and Huston, A. C. Identification as a process
of incidental learning. J. Abnor. Soc. Psych..
1961, 6^, 311-318.
Bandura, A., and Kupers, C. Transmission of patterns of
self-reinforcement through modeling. J. Abnor.
Soc. Psych.. 1964, 6£, 1-9.
Bandura, A., and Mischel, W. The influence of models in
modifying delay of gratification patterns.
J. Pers. Soc. Psych.. 1965, 2, 698-705.
Bandura, A., and Rosenthal, T. Vicarious classical con
ditioning as a function of arousal level. J. Pers.
Soc. Psych.. 1966, 2* 54-62.
Bandura, A., Ross, D., and Ross, S. Imitation of film-
mediated aggressive models. J. Abnor. Soc. Psych..
1963, 66, 3-11. —
Baron, R. A., and Kepner, C. R. Model's behavior and
attraction toward the model as determinants of
adult aggressive behavior. J. Pers. Soc. Psych..
1970, 14, 335-344.
Baron, R. A. Attraction toward the model and model's
competence as determinants of adult imitative
behavior. J. Pers. Soc. Psych.. 1970, 14, 345-351.
Bass, B. M. The Orientation Inventory. Palo Alto: Con
sulting Psychologists Press, 1962.
Bringham, T. A., and Sherman, J. A. Experimental analysis
of verbal imitation in preschool children.
J. Applied Behavior Analysis. 1968, 1, 151-158.
deCharms, R., and Rosenbaum, M. E. Status variables and
matching behavior. J. Pers.. I960, 28, 492-502.
Easterbrook, J. A. The effect of emotion on cue utiliza
tion and the organization of behavior. Psych. Rev.,
1959, 66, 183-201.
Oelfand, D. M. The influence of self-esteem on rate of
verbal conditioning and social matching behavior.
J. of Abnor. Soc. Psych.. 1962, 6j5, 259-262.
119
Gerst, M. S. Symbolic coding processes In observational
learning. J. Pers. Soc. Psych., 1971, 1£, 7-17.
Grusec, J., and Mischel, W. The model's characteristics
as determinants of social learning. J. Pers. Soc.
Psych.. 1966, 3, 211-215.
Hays, W. Statistics. New York: Holt, Rinehart, and
Winston, 1963*
Hicks, D. J. Imitation and retention of film-mediated
aggressive peer and adult models. J. Pers. Soc.
Psych.. 1965, 2, 97-100.
Holt, E. B. Animal Drive and the Learning Process.
Vol. 1~ New Y'ork: riolt, 1931.
Jabukzak, L. P., and Walters, R. Suggestibility as de
pendency behavior. J. Abnor. Soc. Psych.. 1959,
5£, 102-107.
Kagan, J., and Mussen, P. Dependency themes on the TAT
and group conformity. J. of Consult. Psych..
1956, 20, 29-32.
Kanareff, V. T., and Lanzetta, J. Effects of task defini
tion and probability of reinforcement upon the
acquisition and extinction of imitative responses.
J. Exptal. Psych.. I960, 60, 340-348.
Kanner, J., and Sulzer, R. Overt and covert rehearsal of
fifty per cent versus one hundred per cent material
in film learning. In A. A. Lumsdaine (ed.), Stu
dent Response in Programmed Instruction. Washing
ton, D.C.: National Academy of Sciences— National
Research Council, 1961. Pp. 111-112.
Kausler, D., and Trapp, E. Motivation and cue utilization
In intentional and incidental learning. Psych. Rev.,
I960, 67, 373-379.
Lefkowitz, M., Blake, R., and Mouton, J. Status factors
in pedestrian violation of traffic signals. J.
Abnor. Soc. Psych.. 1955, £1, 175-183.
Maccoby, E., and Wilson, W. Identification and observa
tional learning from films. J. Abnor. Soc. Psych.,
1957, ££, 76-87.
120
Mandler, Q. Organization and memory. In K. W. Spence and
J. T. Spence (eds.), The Psychology of Learning and
Motivation. New York! Academic Press, 1967*
Vol. 1, pp. 328-373.
Mandler, 0. Association and organization: Facts, fancies,
and theories. In T. R. Dixon and D. L. Horton
(eds.), Verbal Behavior and General Behavior Theory.
Englewood Cliffs, N. J.: Prentice-Hall, 1968.
Pp. 109-119.
Margolius, G., and Sheffield, F. Optimum methods of com
bining practice with filmed demonstration in teach
ing complex response sequences: Serial learning of
a mechanical assembly task. In A. A. Lumsdalne
(ed.). Student Response in Programmed Instruction.
Washington, D.C.: National Academy of Sciences—
National Research Council, 1961. Pp. 33-53.
Mausner, B. Studies in social interaction: III. Effect
of variation in one partner’s prestige on the inter
action of observer pairs. J. Applied Psych., 1953,
II, 391-393.
Mausner, B. The effect of prior reinforcement on the inter
action of observer pairs. J. Abnor. Soc. Psych.,
1954, 49, 65-68. (a)
Mausner, B. The effect of one partner’s success in a rele
vant task on the interaction of observer pairs.
J. Abnor. Soc. Psych.. 1954, 49, 557-560. (b)
Mausner, B., and Bloch, B. A study of the additivity of
variables affecting social interaction. J. Abnor.
Soc. Psych.. 1957, £4, 250-256.
McDougall, W. An Introduction to Social Psychology.
London: Methuen, 1908.
McGuire, W. Audience participation and audio-visual in
struction: Overt-covert responding and rate of
presentation. In A. A. Lumsdalne (eds.), Student
Response in Programmed Instruction. Washington, D.C.
National Academy of Sciences— National Research
Council, 1961. Pp. 431-438.
Michael, D., and Maccoby, N. Factors influencing the
effects of student participation on verbal learning
from films: Motivating versus practice effects,
121
"feedback,” and overt versus covert responding. In
A. A. Lumsdalne (ed.), Student Response in Pro
grammed Instruction. Washington, DC.: National
Academy of Sciences— National Research Council,
1961. Pp. 271-293.
Miller, G. The magical number seven, plus or minus two:
Some limits on our capacity for processing informa
tion. Psych. Rev.. 1956, 6^, 81-97.
Miller, N. E., and Dollard, J. Social Learning and Imlta-
tion. New Haven, Conn.: Yale University Press,
TpT.
Mischel, W., and Grusec, J. Determinants of the rehearsal
and transmission of neutral and aversive behaviors.
J. Pers. Soc. Psych.. 1966, %.» 197-205.
Morgan, C. Habit and Instinct. London: E. Arnold, 1896.
Mowrer, 0. H. Learning Theory and the Symbolic Processes.
New York" Wiley, I960.
Perry, H. The relative efficiency of actual and imaginary
practice in five selected tasks. Archives of Psy
chology , 1939, No* 2**3.
Rosenbaum, M., and Tucker, I. The competence of the model
and the learning of imitation and non-imitation.
J. Exptal. Psych., 1962, 6j£, 183-190.
Rosenblith, J. F. Learning by immitation in kindergarten
children. Child Development. 1959, ^0., 69-80.
Rosenblith, J. F. Imitative color choices in kindergarten
children. Child Development. 1961, ^2, 211-223.
Rosenthal, T., Zimmerman, B., and Durning, K. Observa-
tionally induced changes in children’s interrogative
classes. J. Pers. Soc. Psych.. 1970, l£, 681-688.
Ross, D. Relationship between dependency, intentional
learning, and Incidental learning in preschool
children. J. Pers. Soc. Psych., 1966, 37^-381.
Sheffield, F., and Maccoby, N. Summary and interpretation
of research on organizational principles in construc
ting filmed demonstrations. In A. A. Lumsdalne
(ed.), Student Response in Programmed Instruction.
122
Washington, D.C.: National Academy of Sciences—
National Research Council, 1961. Pp. 117-131.
Tulving, E. Subjective organization and effects of
repetition in multi-trial free recall learning.
J. Verbal Learning and Verbal Behavior, 1966, £,
193-197.
Tulving, E. Theoretical issues in free recall. In T. R.
Dixon, and D. L. Horton (eds.), Verbal Behavior
and General Behavior Theory. Englewood Cliffs.
JTT7! Prentice-Hall, 196b. Pp. 2-36.
Twining, W. Mental practice and physical practice in
learning a motor skill. Research Quarterly, 1949,
2£, 432-435.
Vandell, R., Davis, R., and Clugston, H. The function of
mental practice in the acquisition of motor skills.
J. of General Psych., 1943, 2£, 243-250.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Weight Reduction As A Function Of The Timing Of Reinforcement In A Covertaversive Conditioning Paradigm

PDF

An Examination Of Positive And Negative Reinforcement In Classical And Operant Conditioning Paradigms In The Primary Psychopath

PDF

Modification Of Low Self-Confidence In Elementary-School Children By Reinforcement And Modeling

PDF

Looking Back After Coming Down: Conformity And Commitment In Campus Protest

PDF

Verbal And Imaginal Processes In Learning

PDF

Death Anxiety In Leukemic Children

PDF

The Development And Evaluation Of Three Therapeutic Group Interventions For Widows

PDF

Dissonance and Self-Perception Analysis of "Forced Compliance": When Two Theories Make Competing Predictions

PDF

The Effects Of Diphenylhydantoin On The Galvanic Skin Responses Of Psychopathic And Normal Prisoners

PDF

The Effects Of Nonreward Dissonance And Secondary Reward On Extinction And Attractiveness

PDF

The effects of an RNA polymerase in the improvement of verbal learning and memory

PDF

Tension And Anxiety In Deconditioning

PDF

Relationship Between Perceptual Style, Achievement, And Childrearing Practices In Elementary-School Boys And Girls

PDF

Diagnostic indices of hyperactivity and an investigation of verbal mediation training as an alternative to pharmacotherapy

PDF

Self-Sacrifice, Cooperation, And Aggression In Women Of Varying Sex-Role Orientations

PDF

The Effects Of Repression-Sensitization, Race, And Levels Of Threat On Extensions Of Personal Space

PDF

The effects of physical proximity and body boundary size on the self-disclosure interview

PDF

The Effect Of Personalized Emotional Stimuli On Asthmatic Reactions

PDF

Strategies Of Marital Communication

PDF

Prognostic Expectancy Effects In The Desensitization Of Anxiety Over Invasion Of Body Buffer Zones

Asset Metadata

Creator Vincent, Steven Max (author)

Core Title Attention, retention, and incentive processes in observational learning

Degree Doctor of Philosophy

Degree Program Psychology

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

Tag OAI-PMH Harvest,psychology, experimental

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Frankel, Andrew Steven (committee chair), Marston, Albert R. (committee member), Tiber, Norman (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-801948

Unique identifier UC11364469

Identifier 7314452.pdf (filename),usctheses-c18-801948 (legacy record id)

Legacy Identifier 7314452

Dmrecord 801948

Document Type Dissertation

Rights Vincent, Steven Max

Type texts

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

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

Repository Name University of Southern California Digital Library

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