Anda di halaman 1dari 2

J. Behov. ?-h@r.1 Exp. P&G/.

Primed in Great Britain.

Vol. 12, No. 2, PP. 113-l 14, 1981

A METHODOLOGICAL

0005.7908/81/020113-02 $02.W/O
b 1981 Pergamon Press Ltd.

CRITIQUE OF BANDURAS

THEORY OF BEHAVIOR

SELF-EFFICACY

CHANGE

WARREN W. TRYON
Fordham University
Summary-This brief theoretical note identifies a major methodological problem in Banduras
argument in favor of self-efficacy theory and provides two methods of controlling for it. His
data could likely be accounted for by social contingencies operating within his highly structured
behavioral approach situation.

Bandura (1977) introduced the self-system as a


unifying central mediator that is influenced
by performance
accomplishments,
vicarious
experiences, verbal persuasion and physiological
states. Moreover, the self-system is hypothesized
to distinguish between efficacy expectations
and outcome expectations. Efficacy expectations refer to whether the person believes he can
or will perform certain behaviors while outcome
expectations refer to anticipated effects the
behaviors are likely to have on the environment
if performed.
Bandura reported data demonstrating that
efficacy expectations correlate better with actual
performance in a behavioral approach test
than scores derived from performance measures
during treatment: such as the number of
hierarchy steps completed. He then concluded
that these results validated his hypotheses
concerning the self-system being the central
mediating construct unifying all behavior change
data.
A re-examination of Banduras procedures
reveals a very troublesome point touched upon
previously by Borkovec (1978). Efficacy expectations were operationally obtained by asking
each subject whether or not he would actually
perform each of the behaviors involved in the
behavioral approach test. The subjects were
given the behavioral approach test, and 83-89070

congruence between the subjects verbal and


motor behaviors was reported. Bandura was
quick to interpret this result as proving the
validity of self-efficacy theory. He did allow for
the possibility that some other superordinate
mediator might account for his results; but
omitted any discussion of how alternative
such as social reinforcement
explanations,
contingencies, might do so.
Consider the test situation again. A subject
states that he will perform behavior X and not
behavior Y. He is then asked to do behaviors X
and Y. One would expect no less than 83-89%
congruence with his statement as most people
have been shaped by social contingencies to
congruence between their verbal and motor
behaviors, especially when they are volunteering
in highly structured situations closely monitored
by authority figures. Bandura appears to
hypothesize that verbal and motor behaviors
are orthogonal !
Bandura is either unaware of or has ignored
the existing literature on the susceptibility of
behavioral approach tests to social contexts.
Bernstein (1973) reported that subjects avoided
phobic objects much more in a clinical situation
than in an experimental context unrelated to
fear assessment. The effects of context on
behavioral assessment have also been documented
by Bernstein and Nietzel (1973; 1974); Smith,

Reprint requests should be sent to: Warren W. Tryon, Department of Psychology, Fordham University, Bronx, New
York 10458, U.S.A.
113

114

WARREN

Diener and Beaman (1974), Tryon and Tryon


(1974). Any acceptable demonstration
of selfefficacy
theory
must squarely
address
and
resolve this crucial methodological
problem.
One important
control would be an experimental condition
that minimizes the effects of
social demand
on theoretical
predictions.
It
would
separate
the assessment
of efficacy
expectations from the evaluation of performance.
It would minimize cues concerning
surveillance
of behavioral performance.
The subjects would
be enrolled in what are outwardly two separate
experiments, one involving efficacy expectations,
the other involves performance
measures. Or,
efficacy expectations
could be obtained as part
of an interview having a much wider scope and
then performance
measures could be obtained
subsequently
in an unobtrusive
manner.
Another important
control condition
would
evaluate self-efficacy theory by maximizing
the
effects of social demand
but in a direction
opposite to that predicted by the theory. This
control condition would establish the magnitude
of the effect of self-efficacy position relative to
that of social demand. One approach would be
to tell subjects that the experimenter
is replicating professor Xs theory that people always do
what they say they will do. The experimenter
would go on to say that while this makes
good common
sense, such a theory is often
wrong as numerous
reasons exist why people
do not always behave as they say they will. The
experimenter
would explicitly say that this is the
alternative hypothesis under study. Then subjects
would be asked what they will do and their
behavior thereafter observed.
Another
approach
would be to allow the
social context to occasion
behaviors
incom-

W. TRYON

patible with self-efficacy


theory. Efficacy expectations would be assessed as part of a clinical
experiment
on phobias conducted
at the University health center by a staff
member
(cf. Bernstein,
1973). This would minimize
statements
of approach.
Then the subjects
would be taken to an experimental
psychology
laboratory
where the experimenter
was conducting a physiological
study of the effects of
fear. The subjects
would be instructed
to
approach the phobic object as closely as possible
to allow the experimenter
to obtain a valid
sample of their fear reaction
(cf. Bernstein,
1973). I predict that this control
condition
would demonstrate
that the notion
of selfefficacy judgements,
as a unifying theory of
behavioral
change
has serious limitations.
The apparent unification
is likely to be attributable to the controlling influence of social demand
across experimental
conditions.

REFERENCES
Bandura A. (1977) Self-efficacy:
Toward a unifying theory
of behavioralchange,
Psychol. Rev. 84,191-215.
Bernstein D. A. (1973) Situational
factors in behavioral
fear assessment:
A progress
report,
Behav.
Ther. 4,
41-48.
Bernstein
D. A. and Nietzel M. T. (1973) Procedural
variation
in behavioral
avoidance
tests, J. Consult.
Clin. Psychol. 41, 165-174.
Bernstein D. A. and Nietzel M. T. (1974) Behavioral avoidance tests: The effects of demand characteristics
and
repeated
measures
on two types of subjects,
Behav.
fher. 5,183-192.
Borkovec T. D. (1978) Self-efficacy:
Cause or reflection
of behavior change? Adv. Behov. Res. Ther. 1, 163-170.
Smith R. E., Diener E. and Beaman A. L. (1974) Demand
characteristics
and the behavioral avoidance measures of
fear in behavior
therapy
analogue
research,
Behav.
Ther. 5,172-182.
Tryon W. W. and Tryon G. S. (1974) Desensitization
and
demand characteristics,
Behov. Ther. 5,297-303.

Anda mungkin juga menyukai