THEORETICAL

WEAKNESSES LEAD TO PRACTICAL PROBLEMS:

THE EXAMPLE OF SELF-EFFICACY THEORY
CHRISTINA
University

of Newcastle,

LEE
Australia

Summary

Banduras
self-efficacy
concept has received widespread
acceptance
from
psychologists dealing with a varied range of behaviors. This paper argues that, like many
related models, its ability to explain human behavior is largely illusory. It has its basis in the
argument that behavior arises from complex interactions between unobservable
variables
which are vaguely defined and cannot be assessed. For example. the process by which efficacy
expectations arise from various sources of information,
and the interaction of self-efficacy
with levels of skill and motivation,
are argued to be important,
but there is no model to
explain how these processes occur. Since unambiguous predictions cannot be made, it is
imoossible to test the model in a scientific sense. This weakness seriously compromises the
practical usefulness of the theory.

essentially non-scientific nature of the models

they use. but continue to use them anyway.
Practitioners get results while using the models
and, naturally although illogically, assume that
the results occur because of the models. Besides, there do not appear to be any viable
alternatives.
Theoretical
precision may be
viewed as a luxury that the busy psychologist
cannot afford.
This paper argues that matters of scientific
verifiability and precise measurement are not
esoteric trivia for the academic psychologist to
play with, but essential practically. In particular, a reliance on unobservable variables and a
lack of precision are two major weaknesses
which lead to confusion and result in practical
problems. I will use Banduras theory as an
example, to illustrate that lack of a scientific
basis leads inevitably to a lack of internal
clarity. I should make clear that Banduras
theory is used simply because it is so popular
and well-known, not because it has particular
faults by comparison with other models.
Lack of clarity in many theories means that
such theories eventually fail practically as well

Self-efficacy theory was first proposed just

over a decade ago (Bandura, 1977) and has
quickly gained widespread acceptance. It is
because of this rapid acceptance that I intend
in this paper to use it to illustrate a general
argument about explanations of human behavior. Self-efficacy theory, in common with
numerous other cognitive, social-learning, and
systems-based frameworks for understanding
human behavior, has apparent practical appeal. However, there are also sound practical
reasons for rejecting these theories.
Such theories have frequently been criticised
because they rely for their explanatory power
on unpredictable
interactions between variables which are unobservable
and vaguely
defined (e.g., Eysenck, 1978; Lee, 1987; Skinner, 1977, 1987). This means they are prescientific and cannot be falsified (Lang, 1978).
While psychologists might be convinced by
the logic of such an argument, however, the
behavioral effects of such criticisms are negligible. Behaviorists remain a small proportion of
psychologists (Baer, Wolf, & Risley, 1987). It
appears that many psychologists accept the
Requests for reprints should be addressed
N.S.W.,
2308, Australia.

to Christina

Lee, Ph.D.

11.5

Department

of Psychology,

University

of Newcastle,

116

CHRISTINA

as philosophically.
The view that the scientific
basis of a theory is irrelevant
to its practical
usefulness
cannot be sustained,
and there are
practical reasons why pragmatic psychologists
should consider issues of theories and theory
construction.

Self-efficacy

Theory:

Widespread

Acceptance

Banduras (1977) initial self-efficiency

paper
generated
a strong
positive
reaction
(e.g.,
Kazdin,
1978: Kendall
& Korgeski,
1979;
Wilson, 1978), together with critical and theoretical debate (e.g., Borkovec,
1978; Eysenck,
1978; Lang, 1978; Tryon,
1981). The model
was accepted as potentially
useful in a range
of clinical settings, to assess progress and assist
in the selection
of therapeutic
goals. However, some basic questions concerning
aspects
of the theory as an explanation
of human
behavior
were
raised.
The difficulties
involved in unambiguous
measurement
of selfefficacy (e.g., Kazdin,
1978; Wilson,
1978)
were discussed
and the point
made that,
without
precise measurement,
theory-testing
cannot produce clear-cut
results.
Bandura
has readily
replied
to criticisms
(e.g., Bandura.
1978, 1980, 1984, 1986), responding
to specific points and expanding
on
the theory. He has argued consistently
that the
theory identifies
a causal mechanism
(selfefficacy) which is the proximal determinant
of
the initiation
of, and persistence
with, behavior. Although a complex process of reciprocal influence
among variables
is an essential
feature
of Banduras
model,
self-efficacy
is
described
as the most immediate
cause of
behavior.
Critical debate has continued,
but recently
authors
have restricted
themselves
more to
matters of emphasis (e.g., Eastman & Marzilher, 1984). of method (e.g., Cervone.
1985;
Kirsch, 1986), or of terminology
(e.g., Kirsch,
1986) within the framework of self-efficacy and
social-learning
theories.
Questions
regarding
the validity
of the basic propositions,
the

LEE

assumptions
on which the theories are based,
or the predictive
value of the assessment
of
cognitions
in general,
are less often asked.
Meanwhile,
the number of papers which use
self-efficacy measures or invoke the concept to
explain behavior
has grown exponentially
in
the past decade. Self-efficacy data now appear
in such diverse publications
as Humarz Refations, Addictive Behaviors, and the Journal of
Vocationaf Behavior, as well as the more
centrally
behavioral
and cognitive-behavioral
journals.
Self-efficacy
has been measured and
reported
in connection
with a wide range of
behaviors,
including
the traditional
clinical
areas (e.g., Katz, Stout, Taylor,
Horne,
&
Agras, 1983), intellectual
development
(e.g.,
Schunk,
1983), health-related
activities (e.g.,
OLeary,
1985), and sporting
performance
(e.g., Lee, 1982). Such unexpected
topics as
the avoidance
of professional
burnout (Meier,
1983), adjusting to retirement
(Holahan,
Holahan. & Belk, 1984), and the treatment
of child
molestors
(Segal & Marshall,
1986) have also
been examined
within a self-efficacy
framework. To generalize,
these and many similar
found efficacy
investigations
have, overall,
expectations
to be reasonably
accurate
predictors of outcome.
Other social-learning
theorists have also included the concept of selfefficacy in the development
of complex models
of the processes
of behavior
change
(e.g..
Marlatt & Gordon,
1985).
The widespread
acceptance
of the concept is
further
demonstrated
by the appearance
of
papers which make use of the term in a very
general sense, more or less as a fashionable
synonym
for self-confidence
(e.g., Barling &I
Snipelisky,
1983; Devins et al., 1982), without
adhering to Banduras specific definition of the
term within a particular
theoretical
context.
This development
can be seen as an indication of a widespread
familiarity
with, and
acceptance
of, the term, though not the concept or the implications
of its theoretical
basis.
This process is analogous
to the way in which
Rotters concept of locus of control, developed
within
a social-learning-theory
framework,

Self-efficacy

came to be taken up and used by trait theorists

in a way quite contrary to that intended by its
original author (Rotter,
1973). The fact that a
term, defined within the context of a particular
theory, has come to be used generally outside
that context does not in itself strengthen
or
weaken the theory. It merely indicates a strong
awareness
within the research community
of
the terminology.
Authors
who use the term in the sense in
which it was originally
intended
find results
which are generally
supportive
of the model,
and criticism
of the theory,
although
still
appearing
(e.g.,
Kirsch,
1985), has slowed
markedly.
Are we to conclude from this state
of affairs that self-efficacy theory has answered
all its detractors,
and now has only to beware
of imitations?

Levels of Explanation
In assessing
a theory,
it is important
to
distinguish
between
its ability to develop
a
believable
but essentially
metaphorical
description of a process and its ability to explain
that process.
Although
self-efficacy
theory
appears
to be a valuable
tool clinically
and
empirically,
much of its apparent
explanatory
strength is illusory and is based on a confusion
of description
with explanation.
The tendency
to take a reasonable-sounding
description
of
how events might occur, and then assume it to
be equivalent
to an explanation
of those
events, is understandable
but it is not logical.
The self-efficacy
concept allows us to predict
behavior in many settings with a fair degree of
accuracy.
But at the level of explaining
the
processes involved in behaving and in behavior
change.
rather
than simply
providing
one
possible
description
of them,
it does not
increase our understanding
of human activity.
To state that a person behaved in a particular
way partly because of his or her level of selfefficacy is not a scientific explanation
of that
behavior.
The fact that self-efficacy
has no
empirical
existence
is seen by many authors

117

(e.g.,
Eysenck,
1978; Lee,
1987; see also
Skinner,
1977, 1987) as invalidating
the theory
as a casual explanation.
The point I wish to
make is that, even if one chooses to reject the
arguments
of these writers
concerning
the
relevance of scientific objectivity,
this aspect of
the theory means that its practical usefulness is
also limited. While the theory has strengths as
a metaphor for describing human behavior, its
weakness is that it is not a model for explaining
behavior.

Self-efficacy

Theory: A Reliance
Unobservable

on the

The theory emphasises

individuals
expectations and beliefs as crucial determinants
of
behavior.
In particular,
it argues that a persons efficacy
expectations,
beliefs about
whether
he or she has the capabilities
to
perform
particular
actions,
are major determinants of whether that person attempts those
actions.
They also determine
the extent to
which he or she will persist in the face of initial
failure,
the
level
of success
eventually
achieved,
and the level of intrinsic
interest
developed.
Self-efficacy
theory is based on the premise
that people develop specific expectations
concerning their ability to perform any task, and
that it is these expectations
which determine
action. Provided the individual
has appropriate skills and adequate
incentives
(Bandura,
1977, p. 19-t) for the performance
of a particular task, behavior can be predicted with a high
degree of accuracy from efficacy expectations.
Although
Bandura
(e.g., 1981, 1983, 1983.
1986) has argued consistently
against simple
models which propose unidirectional
determinants of behavior,
in favor of reciprocal influences between different classes of variable, he
makes it clear that he does regard efficacy
expectations
as causally
prior to behavior:
Self-percepts
of efficacy are not simply inert
predictors of future behavior.
Peoples beliefs
about their capabilities
influence
how they

118

CHRISTINA

behave
. . .. (Bandura,
1984, p. 242).
Although efficacy expectations
are not the only
causal factor in determining
behavior, they are
definitely seen as playing a central causal role.
An alternative view, that they are epiphenomenal reflections
of some underlying
physiological mechanism
(Eysenck,
1978) is explicitly
rejected
(e.g., Bandura,
1978).
Efficacy expectations
themselves
are seen to
arise from: the previous
experiences
of the
individual,
vicarious
experience
gained from
observation
of others, verbal inputs, and physiological state. These four interact to produce
the efficacy expectation,
a synthesis of information from all available
sources,
weighted
according
to the relative
salience
of each
source. It is not clear whether
there is any
degree of conscious awareness of this process,
which is described
in general terms.
The model certainly
has advantages
over
less complex models of behavior. In particular,
it suggests a potential
common
mechanism
whereby a variety of experiences
or therapeutic strategies could influence behavior in similar ways. However, the major flaw of proposals
of this kind is that they are based on undefined
and unobservable
interactions
between imprecisely defined variables. It should be noted that
this flaw is by no means unique to self-efficacy
theory, but is also a weakness
of a range of
other cognitive
models of human
behavior
(e.g., Guidano
& Liotti, 1983).
Many influential
theoreticians
(e.g., Mahoney, 1974) have argued that it is both legitimate and necessary
to infer the existence
of
unobservable
processes in order to understand
human behavior. However, making such inferences causes more difficulties
than it solves;
the unverifiable
nature of the variables and of
the resultant models renders them unscientific,
and their vagueness
and ambiguity
render
them unworkable
(e.g., Lee, 1987; Skinner,
1977).
The Need for Precision
The brief description

in the previous

section,

LEE

of the process by which efficacy expectations

develop,
highlights
a major weakness
of the
theory as an explanation
of human behavior.
This is that the process by which the synthesis
of information
takes place is not specified.
There is no model by which one could predict
the relative
weighting
of various
types of
information.
Therefore,
there is no way of
predicting
from a knowledge
of these inputs
what efficacy expectation
will result, and no
way of developing
techniques
which will reliably alter efficacy expectations.
Because of the
large number of potential influences described,
and because of the lack of a rigorous model,
almost any level of efficacy could potentially
arise from almost any set of circumstances.
Without the development
of a framework
for
predicting
how efficacy expectations
develop
and change as a result of the input of information, the theory is no more than a metaphorical
description
of a hypothetical
process.
The
practical
usefulness
of this description
is
severely limited,
and the practical
limitation
derives directly from the theorys inadeqaucy
at a formal level.
The problem of undefined
and unobservable
interactions
is not confined to the development
of the efficacy expectation.
A similar problem
arises in describing
how efficacy expectations
interact
with other variables.
For example,
Bandura argues that skill levels and incentives
are also important
influences
on behavior.
Self-efficacy
is described as one component
of
a complexly
interacting
system of variables,
but as that component
which has the most
direct influence on behavior (Bandura,
1984).
Such a model reflects the complexity
and
variability
we perceive
in human
behavior.
However, there is no framework for specifying
how the efficacy expectations
will interact with
other relevant variables, and thus of predicting
what behavior will result from varying levels of
skill, efficacy and incentive.
Further,
I would
argue that the vagueness of the basic constructs
precludes the successful development
of such a
framework.
Again,
therapeutic
applications
must perforce be less than reliable, because of

Self-efficacy

a weakness at the level of formal theory.

Thus, the theory postulates a number of
unobservable processes and argues that they
interact in an unspecified manner to produce
observable behavior. The lack of precision
means that hypotheses cannot be developed
which will test the adequacy of the theory. It
also means that techniques cannot be developed which will reliably assist in dealing with
human problems.

Distinctions Between Efficacy and Other

Expectations
Even the end product of the hypothesised
process, the efficacy expectation itself, has
proved difficult to measure and to distinguish
from other types of self-report. The difficulties
involved in unambiguous measurement of selfefficacy (e.g., Kazdin, 1978; Wilson, 1978)
were discussed when the theory was first
proposed and the point made that, without
precise measurement,
theory-testing
cannot
produce clear-cut results. Much of the theoretical research relating to self-efficacy theory in
recent years has focused on the relative importance of efficacy expectations and various other
cognitions in the prediction of behavior (e.g.,
Hackett, 1985; Kavanagh & Bower, 1985;
Maddux, Norton, & Stoitenberg, 1986) and on
how exactly efficacy expectations differ from
other cognitions (e.g., Godding & Glasgow,
1985; Kirsch, 1982; Leary & Atherton, 1986).
Although the research and theoretical arguments have been cogent and well-articulated,
no general agreement has been reached. This
lack of a firm conclusion does not stem from
any deficiency in the quality of the research;
rather, it stems from the lack of theoretical
clarity and objectivity.
The relationship between efficacy and outcome expectations, for example, has been a
subject of considerable
contention.
While
several theories explain behavior in terms of
expectancies, self-efficacy theory differs from
others (e.g., Abramson, Seligman, & Teas-

119

dale, 1978) in its focus on ability rather than on

the outcome of particular actions. Efficacy
expectations
and outcome
expectations
have been distinguished, although there is
considerable disagreement as to exactly how.
Bandura (e.g., 1978, 1982) has argued that
efficacy expectations and outcome expectations are conceptually distinct and can be
distinguished empirically by the way in which
they affect behavior. Others (e.g., Eastman &
Marzillier, 1984), however, have argued that
the distinction between the two, in practical
terms, is unclear. If the theory is to make an
important distinction between them, then an
inability to do so reliably weakens the theory
considerably. This again emphasises the practical problems which arise from theoretical
weaknesses.

Problems of Definition
When one actually comes to measure efficacy and outcome expectations separately,
problems of definition arise. Maddux and
Barnes (1986) argue that the efficacy expectation refers to ones ability to perform a
particular behavior, and outcome expectancy
to what specific outcome is expected. While
this may be appropriate in some behavioral
domains (Lee, 1984a), in many therapeutic
situations there is no question as to what the
observable outcome of a particular behavior
will be. For example, many investigations of
self-efficacy have dealt with snake phobia
(e.g., Bandura, Adams, & Beyer, 1977; Lee,
1984b) and have involved participants in behavioral avoidance tests with a snake. In
making judgments concerning items in this
test (e.g., place gloved hand in cage), there
is no question as to what the result of that
action will be. The salient issue is not what will
happen but how the target person will react to
performing the task. Lee (1984a, 1984b) therefore has defined outcome expectation as the
expected valence of the result, and asked
subjects to rate how positive or negative they

CHRISTINA

110

would feel if they were to perform the task.

Manning
and Wright (1983) compared
efficacy and outcome expectations
in a study of
women experiencing
childbirth.
Their definition of outcome
expectation
focused on the
belief that certain techniques,
if used. would be
effective in controlling
pain. It is difficult to see
whether this definition
matches Madduxs or
Lees, because the notion of controlling
pain is
simultaneously
concerned
with the actual outcome and the valence of that outcome.
Meier
(1983),
looking
specifically
at occupational
burnout,
has gone one step further and argued
for the existence of three types of expectation:
outcome
expectations,
efficacy expectations,
and reinforcement
expectations.
In this framework, a distinction
is made between the practical result of an action (outcome
expectation)
and the extent to which the person finds the
action and its result to be personally rewarding
(reinforcement
expectation).
As I will argue
later, such a distinction
is not always as clearcut as it might seem.

Theoretical

Difficulties Translate
Practical Problems

Into

It is unlikely that the debate will reach any

conclusion
if it continues at a semantic level, or
if different researchers
continue
to define and
measure variables in different ways. Its basis,
both at a practical and at a theoretical
level,
can be questioned.
Definitions
of what constitutes
a behavior,
an outcome,
and the reinforcing
value of that
outcome (if one is to accept Meiers suggestion
of a third class of expectation)
are not at all
clear cut. Take the example of the high jumper
used originally
by Bandura
(1978) and repeated in a variety of forms since then (e.g.,
Bandura,
1986; Eastman
6t Marzillier,
1984;
Maddux & Barnes,
1986). One could argue,
with Bandura,
that . . . the belief that one
can jump six feet is an efficacy judgment;
the anticipated
social recognition,
applause,
trophies,
and self-satisfaction
for such a per-

LEE

formance constitute the outcome expectations*

(Bandura,
1986, p. 391). In this model, the
reinforcement
expectation
would be the belief
that the social and personal
consequences
would be positively
reinforcing.
Alternatively,
and with equal logic, one
could argue that efficacy expectations
refer to
ones confidence
that one can perform the runup, take-off,
and jump actions with correct
form. Outcome
expectations,
in this framework, would refer to the belief that somebody
doing this would clear the bar, and reinforcement expectations
to the belief that clearing
the bar would result in winning the competition, satisfying
a goal, or being admired
by
others. (Incidentally,
this breakdown
accords
better with the behavioral
sport psychologists
focus on process and effort rather than results;
see, for example,
Allison & Ayllon,
1980.)
It is difficult to see how one could mount a
logical argument
to distinguish
the relative
of these two arguments.
The
correctness
difference lies in how we choose to define any
particular act. If an event can be a component
of a behavioral
act, a component
of an outof a consequence,
come, or a component
depending
on the way one chooses to describe
it, a model which makes a point of distinguishing
between
expectations
concerning
those
concerning
outcomes,
and
actions,
(perhaps)
those concerning
the reinforcing
value of those outcomes
must break down. It
becomes completely
impossible
to develop an
adequate
explanation
for behavior
when the
variables purported
to be affecting each other
cannot be distinguished
from each other.
This brief example
indicates
the largely
semantic basis of any effort to separate these
different
types of expectation,
and highlights
an important
weakness,
not only of selfefficacy theory, but of many modern cognitive
theories which propose the existence and interaction of various beliefs and expectations
as
an important cause of observable
behavior, but
fail to provide adequate
means of assessing
them. Staddon
(1984) has argued that Banduras descriptions
of the causal links between

Self-efficacy

internal variables are vague and ill-defined,

and give the illusion of precision without the
reality (p. 507). This problem is by no means
unique to self-efficacy theory. Many cognitivebehavioral theories rely on complex but undefined relationships
between hypothetical
cognitive structures in order to deal with the
complexity and variability of human behavior
(Lee, 1987). Although they may give the
impression of dealing with variables which
appear important in understanding
human
behavior, their lack of definitional rigour renders them something less than scientific. It
seems premature to reject a behavioral and
empirical approach to psychology in favor of
an approach which is characterized by vagueness and lack of scientific rigour.
Taking Things at Face Value
A further problem with the development of
vague descriptive framework for why people
behave the way they do is that data which
appear to support them could equally support a
range of alternative models. For example, a
large amount of the self-efficacy literature
demonstrates that efficacy expectations correlate with actual behavior (e.g., Di Clemente,
Prochaska,
& Gibertini,
1985; Weinberg,
Gould, Yukelson, & Jackson, 1981) or behavioral intentions (e.g., Hackett, 1985; Maddux, Sherer, & Rogers, 1982). Other empirical
studies show that efficacy expectations parallel
behavioral and emotional changes during therapy (e.g., Bandura, Adams, & Beyer, 1977;
Williams, Dooseman, & Kleifield, 1984; Wilhams, Turner, & Peer, 1985) or that they
correlate with physiological measures (e.g.,
Bandura, Taylor, Williams, Mefford, & Barchas, 1985; Barrios, 1983).
While such findings are consistent with selfefficacy theory, they are equally consistent
with a variety of other models. The central
tenet, that efficacy expectations are fundamental causes of other phenomena, is not tested.
Biglan (1987) argues that self-efficacy research
results could equallv
through a
1
s be exolained
1

121

model which proposes that all the variables.

expectations,
behaviors,
and physiological
changes, result directly from environmental
events. The concept that all such variables are
parallel responses to causal events in the
environment has also been argued more generally by Wilkins (1986), who takes the view
that expectancies of any kind are best seen as
reflections of behavior, not as causes.
Lang (1978) has made the related point that
self-reports of mental processes need not have
any veridical relationship to the actual causes
of motor behavior, cognition, or emotion.
Social psychology has a number of theories
dealing with the pitfalls of accepting selfreports at face value. Self-perception theory
(e.g., Bern, 1972), for example, takes the view
that individuals main source of information
about their motives and beliefs is through
observation of their own behavior. Individuals
are then assumed to concoct expectations and
beliefs which accord with what they see themselves doing. While there is considerable de:
bate about that issue, there is evidence (e.g.,
Nisbett & Wilson, 1977) to suggest that in at
least some situations, individuals are unaware
of the cognitive processes involved in making
behavioral choices, and are unable to recognize or describe the stimuli which have caused
them to alter their behavior. Such arguments
suggest that accepting self-reports at face value
may be inappropriate.

Conclusion
The arguments presented in this paper are
not simply scientific purist ones; I have
attempted to illustrate that there are good
practical reasons why non-verifiable theories
will eventually fail to provide practical techniques that can be used. They make no clear
distinctions between various cognitions, nor do
they provide any clear model of the relationships between these c%nitions and observable
antecedents and consequences. However, they
declare that we need to change these cogni-

122

CHRISTINA

tions in order to have a therapeutic effect on

individual problems. Practically they break
down, and their practical limitations are due to
their scientific failings.
My argument is that self-efficacy theory, and
other similar theories, do not aid our understanding of human behavior. At the level of
theoretical explanation,
the statement that
people behave in certain ways because of their
efficacy expectations is not testable. At the
level of practical techniques, it is not useful.
Banduras self-efficacy theory is a vague
descriptive model, not an explanatory theory.
It provides clinicians and researchers with a
non-scientific framework which describes, in
metaphorical terms, a hypothetical process by
which behavior might result from current
environmental,
and other, influences. There
are other, equally plausible, explanations for
the research findings which are claimed to
support this model.
The main practical weakness of this and
similar theories is their inability to make
precise predictions. The undefinable nature of
the processes assumed to be involved means
that there is no clear relationship between
observable
inputs and observable
results.
Thus, precise predictions cannot be made and
crucial tests of the theories are impossible. I
argue that this is not simply an academic
point of view, but one which must be considered if practical and useful advances in
psychological techniques are to be made.
References
Allison, M. G.. & Ayllon, T. (1980). Behavioral
coaching
in the development
of skills in football, gymnastics,
and
tennis. Journal of. Applied
Behavior Analysis, 13, 297..
314.
Abramson,
L. Y., Seligman, M. E. P., & Teasdale,
J. D.
(1978). Learned
helplessness
in humans:
Critique and
reformulation.
Journal of Abnormal Psychology, 87,4974.
Baer, D. M.. Wolf, M. M., & Risley, T. R. (1987). Some
still-current
dimensions
of applied
behavior
analysis.
Journal of Applied Behavior Analysis,
20, 313-327.
Bandura,
A. (1977). Self-efficacy:
Toward
a unifying
theory of behavior
change.
Psychological Review, 84,
191-215.

LEE
Bandura, A. (1978). Reflections on self-efficacy.
Advances
in Behuviour Research and Therapy, 1, 237-269.
Bandura,
A. (1980). Gauging
the relationship
between
self-efficacy
judgment
and action.
Cognirive Therapy
and Reseurch, 4. 263-268.
Bandura,
A. (1981).
In search of pure unidirectional
determinants.
Behavior Therapy, 12, 30-40.
Bandura,
A. (1982). Self-efficacy
mechanism
in human
agency. American Psychologisr, 37, 122-147.
Bandura.
A. (1983). Temporal
dynamics and decomposition of reciprocal
determinism:
A reply to Phillips and
Orton.
Psychological
Review. 90, 166-170.
Bandura,
A. (1984). Recycling
misconceptions
of perceived self-efficacy.
Cognirive Therapy and Research, 8,
231-255.
Bandura,
A. (1986). Social foundurions of thought and
action. Englewood
Cliffs. NJ: Prentice-Hail.
Bandura, A., Adams, N. E., 91 Beyer, J. (1977). Cognitive
processes
mediating
behavioral
change.
Journal of
Personaliry and Social Psychology,
35, 125-139.
Bandura.
A., Taylor,
C. B.. Williams, S. L., Mefford.
1. N., & Barchas, J. D. (1985). Catecholamine
secretion
as a function of perceived coping self-efficacy.
Jourrrul of
Consulring and Clinical Psychology, 53. 406-414.
Barling,
J.. & Snipelisky,
B. (1983).
Assessing
the
determinants
of childrens academic self-efficacy
beliefs:
A replication.
Cogniuve Therapy and Research, 7. 371376.
Barrios, B. A. (1983). The role of cognitive mediators
in
heterosocial
anxiety:
A test of self-efficacy
theory.
Cognidve Therapy and Research, 7. 543-554.
Bern, D. J. (1972). Self-perception
theory. Advances in
Experimeuml Social Psychology, 6, l-62.
Biglan, A. (1987). A behavior-analytic
critique of Banduras self-efficacy
theory. Behavior Anal.vsr, 10. l-15.
Borkovec,
T. D. (1978). Self efficacy: Cause or reflection
of behaviour
change? Advances in Behaviour Research
and Therapy, 1, 163-170.
Cervone,
D. (1985). Randomization
tests to determine
significance
levels for microanalytic
congruences
between self-efficacy
and behavior.
Cognirive Therapy and
Research, 9, 357-365.
Devins, G. M.. Binik. Y. M., Gorman,
P., Dattel. M..
McClosky. B., Oscar, G., & Briggs, J. (1982). Perceived
self-efficacy,
outcome expectancies,
and negative mood
states in end-stage
renal disease. Journal of Abnormal
Psychologv. 91, 241-244.
Di dIeme&.
C. C., Prochaska.
J. 0.. & Gibertini,
M.
(1985). Self-efficacv
and the stages of self-chance
of
smoking. Cognitive*Therapy and Research. 9, 181-200.
Eastman,
C.. & Marzillier,
J. S. (1984). Theoretical
and
methodological
difficulties
in Banduras
self-efficacy
theory.
Cognitive Therapy and Research, 8, 213-229.
Eysenck, H. J. (1978). Expectations
as causal elements in
behaviour
change. Advances in Behaviour Research and
Therapy. 1, 171-175.
Godding,
P. R.. & Glasgow,
R. E. (1985). Self-efficacy
and outcome
expectations
as predictors
of controlled
smoking status. Cognitive Therapy and Research, 9,58%
590.
Guidano,
V. F., & Liotti, G. (1983). Cognifive processes
and emotional disorders. NY: Guilford.
Hackett,
G. (1985). Role of mathematics
self-efficacy
in

Self-efficacy
the choice of math-related
majors of college women and
men: A path analysis. /ournal
of Comrseling
Psychology, 32. 47-56.
Holahan.
C. K., Holahan,
C. J., & Belk. S. S. (198-l).
Adjustment
in aging: The roles of life stress, hassles. and
self-efficacy.
Health Psychology, 3, 315-328.
Katz, R. C., Stout. A., Taylor,
C. B.. Horne,
M.. &
Agras. W. S. (1983). The contribution
of arousal and
performance
in reducing spider avoidance.
Behavioural
Psychorherapy. II. 127-138.
Kavanagh,
D. J., & Bower, G. H. (198.5). Mood and selfefficacy:
Impact
of joy and sadness
on perceived
capabilities.
Cognitive Therapy and Research, 9, 507526.
Kazdin. A. E. (1978). Conceptual
and assessment
issues
raised by self-efficacy
theory. Advances in Behaviour
Research and Therapy, I. 177-185.
Kendall, P. C., & Korgeski, G. P. (1979). Assessment
and
cognitive-behavioral
interventions.
Cognitive Therapy
and Research, 3, 1-21.
Kirsch, 1. (1982). Efficacy expectations
or response predictions: The meaning of efficacy ratings as a function of
task characteristics.
Journal of Personality and Social
Psychology, 42, 132-136.
Kirsch, I. (1985). Self-efficacy
and expectancy:
Old wine
with new labels. Journal of Personality and Social
Psychology, 49, 825-830.
Kirsch, I. (1986). Early research on self-efficacy:
What we
already know without
knowing we know. ~Journal of
Social and Clinical Psychology. 4. 339-358.
Lang. P. J. (1978). Self-efficacy
theory:
Thoughts
on
cognition
and unification.
Advances in Behavi&r Research and Therapy, I, 187-192.
Leary. M. R., & Atherton,
S. C. (1986). Self-efficacy,
social anxiety, and inhibition
in interpersonal
encounters. Journal of Social and Clinical Psychology, 4, 256
267.
Lee, C. (1982). Self-efficacy as a predictor of performance
in competitive
gymnastics.
Journal of Sport Psychology.
4, -Io5--109.
Lee, C. (1981a).
Accuracy
of efficacy
and outcome
expectations
in predicting
performance
in a simulated
assertiveness
task. Cognitive Thercpy and Research, 8.
3748.
Lee, C. (198lb).
Efficacy
expectations
and outcome
expectations
as predictors
of -performance
in a snakehandling task. Cognitive Therapy and Research, 8, 50%
516.
Lee, C. (1987). Affective
behavior
modification:
A case
for empirical investigation.
Journal of Behavior Therapy
and E.iperimental Psychiatry, 18, 2b3-213.
..
Maddux. J. E.. & Barnes. J. f~1986). The orrhoeonalitv and
relative predictive urility of self-efficacy expectancy,
outcome expectancy, and outcome value: A review of
empirical studies. Unpublished
manuscript,
George
Mason University.
Maddux.
J. E., Norton,
L. W., & Stoltenberg.
C. D.
(1986). Self-efficacy
expectancy,
outcome
expectancy,
and outcome
value:
Relative- effects
on behavioral
intentions. Journal of Personality and Social Psychology,
51, 783-789.
I

123

Maddux. J. E., Sherer. M.. & Rogers. R. W. (1982). Selfefficacy

expectancy
and outcome
expectancy:
Their
relationship
and their effects on behavioral
intentions.
Cognitive Therapy and Research. 6. 207-2 11.
Mahoney,
M. J. (1973). Cognition and behavior modificalion. Cambridge,
MA: Ballinger.
bl. M.. & Wright, T. L. (1983). Self-efficacy
Manning,
expectancies,
outcome expectancies,
and persistence
of
pain control
in childbirth.
Journal of Personaliry and
Social Psychology, 45, 2-1113 1.
Marlatt, G. A.. & Gordon,
J. R. (Eds.). (1985). Relapse
preveniion. NY: Guilford
Press.
Meier, S. T. (1983). Toward a theory of burnout.
Human
Relations, 36. 899-910.
Nisbett, R. E., & Wilson, T. D. (1977). Telling more than
we can know: Verbal
reports
on mental
processes.
Psychological Review, 84. 231-256.
OLeary,
A. (1985). Self-efficacy
and health. Behaviour
Research and Therapy, 4, 337--151.
Rotter. J. B. (1973). Some problems and misconceptions
relating
to the construct
of internal
versus external
control
of reinforcement.
Journal of Consulfing and
Clinical Psychology, 43. 56-67.
Schunk.
D. S. (1983).
Reward
contingencies
and the
development
of childrens skills and self-efficacy.
Journal of Educarional Psychology, 75, 51 l-518.
Segal. Z. V.. & Marshall,
W. L. (1986). Discrepancies
between
self-efficacy
predictions
and actual performante in a populatidnof
rapists and child molesters.
Cognitive Therapy and Research. IO, 363-375.
Skinner,
B. F. (1977).
Why I am not a cognitive
psychologist.
Behaviorism, j, l-10.
Skinner. B. F. (1987). Whatever
happened
to psychology
as the science of behavior?
American Psychologist. 42,
780-786.
Staddon.
J. E. R. (1981). Social learning theory and the
dynamics of interaction.
Psychological Review. 91. 502507.
Tryon.
W. W. (1981).
A methodological
critique
of
Banduras self-efficacy
theory of behavior change. Journal of Behavior Therapy and Experimental Psychiarry.
12, 113-111.
Weinberg,
R. S.. Gould, D., Yukelson, D., &Jackson,
A.
(1981). The effect of pre-existing
and manipulated
selfefficacy
on a competitive
muscular
endurance
task.
Journa-I of Sporr Psychology, 4, 345-35-L
Wilkins. W. (1986). Invalid evidence for exoectancies
as
causes: comment on Kirsch. American Psy;hologist. 41,
1387-1389.
Williams, S. L.. Dooseman.
G.. & Kleifield,
E. (198-t).
Comparative
effectiveness
of guided mastery and exposure treatments
for intractable
phobias.
Journal of
Consulring and Clinical Psychology, 52. 505-518.
Williams,
S. L., Turner,
S. M.. & Peer, D. F. (1985).
Guided mastery and performance
desensitisation
treatments for severe acrophobia.
Journal of Consulting and
Clinical Psychology, 53. 237-237.
Wilson. G. T. (1978). The importance
of being theoretical:
A commentary
on Band&as
Self-efficacy:
Toward a
unifying
theory of behavioral
change*. Advances in
Behaviour Research and Therapy, I. 217-230.