Embed Size (px)
Citation preview
W&M ScholarWorks W&M ScholarWorks
Dissertations, Theses, and Masters Projects Theses, Dissertations, & Master Projects
1979
Motivated Attributions of Responsibility: Defensive Attribution Motivated Attributions of Responsibility: Defensive Attribution
Glenn Paule-Carres College of William & Mary - Arts & Sciences
Follow this and additional works at: https://scholarworks.wm.edu/etd
Part of the Psychology Commons
Recommended Citation Recommended Citation Paule-Carres, Glenn, "Motivated Attributions of Responsibility: Defensive Attribution" (1979). Dissertations, Theses, and Masters Projects. Paper 1539625077. https://dx.doi.org/doi:10.21220/s2-mpnf-1j25
This Thesis is brought to you for free and open access by the Theses, Dissertations, & Master Projects at W&M ScholarWorks. It has been accepted for inclusion in Dissertations, Theses, and Masters Projects by an authorized administrator of W&M ScholarWorks. For more information, please contact [email protected].
MOTIVATED ATTRIBUTIONS OF RESPONSIBILITY: DEFENSIVE ATTRIBUTION
A Thesis Presented to
The Faculty of the Department of Psychology The College of William and Mary in Virginia
In Partial Fulfillment Of the Requirements for the Degree of
APPROVAL SHEET
This thesis is submitted in partial fulfillment of
the requirements for the degree of
Master of Arts
Glenn Paule-Carres
Approved May 7 , 1979
!/ ^ u y-Ptter L. Derks ,Ph.D
Donelson Forsyth, Ph.D.
Shaver, Ph.D
E. Rae Harcum, Chairman Department of Psychology
TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS........................ ivLIST OF TABLES.... , .................. . .......... vLIST OF FIGURES.......... . viABSTRACT................ . .vliINTRODUCTION.............. 2METHOD. ........ 23RESULTS AND DISCUSSION .............................. . 35GENERAL DISCUSSION............. 54APPENDIX......... 59BIBLIOGRAPHY................. 63
ill
ACKNOWLEDGMENTS
The writer wishes to express his appreciation to Professor Kelly Shaver, under whose guidance this investigation was conducted, for the interest he has shown for the writer and this investigation. Thanhs are due also to Professors Larry Ventis, Peter Derks, and Donelson Forsyth for their comments and advice; to Professor Cynthia Null for her statistical aid; and to Dan Ozer and Robert Fulcher for their assistance in conducting the experimental sessions. Special appreciation is expressed to EvalyneCarres for her help in typing. Lastly, thanks to Pattimae
%
and .Larissa Paule-Carres who endured my absence during the early stages of this project and who have consistently provided support and encouragement during the course of this work.
iv
LIST OF TABLES
Table Page
1. Mean Attributions of Similarity .... . . .......... 372. Mean Attributions of Perpetrator
Responsibility. ................... 403. Mean Attributions of Perpetrator
Responsibility......... 434. Mean Attributions of Responsibility...... 465. Mean Attributions Relevant to
Derogation of Victim................. . 476. Mean Attributions of Similarity to Viet i m f s
Personality Traits. ...... 497. M e a n .Attributions of Effect of Accident
and Perpetrator Likeability ...... 508. Mean Attributions of Victim Blameworthiness... 51
v
LIST OF FIGURES
Figure Page
1. Bimodal Distribution of Builder'sReward Ratings................... 56
Vi
ABSTRACT
An experiment was conducted to investigate the attributions of responsibility made by witnesses of an accident. Using methodological improvements including those suggested by Fishbein and Ajzen (1973) and Chaikin and Darley (1973), the defensive attribution model of Shaver (1973) was tested against the just world hypothesis of Lerner (1966, 1967).
Eighty male undergraduates from the College of William and Mary viewed a videotaped accident and were asked to assign responsibility for the event using Heider's levels model of responsibility. A 2 x 2 x 2 factorial design was devised in which the subjects' alleged personal similarity to the accident participants, their identification with the accident participants, and their possibility of being in a similar situation were manipulated.
As predicted by the defensive attribution model results showed that subjects made lessened attributions of responsibility to the perpetrator when they felt a high possibility of being in a similar situation in the future. Results were inconclusive in regards to whether the victim would be leniently or harshly judged as a function of personal similarity. Suggestions were made concerning subject selection criteria and manipulation salience for future research of the defensive attribution phenomenon.
vii
MOTIVATED ATTRIBUTIONS OF RESPONSIBILITY
DEFENSIVE ATTRIBUTION
INTRODUCTION
Attribution theory deals with the active attempt of a
person to bring causal order and meaning into his percep
tions of events (Jones, Kanouse, Nisbett, Valins, & Weiner,
1972 ; Shaver, 1975). A person employs ̂ the process in order
,to better understand and predict observed behavior. Such
understanding and prediction is facilitated by the recognizing o f underlying regularities of behavior and by the
developing of causal chains which link events with their
antecedents (Heider, 1958). Having ascertained the event
and formulated hypotheses as to its occurrence, an observer
will compare the possible causes of an event and make an in
ference as to the probable cause of the 'observed behavior.
Within this theoretical framework, causes of events can be
attributed to any combination of three factors -- chance, en
vironmental circumstances, and personal dispositions. An
attribution of causality is based, however, on more than the
objective behavioral information available. The attributor
brings to the task assumptions, expectations, and personal
needs which will affect the attribution process by filling
in gaps in information, producing shifts in attention or
emphasis, and relating behavioral information to comparative
standards (.Jones et al. , 1972). Because of these subjective
2
3
influences, observers of the same event frequently draw
different and/or incorrect inferences about the causes of
the event.
Major attribution theorists have recognized the occur
rence of such "biased" attributions or "misattributions".
Heider (1958), the acknowledged founder of attribution the
ory, concluded that one factor in the attribution process
was the "affective significance" of an event. Personally
acceptable causes of events are sought, those which would
put the attributer in a good light. Jones & Davis (1965)
identified "hedonic relevance", the positive or negative
consequences of observed behavior for the perceiver, as an
instance of personal needs affecting attributions. Kelley
(1967) outlined five general instances in which attributional
errors could arise. These are cases in which: (1) the rel
evant situational constraints on the actor are ignored;
(2) "egocentric" assumptions are made in the absence of com
plete information; (3) the effects have affective signifi
cance for the perceiver; (4) the surrounding situation is
misleading; and (5) experimental manipulations deliberately
induce attributional errors.In spite of their recognition of "altered" or "distort
ed" attributions, the major theorists and researchers of
attribution processes have emphasized the objective, cogni
tive, information processing components of the phenomenon
(Brewer, 1977; Heider, 1958; Jones & Davis, 1965; Jones &
4
Nisbett, 1971; Kelley, 1967, 1972; Miller, 1978; Miller &
Ross, 1975; Ross, 1977). When persons produce discrepant
attributions of causality to an event, these researchers
hypothesize objective differences among the individuals as
resulting in the differing attributions. An example of this
emphasis is giyen ,,by Jones and Nisbett !s (1971) general hy
pothesis concerning the divergent perspectives of the
actor — that is, the individual who performs a given behav
ior -- and the observer of the behavior. They proposed that actors tend to perceive their behavior as a response to
situational cues, while observers tend to perceive the behav
ior as a manifestation of a disposition or quality possessed
by the actor. They suggested that the reasons for such
actor-obseryer differences is probably a simple perceptual
issue*. The actor's attention at the moment of action is
focused on the environment, the situational cues with which
his behayior is coordinated. For the observer, however, the
salient cues are the behavior of the actor.<
Jones and Nisbett offered another possible exp1anation
for a differential perceptual orientation of actors and ob-
seryers. They claimed that it stems from a difference in
the nature and extent of information that actors as opposed
to observers possess. In general, the actor knows more about
his past behavior and his present experiences than does the
obseryer. He is more likely to know about the extenuating
circumstances surrounding his behavior, and more likely to
5
locate causality away from himself.
Direct tests of this hypotheses (Harvey, Harris &
Barnes, 1975; Miller, 1975; Nisbett, Caputo, Legant, &
Maracek, 1973; Wolfsoa & Salancik, 1977) as well as indirect
sources (Duval & Wicklund, 1973; Jones & Harris, 1967; Jones,
Rock, Shaver, Goethals, & Ward, 1968; Regan & Totten, 1975;
Storms, 1973; Taylor & Fiske, 1975) have provided empirical
support for this notion of actor-observer differences.
Yet, despite the impressive track record of confirmation
of the Jones and Nisbett hypotheses, other empirical evi
dence suggests possible limitations (cf. Monson & Snyder,
1977). The general conclusion drawn by these researchers
is that attributions of actors and observers do differ, although not always in the direction suggested by Jones and
Nisbett (Apsler & Friedman, 1975; Cialdini, Braver, & Lewis,
1974; Langer & Roth, 1975). Equally interesting is the grow
ing tendency to explain differences in terms of motivational
distortions (Harvey, Harris, & Barnes, 1975; Miller & Norman,
1975; Monson & Snyder, 1977) or in terms of the conceptually
different tasks facing actors and observers when explaining
an event, with implications of motivated distortions (cf.
Buss, 19 78).
Another example of the emphasis on the objective, ra
tional aspects of the attribution process is a line of re
search derived from Shaw and Sulzer's study (1964) on attri
‘6
butions of responsibility. Shaw and Sulzer's approach was
to view attributions of responsibility in terms of the
attributer's level of cognitive sophistication (Sulzer, 1971)
or in terms of their stage of development (Fishbein' & Ajzen,
1973). Heider (1958) originally identified five levels in
which attributions of responsibility to a person vary as the
relative contributions of person and environment change.
These levels represent a progression from relatively primi
tive to relatively sophisticated cognitive processes. The
levels have been restated and labeled by Shaw and Sulzer
(1964) and by Sulzer (1971) as follows:
Level 1 Association: The person is held respon
sible for any outcome that he is connected
with in any way.
Level. 2. Causality: The person is held responsible
for any effect that he produced by his
actions, even though he could not have fore
seen the consequences of his actions.
Level 3 Foreseeability: The person is held respon
sible for any foreseeable effect that he
produced by his actions even though the
effect was not a part of his goals or in
tentions .
Level 4 Intentlonality: The person is held respon
sible for any effect that he produced by
7
his actions, foreseeing the outcome and
•intending to produce the effects.
Level 5 Justifiability: The person is held only
partly responsible for any effects that he
intentionally produced if the circumstances
were such that most persons would have felt
and acted as he did. Responsibility for
the act is shared by the coercive environ
ment .In studies of this levels model of attributions of re
sponsibility, subjects read a standard set of stories
created by Shaw and Sulzer (1964). For each story, the sub
jects are asked to assess the responsibility of the stimulus
person for causing negative or positive consequences. Levels
model researchers are concerned with the differences in
attributions of responsibility made by subjects as a func
tion of their "different developmental levels or cultural backgrounds 11 .
A different approach to attribution research is rep
resented by investigations in which the focus is on attri
butions which are motivationally altered or distorted.
Specifically, these studies stress the importance of a
particular motivation--perceiver self-protection.
Studies dealing with self-protective attributions
(variously) labeled ego-defensive, ego-protective, or ego-
biased attributions, can be grouped into three general areas:
studies of the locus of causality attributed for success
and failure (Bradley, 1978; Feather, 1969, 1971; Feather &
Simon, 1971; Sicoly & Ross, 1977; Streufert & Streufert,
1969; Weiner, Frieze, Kukla, Reed, Rest & Rosenbaum, 1971;
Weiner & Kula, 1970 ; Wolfson & Salancik, 1977 ; Lerner’s
research on the need to believe in a just world (Lerner &
Simmons, 1966; Lerner & Matthews, 1967; Lerner & Miller,
1978; and studies of responsibility assignment for accidental
occurrences (Chaiken & Darley, 1973; McKillip & Posovac,
1972; Schroeder & Linder, 1976, Shaver, 1970, 1971;
Sorrentino & Boutilier, 1974; Walster, 1966, 1967). It
is the last two areas of research which are of particular
interest in this paper. These studies have generated several
hypotheses regarding self-protective attributions which would
not have been predicted nor readily explained by the tradi
tional objective information processing models of attri
bution .*
Lerner (1966, 1970) hypothesized that people have a need
to believe that they live in a just world, one in which suf
fering people deserve what they get. According to Lerner,
their suffering has three possible causes: the victim’s ownovert behavior, his low intrinsic personal worth, or
"chance". The latter possibility is unacceptable to the
perceiver of misfortune, for were he to attribute misfortune
9
to the chance factors of a capricious world, this would imply
that such suffering was beyond anyone’s power to control.
He, too, could fall victim to such fate. !
The observer of such suffering prefers, rather, to convince himself that the victim was in some way behaviorally responsible for his own misfortune, or that the victim was characterologically responsible for his suffering (i.e.,Though innocent by deed, the victim deserves his fate by
virtue of his undesirable personal qualities.)
The general paradigm employed in just world studies has
observers evaluate the personal worth of a suffering victim
under conditions which make it clear that nothing about the
victim’s overt behavior is responsible for his plight. As
predicted, the results show derogation of the victim, indi
cating that perceivers believe the victim "deserves what he
gets". Assuming that the observer thinks of himself as a
good'person, the self-protective belief in justice removes the
threat posed by the negative occurrence.
Walster (1966) provided the prototype for a line of re
search regarding the motivationally distorted attributions
of responsibility for an accident with a study in which the
perpetrator of an accident, rather than the victim as in
Lerner and Simmons (1966), was the focus of attention. She
initially found that the more severe the consequences of an
accident, the more the perpetrator would be judged respon
sible. According to Walster, the perceiver of the accident
10
would be threatened were he to attribute responsibility for
the accident to chance factors. Such an attribution implies
that he, too, could be in a similar situation of causing\the negative consequences befalling the victim. In a self-
protective move, the perceiver judges the perpetrator as
responsible for the accident, while denying that he could be
in a similar situation, or that if he were, that he would
act differently than the perpetrator.
Whereas both Lerner's just world hypothesis and Walster's
hypothesis predict that a perceiver would]attribute respon
sibility to the victim or the ^perpetrator (with self-
protective denial of situational and/or personal similarity✓
to the victim or perpetrator), in a series of experiments
Shaver (1970) obtained and explicated results which were not *
in keeping with these predictions. Shaver found the height
ened probability of occurrence, especially in the form of
personal similarity to the potential perpetrator, lessened
the observer's attributions of responsibility and increased(
his ascriptions of carefulness to the perpetrator. Shaver
suggested that his finding may reflect "defensive attribu
tion" (e.g., I may cause a similar accident in the future.
If I’ do, I will want people to think chance was responsible
and not I.) The subjects exhibiting defensive attributions
were, according to Shaver, concerned with avoiding not only
potential threats to their self esteei and/or physical
safety (as suggested by Walster and Lerner) but also poten-
11
tial blameworthiness. The lessened attributions of respon
sibility to the perpetrator reported by Shaver indicated
that a person would prefer to belieye in a capricious world
rgther than belieye that he himself could potentially be
responsible-blameworthy for negatiye consequences to others.
Sh.ayer (1970, 19731 formulated a defensive attribution model
Which, encompassed the various self-protective attribution hypotheses and suggested the probable parameters of the
ph-enpmouon.
Shayerfs defensive attribution model stresses two dimen
sion's, which, are factors in the occurrence of self-protective
attributions of responsibility; situational possibility and
personal similarity. Situational possibility refers to the
ohseryer!s perceiyed probability of finding himself in
similar circumstances as the stimulus person (victim or
perpetrator) of an accident. As originally identified by
Walster .(19-56) , a prerequisite for the occurrence of self-
pro tec tiye attributions of responsibility is the possibility
tb&t the accident could befall the perceiver. As an ex-
a^mple, an account of a window washer’s fall might be of
little import to a store clerk, but would probably be quite
threatening to another window washer. The higher the pro
bability of the perceiver's confronting a similar situation,
the more threatening will be his use of defensive attribu-
tiqns. of responsibility.
The direction of self-protective attributions of respon
sibility (higher assignment of responsibility to the stimulus
person a la Walster or Lerner, or lessened assignment of res
ponsibility to the stimulus person a la Shaver) is influenced
by the second dimension— personal similarity. Personal
similarity refers to a concept of congruence of experiences,
attitudes, beliefs, and personality traits. As an example,
a female college junior might feel little similarity to a
male high school senior, somewhat more similarity to a female
college freshman, and perhaps much similarity to a female
college senior. Given situational possibility, the less
similar the observer of an accident feels to the stimulus
person, the more heavily he will attribute responsibility
for the accident to the stimulus person. Simultaneously,
the observer will assert his dissimilarity and might also
derogate the suffering victim (Lerner & Simmons, 1966).
Thus, the ob.server has avoided the threat of a capricious
world by attributing responsibility for the accident to the
stimulus person and, likewise, avoided the threat of this
potentially harsh standard of judgment to himself by
asserting his dissimilarity to the stimulus person (I
would not act in the same way. I fm not deserving of suffer-
ing as he was.)
Given high situational possibility, the more similarity
the observer of an accident perceives between himself and
the stimulus person, the less he will assign responsibility
13
for an accident to the stimulus person and the less he will
derogate the stimulus person. The observer cannot assume
that he would act differently or be less deserving of suf
fering than the stimulus person since he perceives himself
as being similar to the stimulus person. He establishes,
therefore, a lenient standard of judgment (i.e., lessened
attributions of responsibility) by which he would want to
be evaluated if and when he is also involved in a similar
misfortune. Shaver’s defensive attribution formulation has
been supported by evidence from a number of researchers
(Chaikin & Darley, 1973; McKillip & Posovac, 1972; Schroeder
& Linder, 1976; Shaver, 1970, 1971; Sorrentino & Bouti-
lier, 1974).
Despite the empirical evidence and intuitive appeal of
a defensive attribution phenomenon, there remains a hesi
tancy to accept the notion of motivation having an effect
on attributions. A major critique of the motivated attri
bution of responsibility literature comes from Fishbein
and Ajzen (1973). They portray the research as suffering
from two principal' deficiencies--(1) the lack of a guiding
theoretical schema and (2) the contextual ambiguity of
the accident situations typically presented to the sub
jects. Emphasizing the success of Shaw and his associates
(Shaw and Sulzer, 1964; Sulzer, 1971) in cross cultural
and developmental investigations of attributions of respon
sibility using Heider’s levels model of responsibility, they
1.4
suggested that researchers should likewise utilize the
Heiderian framework. Fishbein and Ajzen proposed that assign
ments of responsibility, rather than being contradictory and
uninterpretable, would follow predictable patterns when the
contextual level of the stimulus person's behavior in the
accident is clearly specified and when the subjects are
told at what level of responsibility they are to respond to
the question "Is the stimulus person responsible for the
accident?"In a rebuttal to their criticisms, Shaver (1973) pre
sented his model of defensive attribution mentioned above
and suggested that it be used as a theoretical guideline for
further research in the area. He also pointed out Fishbein
and Ajzen's lack of appreciation for the content and goals
of motivated distortion of attributions of responsibility re
search. Motivated distortion researchers are not interested
in the ability of subjects to make rational judgments in the
face of undeniable evidence as is the case of Shaw and
associates. Rather, they attempt to determine what factors
must be present to' motivate an observer to distort his per
ception of the causal chain of events leading to an accident
in order to avoid feelings of vulnerability or empathic cul
pability. Walster (1966) originally noted that such pro
tective tendencies will be strongest when the "objective
evidence as to the responsibility of the stimulus person is
ambiguous" (p. 74). Shaver described the ambiguous context
15
of motivated distortion studies (which Fishbein and Ajzen
condemned as a methodological deficiency but which Walster
claimed as important to eliciting the phenomenon) as falling
between Heider1s levels of Causality and Foreseeability. The
stimulus person is readily seen as having caused the accident,
but whether or not he foresaw or intended the results of
his action is unknown. This ambiguity allows the perceiver
to exaggerate according to his needs the accountability and
culpability of the stimulus person for causing the accident.
Rather than a fault of the method, contextual ambiguity is
a methodological necessity in this area of research.
Schroeder and Linder (1976) empirically verified and
elucidated the role of contextual ambiguity in defensive
attribution research. They proposed that an initial attri
bution of causality to the stimulus person is essential to
the eliciting of defensive attributions. How is it that
Walster considered contextual ambiguity helpful in elicit
ing defensive attributions, while Shaver maintained that
the stimulus person is readily seen as having caused the
accident, and Schroeder and Linder contended that the
initial attribution of causality to the stimulus person is
absolutely essential for the occurrence of defensive attri
butions? Schroeder and Linder pointed out that several
attributions are being made in assigning responsibility
for an accident. As noted by Heider (1958), a hierarchy of
attribution processes may exist. An observer's naive
16
analysis of the sequence of events yields only the informa
tion that some entity (person, environment, or chance) has
caused some effect. Additional meaning may then be attached
to the causal entity by the observer; attributions concerning
the entity’s abilities, desires, intentions, and respon
sibility for the effect. Assignments of responsibility,
then, may be considered second order attributions which
follow an observer’s first order to determination of the
most probable cause of an accident. It is during this second
ary processing that self-protective tendencies may be mani
fested. If initial causality can be objectively assigned
to entities in the event other than the stimulus person
(e.g., chance or environmental factors), no threat to the
observer’s self-image or potential blameworthiness should
be aroused, and attributions of responsibility should not
be defensively distorted. Only when the stimulus person is
perceived as the probable cause of the accident, and the
observer feels similar to the st imulus person and feels
that he could be in a similar situation, will distorted
attributions of responsibility be aroused. Defensive assign-iments may, then, be conceptualized as a two-stage process.
The observer must first perceive the stimulus person as a
primary causal agent of events for which the observer would
not want to be blamed. Then, motivated by a need to avoid
feelings of potential blameworthiness or vulnerability, the
observer assigns a lesser degree of responsibility to the
17
stimulus person.
Germane to the issue of a two step attribution process
is the distinction made by Shaver (1975) among the various
meanings of the term "responsible" -- causality, legal
accountability, or moral accountability. As causality, respon
sibility refers simply to the production of effects. As legal
accountability, responsibility refers to not only the effects
which a person produces but also to effects which he does not
directly produce (e.g., legal responsibility of parents for the actions of their children). As moral accountability,
responsibility is a value judgment made by a perceiver of an
event, and as such, is not always totally rational and ob
jective. Heider’s hierarchy of attributions and Schroeder
and Linder’s* first and second order attributions can be viewed
as employing these various interpretations of the term respon
sibility. First order attributions correspond to Shaver’s
"causality" meaning of responsibility; second order attribu
tions to his moral accountability meaning. In this regard,
the suggestion of Fishbein and Ajzen to utilize Heider’s
levels model of responsibility is well taken. If an ob
server is asked to assign responsibility for an accident,
would he respond with a first order responsibility attribu
tion (causality) or with a second order attribution (moral
accountability)? By using a form of the levels model, the
interpretation of the term responsible would be controlled
by the investigator. The responses of the observers may
18S
no longer be ambiguous (contradictory and uninterpretable)
even though the context of the second order attribution re
mained so.
Other prevailing methodological problems in distorted
attribution research have been discussed by Chaiken and
Darley (1973) and Vidmar and Crinklaw (1974). Chaikin and
Darley pointed out that the accident descriptions which were
used tended to confuse the roles of perpetrator and vietim
of the accident. In the auto accident story used by Walster
(1966) and Shaver (1970), for example, the teenaged boy
perpetrated the accident by parking his car on a steep hill
and was also a vietim of the resulting accident in that his
car was damaged. In addition, reading descriptions of acci
dents might not be as threatening an event as actually view
ing an accident via videotape, for example. Finally, ob
server identification with victim or perpetrator of an
accident has generally not been manipulated as a function
of situational possibility. Related to this issue is the
fact that the consequences of the accidents in most defen
sive attribution studies have included personal injury -and/Ior property damage. The probability of such an occurrence
is rather low and too readily denied by observers as being
applicable to them to serve as a viable situational possi
bility manipulation.
The present study included procedures to avoid poten
tial problems as noted by Chaikin and Darley and to consider
the suggestion of Fishbein and Ajzen. First, the victim
of the accident was separated from its perpetrator. Second,
a videotape presentation of the accident was used rather
than a non-involving description of the event. Third, not
only did the subjects actually witness the accident, they
also were led to believe that they would be placed in a
situation in which a similar accident could be caused by
them or could happen to them. There was no subjective
appraisal on the part of the perceiver as to the possi
bility of his being in a similar situation. Fourth, the
accident consequences were not improbable physical injury
or property damage, but the very possible loss of monetary
reward for the task the subjects were to participate in.
Fifth, the subjects were led to identify with either the
victim or the perpetrator by manipulation of both situa
tional possibility and personal similarity. Sixth, as
mentioned above, Shaver distinguished among several meanings
of responsibility and concluded that attribution research
ers are referring to moral accountability when they ask if
the stimulus person is responsible for an accident. Shaver
questioned, however, whether or not subjects responded to
the same interpretation. Using a Heiderian levels model
of responsibility, as suggested by Fishbein and Ajzen, sub
jects responded to dependent measures tapping the various
degrees and interpretations of the term responsibility.
Finally, Shaver (1971, 1973, 1975) referred to personal
ity variables as probably being influential in a person's
assignments of responsibility. In his 1971 study, he suggest
ed that the results may have been confounded by the 'inordinate
concentration in one of the treatment cells of subjects
high on Harmavoidance as measured by the Personality Re-
search Form (Jackson, 1968). Jackson described a person
who scored high on the Harmavoidance Scale as self-protecting,
pain avoiding, attending to danger, withdrawing from danger,
and seeking safety. Harmavoidance seemed to be a trait
which intuitively and empirically (though post hoc, Shaver,
1971) could affect responsibility assignments. Research
involving personality types as defined by Rotter's Internal-
External Locus of Control Scale (Rotter, 1966) has provided
empirical support for the suspicion that personality
variables could affect attributions of responsibility
(Phares, Wilson, & Klyver, 1971; Sosis, 1974). In the
present study, cell populations were matched on Jackson's
Harmavoidance trait and the intuition that this trait could
affect attributions of responsibility was tested in an addi
tional analysis of the data.
From Shaver's defensive attribution model, it was pre
dicted that subjects who felt personally similar to the
perpetrator and who expected to play a role in which the
previous role occupant perpetrated an accident would attri
bute responsibility for the accident less to the perpetrator
21
and more to the situational circumstances or chance in
order to defend themselves from potential blame in the
future. Specifically, those subjects whose fate and person
ality characteristics are similar to the perpetrator would,
a la Schroeder and Linder (1976), attribute causality to the
perpetrator as would the other subjects. They would differ
from the other subjects in that they would then attribute
less Foreseeability and Intentionality to the perpetrator.
How the subjects respond to the multi-interpretable term
Responsibility cannot be predicted for the meaning of the
word responsible is not being directly controlled by the experimenter. One may be able, however, to determine the
implied interpretation of Responsibility by comparing sub
ject responses to the various levels of responsibility,
e*g., Causality, Foreseeability, and Intentionality, with their response to Responsibility. These same fate and per
sonality similar subjects would also be more inclined to
believe in a "capricious11 world rather than one in which
they would be judged "responsible" for negative consequences.
Therefore, they would tend more than other subjects to
blame chance factors or situational circumstances rather
than the perpetrator of the accident. Concerning the vic
tim of the accident, given the just world hypothesis, the
victim of an accident would either be blamed for the acci
dent and/or disliked in order to justify the harm done to
him. However, this effect might not be found among persons
22
who felt similar to the victim and who expected to play a
role in which they might become a victim. That is, by de
valuing the victim in order to maintain a belief in a just
world, they are potentially devaluing themselves. It is
predicted that those subjects whose fate and personality
characteris tics are s imilar would no t devalue the inno cent
victim. Specifically, in response to inquiries into the
Likeability, Blameworthiness and Responsibility of the
Victim, defensively aroused subjects would not perceive the
victum less favorably than would other subjects.
METHOD
Sub j ect s . Ninety male Introductory Psycho logy students
at the College of William and Mary volunteered for a paid
experiment ostensibly investigating the correlations between
personality variables and spatial imagery ability. Eighty-
five subjects were actually contacted and used. One of them
produced unscorable responses, two failed to understand the
instructions, one admitted prior knowledge of the experi
ment, and one was discarded due to experimenter error,
leaving a total of eighty subjects.
Design Overview. A 2 x 2 x 2 factorial design with two
levels of situational possibility (High; Low), two levels of
stimulus person identification (Perpetrator; Victim), and
two levels of personal similarity (Similar-Dissimilar) was
employed. The txiro levels of expected task provide a situa
tional possibility manipulation: the Actors represent a
High situational possibility (High) Group who believe they
will be involved in the same task they are viewing, and
the Observers represent a Low situational possibility
Group (Low) who have no reason to believe that they will ever
be involved in a similar task. The two levels of stimulus
person identification establish a focus of attention on
either the Perpetrator of the accident or the Victim. These
2 3
24
two independent variables determine the situational possi
bility of the accident for the subjects and their potential
role in it. Ten subjects were randomly assigned to each
of the eight conditions.
Stimulus Material. A ten minute videotape purported
to be of two colleagues who also volunteered for the alleged
spatial imagery experiment was prepared. The film was os
tensibly to serve as a demonstration training film of the
experimental task for which the subjects had volunteered.
In reality, the tape served as a vehicle to present the acci
dent to the subjects and contained two students who acted
under the direction of the experimenter. The videotape
presented the following scenario:
The subjects in the videotape were given instructions by
the experimenter for performing the spatial imagery task.
They were sitting at a small table with a screen between them.
One subject, called the Reader (Perpetrator), had a stopwatch,
pencil, tally sheet, and description cards in front of him.
The other subject, called the Builder (Victim), had twenty-
seven WAIS Block Design blocks (Each block was a one inch
cube with two sides red, two sides white, and two sides
divided diagonally, half red-half white.). The perpetrator
was instructed to read the description of block designs from
his cards one at a time, proceeding when the Victim said to.
The Perpetrator was to time the trials, allowing only three
minutes per design. Following completion of the third time
25
trial, the Perpetrator was to go to the Victimfs side of
the screen and score the performance of the Victim. The
Perpetrator was to make note of each properly placed block
in the Victim’s designs, comparing them to the pictures of
the designs on his description cards. The experimenter
would pay the Victim 25£ for each properly placed block in
his structures, thus, the Victim could make as little as
$0.00 or as much as $6.75, depending on his performance as
tallied by the Perpetrator. The Perpetrator was told that
he would be paid $3.00 for his participation.
The video Victim was told to listen to the descriptions
of the structures and attempt to build them using the blocks
provided him. His reward contingency was reiterated --
$0.00 to $6.75, depending on the number of correctly placed
blocks recorded by the Perpetrator. The Victim was en
couraged to do his best and told that he would have only one
chance to perform the task.
Their instructions completed, the subjects in the video
tape proceeded to perform the task. It seemed moderately
difficult, but the Victim did quite well (twenty-four of
twenty-seven.blocks correctly placed). The time expired and
the Perpetrator got up to tally the Victim’s performance.
In getting up from the cramped experimental set-up, the
Perpetrator "accidentally** bumped the table and knocked over
the block structures of the Victim. The tape ended.
26
Procedure. Introductory Psychology students were asked
to participate in a two part study purportedly of the cor
relations between personality variables and spatial imagery
ability. Part 1 would consist of their being administered
the Personality Research Form (Jacksons 1968), completing
this personality inventory, and returning it to the exper
imenter. In Part II, they would be contacted by telephone
and appointments made for them to perform the task ostensibly
being studied. They were told that their total time commit
ment would be approximately one hour, for which they would
be paid.
Ninety Personality Research Forms were returned completed.
The subscale Harmavoidance was scored for each subject and
a stratified random sampling procedure with Harmavoidance %
level .as criterion was used to assign subjects across condi
tions. (A frequency distribution of Harmavoidance scores was
developed and divided into centile groups. This distribu
tion was nearly identical to normative data compiled by
Jackson, 1968.) The subjects were contacted by telephone
and appointments made for them to perform Part II of the
experiment. The contacts went as follows:
I ’m Glenn Paule-Carres, the graduate student for(
whom you completed the Personality Research Form.
I am now conducting Part II of the study and want
to know if you could make a % hour session tomorrow
at x o ’clock in Room 125? I’d like to stress the
27
importance of your keeping this appointment, since
I am coordinating four schedules— yours, your
experimental partner’s, my assistant's, and mine.
In other words, three of us will be expecting you
at x o'clock. If for some reason you cannot make
it, call me and we will try to arrange a new time.
We'll see you tomorrow at x o'clock, Room 125.
Eight appointments were made for each experimental session.
The eight subjects were of the same Harmavoidance level
(centile group) and were randomly assigned to the eight
experimental conditions, with the restriction that each
condition contained only one subject from each Harmavoid
ance level. Of a total of 85 subjects contacted, only two
subjects failed to meet their initial appointment. They
were assigned new times which they did not miss.
Situational Possibility Manipulation. Upon arrival,
the subject was met by the experimenter's assistant. The
assist ant read an introductory atatement whi ch established
two High-Low and Perpetrator-Victim of the three independent
variable manipulations being performed. Together, these
two manipulations comprised the possibility of the subject
being in a similar situation and role as that of the stimulus
person in the videotape. Two of the eight subjects who ar
rived during an experimental session were randomly assigned
to each of four possible High-Low - Perpetrator-Victim
conditions. The assistant read one of four introductory
28
statements to each subject, it being determined by the sub
ject’s condition assignment. The statements were as follows:
High Victim
You are here today to participate in a study of
the relationships between personality traits and
spatial imagery ability. You will be performing
a two-man task with another intro psych student.
He has already finished this training session, so
he will be joining us later. To acquaint you with
the task, we have prepared a short demonstration
film which explains and illustrates the essential
aspects of the task. The subjects in the film
came ,from an intro psych class and are performing
* the same task, with slightly different problems,
of course, which you and your partner will be
doing, so pay attention. You have been randomly
assigned the builder duty in your task. This
name will make sense to you when you see the film.
Remember, you will be the builder in your task.
.In the High Perpetrator condition, the subject was read
the same statement except that the word builder was re
placed by the word reader. In the Low Perpetrator-Vietira
conditions, the subject was told that he would not be per
forming the task. He was merely to observe the film paying
29
particular attention to the builder (or reader).
Similarity Manipulation. Personal similarity was de
fined in terms of personality traits rather than attitudes.
This was done for several reasons. First, personality
traits are presumably more enduring, more basic a guiding
principle of behavior than attitudes. Personality similar
ity would imply a greater probability of like behavior, particularly under the stressful conditions of an accident,
and, therefore, would be more threatening a similarity mani
pulation. Second, studies (e.g., Reader and English, 1947;
Izard, 1960) showed that personality similarity is less
related to attraction than is attitudinal similarity.
Differential attraction to the stimulus person based on si
milarity would present a problem in this experiment, there
fore, it was minimized by employing personality trait simi
larity. Third, attitudinal similarity has been show to pro
duce increases in perceived morality (Byrne, 1961) and in
estimated competence (Jones, Bell and Aronson, 1973), both
of which might influence attributions of responsibility.
The communication of personality similarity or differ
ence was accomplished by using the Personality Research
Form which was administered earlier to the pool of Intro
ductory Psychology students and, then scored on the Harm
avoidance subscale. After receiving the situational possi
bility manipulation from the assistant, the subject was led
into the experiment room where the experimenter was waiting
30
with the video monitor and the spatial imagery experiment
apparatus. The experimental apparatus (table, screen,
block design cards, blocks, etc.) were located in a corner
of the small room. The subject was asked to take a seat
in front of the monitor. The assistant handed the experi
menter a card with the third independent variable manipu-
lation— Personal Similarity. The assistant was unaware of
this manipulation while he interacted with the subject,
and the experimenter was unaware of the Situational Possi
bility Role manipulation performed by the assistant in the
introductory statement. The experimenter read one of two
.possible statements which the assistant randomly assigned,
with the restriction that the factorial design was main
tained. The statements were as follows:
Similar
Take a seat in front of the TV. We are interested
in the effect of personality traits on spatial
imagery ability. Do you remember the Personality
Research Form? (holding up the subject's form) The
PRF is a standard device used to measure personality
traits. The results of these tests allow us to
compare subjects on personality traits and divide
them basically, into groups. Comparing your PRF
results with those of the subjects in the film,
you are similar to the fellows performing the task
in the film. Following this demonstration film,
31
you will answer some questions about the task
situation and the experiment in general. If you
are going to perform the task your partner will
be here by then. O.K.? Watch the film. It is
approximately nine minutes long and ends abruptly,
but don’t be concerned by that. Remember, you
are similar to the subjects in the film.
For the dissimilar condition, the statement was identical
except for the replacing of the word similar with dissimilar.
The experimenter turned on the videotape and left the room.Pretest. In a pretest, twenty subjects received only
the personal similarity manipulation and viewed the video
film up until the point before the accident. The efficacy %
of the similarity manipulation was established. Subjects
who were told that based on their Personality Research Form
profiles they were similar to the video subjects rated
themselves significantly more similar to the stimulus person
than those subjects who were told that they were dissimilar,
t^(18)=4.81, p <. 00 1, (}£=4 . 1 vs. X=5.9). In addition, neither
of the two video subjects was differentially liked (cf.
Regan, Straus, & Fazio, 1974) and the Victim’s performance
was perceived as being near maximum. Questioning the sub
jects showed the cover experiment on spatial imagery to
be both feasible and credible. An additional fifteen sub
jects viewed the tape through the accident and attested to
32
the apparent reality of the accident.
Dependent Variables. Following the videotape, the ex
perimenter joined the subject and led him to another room.
The experimenter presented the subject with a folder containing the questionnaires and the following instructions:
High
Enclosed are two short questionnaires. Fill them
out and then we will get you and your partner to
perform the task. The computer print-out form deals with the major issues of my study. I would
appreciate your quick and honest responses to its
items. The second form has questions dealing with
what happened at the very end of the tape. Budget
cuts prevented us from re-taping the session. The
tape is, nonetheless, a good training demonstration
of the task, so we kept it. We felt that what
happened at the very end did point out some issues
which might be worth investigating, so would you
please answer some questions about it too?
For subjects in the Low condition the "and then we will get
you 'and your partner to perform the task" phrase was deleted
from the instructions.The first questionnaire was a very formal, important
looking computer print-out containing statements about the
bogus e:xperimental task with Likert Scale responses. This
33
questionnaire served several purposes. First, its formal
ity and impressiveness lended credence to the assertion
that the study truly dealt with the bogus spatial imagery
experiment (Chaikin and Darley, 1973). This took the
emphasis off of the "unimportant" second questionnaire
which contained the relevant dependent variables. This,
hopefully, resulted in less defensive, more spontaneous
responses to the second questionnaire than would be expected
if the subjects believed it to be the major concern of the
study. Second, it acquainted the subject with Likert-type
scaling, so that he would have no problem in understanding the technique of responding when he got to the relevant
questionnaire. Third, the identifying information at the
beginning the first questionnaire made the subject ack
nowledge his awareness of the manipulation performed on
him and once more reminded him of the conditions before he
answered the questionnaires.
The second questionnaire contained the dependent vari
ables. It was less formal looking— typed and xeroxed. This
was in line with the assertion that it was of secondary im
portance, only an inquisitive addendum to the real study.
It contained twenty-five items with Likert-type response
continua of unbroken lines with labeled endpoints. These
items asked questions about the accident and the subj ect’s
impressions of the Perpetrator and the Victim, particularly
in regard to the accident.
34
When the subject finished the second questionnaire,
the experimenter returned to the room and fully debriefed,
paid and thanked the subject for participating in the
study.
RESULTS
The response continua of dependent measures were divided
by a scoring key into eight equal intervals. The data were
scored from one to nine, the higher the score, the more of
the variable in question. The data, scored, were then ana
lyzed in a 2 x 2 x 2 (expected task (High-Low) by stimulus
person identification (Perpetrator-Victim) by personal simi
larity (Similar-Dissimilar) ) analysis of variance which was
prepared by the Statistical Package for the Social Sciences
(Nie, Hull, Jenkins, Steinbrenner & Brent, 1975).Manipulation Checks. At the beginning of the first%
questionnaire each subject was asked to supply the following
identifying information:
Name:
Age:
Date:
Circle One: Future Builder Future Reader
Observe Builder Observe Reader
Circle One: Similar to subjects in film
Dissimilar to subjects in film
Eighty-two of eighty-five subjects correctly acknowledged
the manipulation-instructions which they received. Thus,
it was possible to assume that the subjects were aware of
35
36
and understood the manipulation-instructicns which they re
ceived. Because the subjects provided this acknowledgment
before they responded to the questionnaire items, they were
unable to deny (during debriefing, for instance) awareness
and/or understanding of the instructions in order to not
be held responsible for their responses to the dependent
measures .
In a check to see if the accident was perceived as
having negative consequences for the Victim, the subject
was asked "How much did the Builder make as a reward for
■his performance?" with $0.00 and $6.75 serving as endpoints
of the Likert-type scale. The mean response was 4.2. This
compared to a pretest mean of 7.25 for twenty subjects who
viewed the videotape until just before the accident occurred.
This difference was highly significant, £(98) = 4.26,
<.001, but the experimental mean of 4.2 (which equates to
approximately $3.50) seemed rather high reward for a perfor
mance that was apparently negated by the accident.
To insure that the similarity manipulations had been
perceived as intended ("Based on the personality test you
are similar, or dissimilar, to the students in the video
tape."), subjects responded to four items designed to be
manipulation checks. . Table 1 presents the mean responses
to these items. When asked "How similar are the reader's
(Perpetrator) personality traits to yours?", a Sim-Dis
main effect was obtained with the subjects in the Sim
Mean
Attr
ibut
ions
of
Simi
lari
ty
37
43bC•Hca
CO CO CO »M vO rM CO 00•H e • e • • • •
a p m vO «—I m ■ *—1 m•H4-*O•H a St- m O 00 ON ON> •H • • • • • • •
St rM vO rM St rM in
dO CO 1"* CO 1"- ON rM4-J •H • • • • • • •td P co rH St f-M St v—1 SIu4-J<uft a r>. CO CO rM CO t". ONM -H • • • • • • •0) C/3 m pH in CM m r—i StCM
(0 ON CO m CM rM r>. r-'.•H • • • • • • •
a St «—1 St i-M m o sf•H4-1O•H a m vO «-M St St vO O> •H • • • • a • *
C/3 m i-M vO rM m rM vO
ViO CQ t*. rH OO 00 o st co4J y ) • • • • • • •td P -a- CM St rM m CM StVi4-10)ft a CM 00 CM VO o st CMVi t4 • • • • • • •a C/3 vO i—1 vO rM vO rM vOCM
\X P ixi P \x PC/3 CO C/3CO
Vl rO O 4-1 fJ -H cd i—i Im td4j c3 a>CD O 4-Jft CO *HM Vi td0) G) ViCM PM H
504J•H50 i-H tdd toO 4-1 CO *H >4 td •H <U M
> Pm H
a4-1U
(0VIo4J «-fl CO td o) J-i 03 'd 414M d Q) CD 4J d r l *rl V H P <13 <13 4-1PM PQ <3
r̂ .
co
m
mCM
CO
PCO
COCO 03 - co *Td a mm d•H 0) 4J 4-1 *H *H O rH 4-1 •rl 03 4J > pq <3
condition rating themselves as more similar, F.(l, 72) =
4.21, £ <.04. Likewise, a Sim-Dis main effect in the intend
ed direction was obtained on the item "How similar are the
builder's (Victim) personality traits to yours?", F.(l, 72) =
4.42, g<.037, with subjects in the Sim conditions rating
themselves as more similar. Given these main effects, it
appears that the similarity manipulation was effective.
In addition, interaction effects were obtained on both
items. For "reader's personality traits similar", a Per-
petrator-Victim by Sim-Dis interaction was highly signi
ficant, F(l, 72) = 7 . 49 , £<.008, indicating that the Sim-
Dis manipulation was heightened among subjects attending
to the Perpetrator than among those focusing on the Victim.
For "builder's personality traits similar", a High-Low
by Petpetrator-Victim interaction was almost significant,
£(1 , 72) - 3 . 75 , £<.054, with similarity to the Victim
being better, reflected among LSP-Victim subjects. It was
as though the low situational possibility of their being
in circumstances as the Victim allowed them to claim a
higher personal similarity to the Victim than others who
expected to be performing the same task as the Victim.
In response to the question "How similar are the
reader's (Perpetrator) beliefs, attitudes, and values to
yours?", no differences were obtained. To the question "How
similar are the builder's (Victim) beliefs, attitudes, and
values to yours?", a significant Sim-Dis main effect was
39
obtained, F(l, 72) = 7.49, £<.008, with those in the Sim
conditions claiming more similarity. A Perpetrator-Victim
main effect was also obtained which approached significance,
F,(l, 72) = 3.75, £<.054, with those subjects focusing on
the Victim claiming more similarity.
Responsibility of_ Perpetrator.
Causality. Tables 2 and 3 present the mean ratings of
the dependent measures which tested the first prediction—
High Situational Possibility-Similar-Perpetrators would
attribute less responsibility for the accident to the Per
petrator. Included in the tables are the variables
Causality, Responsibility, Foreseeability, and Intentional-
ity. When asked if the Perpetrator caused what happened at
the end of the tape, no differences were obtained and the
high mean responses (7.6 - 8.4) indicated that the Per
petrator was clearly seen as the probable cause of the
accident.
This finding constituted the first order attribution
of responsibility which Schroeder and Linder (1976)
claimed to be essential to the eliciting of defensive
attributions. The Perpetrator was clearly identified as
the causal agent, regardless of the experimental conditions.
Any perceived similarity to him and his situation could be
threatening and could result in defensive attributions of
lessened responsibility. There would be no threat, how
ever, for HSP-Sim-Perpetrators had the Perpetrator not
AO
93 rH m r-» on•H • • • •
0 F5 oo rH rH•H4Ja
•H e f—1 r>. inP > •H #• • • ••H CO CM vO CMrH•H £rQ o•H03 Pa o CO <r Ht vOO 4J •H » • • «P cd Q 00 o r>* CMCO P<U 4JF*S a>
p. o v£> r—I rHP p •H • • • •O a) CO 00 CMP Picd
css PP0) 0)iH PP
cd <0 CO r*̂ CS1 CMH FH *H • • • •s <=> r-̂ ?—i vO eo
UH •HO P
« O\a -H c o 'd- 1^ oFS > *H • • • •o <o r-. F—1 f". rH•HP M3 •HrQ W•H PP O CO vO r̂. CM OsP P • • • •4-1 cd O r-. rH vO CM
pp
fd a)cd p e os m in ma) p •H « • • 0S a)
FMCO f-» rH vO CM
I X Q N QCO CO
P p a)o o rHp p * ocd cd *Hp p COp T3 p<u O a op CO p pp p p COCJ cd <u a)pL, o P p
42
been perceived as having caused the accident.
Responsibility. The other dependent measure presented
in Table 2 is "How responsible is the Perpetrator for what
happened at the end of the tape?". No significant differ
ences were obtained and the mean of 7.0 was similar to a
mean of 7.9 on the Causality variable. It appears as
though responsibility equals causality to these subjects
and, in fact, the two dependent variables are significant
ly positively correlated, r(78) = .32 , £<.01.
Foreseeability and Intentionality. Table 3 contains the other two dependent measures being used to determine
the assignment of responsibility for the accident. When
asked "Do you think the reader (Perpetrator) foresaw what
happened at the end of the tape?", a High-Low by Perpetra-
tor-Victim interaction was obtained, JF( 1, 72) = 4. 33,
£<..039, with the future potential Perpetrators attributing
less foreseeability to the Perpetrator than the other ex
perimental groups. It was predicted that the Sim-High-
Perpetrators would respond in this defensive fashion, but
not predicted for the Dis-High-Perpetrators. Theoretical-[
ly, they could have relied upon their dissimilarity to
the Perpetrator as a threat reducing strategy and proceed
ed to assign foreseeability more in keeping with the other
experimental conditions. Apparently, the situational
possibility was strongly perceived and threat enough to
future potential Perpetrators.
43
CO0)
H
cdH
u•HrH•HJO•HCOdOCM(0(UP$O4Jcd>-<4J<uCMH
a•H4-1 O r-MO • •
>m CO
uo4-1cdH4-10)Ma)Pm
mco
co
CO oco
com
COCO
a0)a4-1cd0)u4-»Ma>•d4-1oaoMCM4Jda)n0)mIM
<u a •HPm •H X)4-1 I"* CO CM O'!M-l U • • • • tnO •H co m CM rH
> 4-1CO dd .d cdo bo o•H •H M •H4-1 td Od 4-J •Hjo cd d•H d •K 4e toM 4-1 rH CO o m4-J 0) • • • • CO4J PM CM H CM fl-l< P4 dCD cdd Pm 0)cd aV
(X « lx PS 4CCO CO
P(O4-1cdu &4-J cda) co pm a)M Ma) o Pm Pm
Mo4-1cd x) M a)4-) noa) dCM <0 U 4-><u dPm M
£.<•01,
Newm
an-K
euls
st
atis
tic
(Win
er,
1971
).
44
When asked "Do you think that the reader (Perpetrator) intended to
push the table?", a High-Low by Perpetrator-Victim interaction was ob
tained, _F(1* 72) = 8.67, jd<!.004, with the future potential perpetrators
attributing less intent to the perpetrator than the other groups. As
with the Foreseeability measure, the lesser attributions of intent from
the Dis-High-Perpetrators was not expected. It might be noted that the
Dis-High-PerpetratorsT rating of personality similarity to the Perpetra
tor of 4.7 might not have been dissimilar enough for them to not have
been threatened by their perceived high possibility of being in a sim
ilar circumstance as the perpetrator.How is it that the High—Perpetrators appear to be using defensively
lessened attributions in the Foreseeability and the Intentionality
measures but not in the Responsibility measure? This discrepancy might
be attributed to the ambiguity of the term responsible. As Shaver (1975)
questioned, what is meant by responsibility? causality or moral account
ability? As noted above, the dependent measures of Perpetrator Causality
and Responsibility are correlated. Similarly, responses to Perpetrator
Foreseeability and Intentionality are highly correlated, _r (78) = .77,
jD'C.OOl. Neither Foreseeability nor Intentionality were correlated with
Causality, :r (78) = . 169, jg_>.l and r_ (78) = .137, jp>. 1 respectively;
but each tended to be correlated with Responsibility, (78) = .232,
£<^.05 and r_ (78) = .,192, jd<.1 respectively. It appears as though the
High-Perpetrators did not hold the perpetrator morally accountable for
the accident as evidenced by their lessened attributions of foresee
ability and intentionality to the Perpetrator. On the other hand, they
perceived him as definitely having caused the accident and their rather
high attributions of responsibility suggest that they tended to inter-
4 5
pret responsibility to mean causality moreso than moral
accountability.
Chance Factors. Table 4 presents the mean ratings of
subjects to the questions "How responsible was pure chance
for the accident?" and "How responsible was the experimental
set-up (table, blocks) for what happened?". The overall
analysis of variance obtained no significant results. Comparing the effects of the High-Sim-Perpetrators to the
average effect of the other treatments, however, yielded a tendency for the High-Sim-Perpetrators to make higher
attributions to "pure chance" and "experimental set-up",
F(l, 72) = 3 . 59 , £<.10; F(l, 72) = 3.70 , p<.10 respect
ively (Winer, 1971).Perogat ion o f Vict im. Table 5 presents the measures
which were relevant to the second prediction being tested,
namely, that High-Sim-Victims would not endorse the just
world belief that the Victim was deserving of and/or
characterologically responsible for his misfortune. When
asked "How likeable a person is the builder (Victim)?"
no significant differences were found and the overall mean
of 5.9 is identical to the overall rating of the likeable
ness of the Perpetrator.
The item "How blameworthy is the builder (Victim) for
what happened at the end of the tape?" provided a Perpe-
trator-Victim main effect, 2?(ls 72) = 8.02 , £<.006 , with
09 0 0 «"» C O• H • • • 9
e O C M < r ' C M• H4-JO• H 6 i H o C M
> • H © • • •
C O C O C M C M
+ j O► 4
i H HO (0 *cr t— 1 h *
. Q 4 J • H • • • •
• H cO Q m C O <T C M03 u
C * JO to< X a C M C M mCO M * H • • • 9
<D <U C O m C O C O H
P $ P 4
o
Q>Hr Q«H
O 03 rH CM CM•H •H • • • •4J 0 m CO CO CMd •H
rC i ■U*H OU •H oo Mt 00V > •H • • * ••M CO Sj- CM r-H<5 ©d
ocd •HCOa)S3 O 03 c- 00 r-> CM
4-J • • • •CO Q m CM CO CMH4JOa e so rH <1- COa •H • • 9a) CO so CM m CMP h
I X <=» I X «CO CO
r~ia) CO <DiH 4-J H43 d rO•H a) . *H03 0 00
O d *H a dO o U d o
<D 0 (X <D i aU cO CO a 4-J CO0 £ i a) X 0) 0)
(U CJ w co X
Tabl
e
47
(0 oo Mf 00 CM CO o\• • • « » •6 O VO rH CM CO CM H■HP0 a•H ■H 0 CM VO M* CO vOP > •H . • • • • •
O CO vO H CM rH CM rHtH> &oU-l t-Jo PO CO vO CO Ml r̂ CM CMd P •H • • • • • •o cd O m Ip rH o CM CM•H PP PCd <D60 (X 0 r—1 CM CM vO OO mO P «H • • • • • •P 0) CO m CM H o rH rH4) Ph(=)OPPd CO vO CM m CM maj •H ■ • • o • •> 0 O »n H CM rH CM rH<0 •HH P<0 OP4 •H 6 On m cn t— VO 00> •H • • • • • •CO CO m rH CM <N CM rHd rdo 60•H •HPd P.a O co VO o Ml- VO CO vO*ri P •H • • • • • •P cd « m CM CM rH CM rHP pP p< 0)<x a r— 1 rH r̂. rH Mt md p •H • • • • « •cd a) CO VO CM rH rH rH oa) Ph
IX Q IX o }X «CO CO CO>v 0)rd > tHp .O0) p •HtH o CO
0 ̂ 0 & 0 d•H cd •H 0) •H oP 0) P 0 P PuO ̂ O cd O CO•H •H •H tH •H <u> ►J > PQ > Pd
48
the subjects attending to the Victim attributing more blame
to him. Specifically, the Sim-High-Victim group attributed
the most blame. It appears as though the belief in a just
world is more potent than defensively lessened attributions.
When asked "How responsible was the builder (Victim)
for what happened at the end of the tape?", no differences
were obtained.
Additional Analyses— Harmavoidance. A second analysis
of variance was performed on the data using Harmavoidance
1evel as an additional condition. It was easy to divide the
members of a treatment cell into Above Median or Below
Median Harmavoidance groups because each cell was represent
ative of the subject population distribution of Harmavoid
ance scores. The resulting analysis was a 2 x 2 x 2 x 2 ,
High-Low by Sim-Dis by Victim-Perpetrator by Above Median-
Below Median. This analysis provided significant results
on four dependent variables: Personal Similarity to Victim;
Perpetrator Likeability; Effect of Accident: and Victim
Blame. Tables 6, 7, and 8 present these results.
The variable "How similar are the Victim’s personality
traits to yours?" obtained an High-Low-LSP by Sim-Dis by
Above Med-Below Med effect, F(l, 64) = 6.43, £_-^.01. The
similarity manipulation made the most difference either
among Below Med-High or among Above Med-Low subjects.
The combination of the two supposed threat evoking condi
tions in the Above Med-High seems to overwhelm the system
49
00V» 4Ji-C *r-f»h tdtd Hr-i Hff0•H 4-1CO *HrH
mh tdo C3vO OCO CO
0) VI*H o a)*Q •H Ptd 4-1H 0 CO
rO ■»•H 0M •ri4-1 4-14-J O< •H
>ntd O0) 4-1S
0)OC3 CO 00 Ocd •H 0 •■o P to —I«H £O o> 1-3td 0 CM O0 iH • •Vi CO to COtdM
atd•H CO rH rH*a •H • •a> P cmS rGto£ •Ho 0 <f o\•H 0 0
tu CO VO rHW
a)o CO VO COPi • •td P <S“ CO
TO S•H oO i-3> 0 h r^td •H • •0 CO vO •—(ntdW
Picd CO CM CO
•H •H • •*d P to oCD rE3s to
•H0) W 0 CJ\ oo> •H • •o CO to rH
rO<
\x pco
50
4J•HrH
4-4 *HO
cd4-J 0)o Ma) •H
m iJCMW Mo
1". CM 4-Jo cd<u VlrH 00 4-1
a 0)cd o CUH iH u
a)P-4*9V* <Odu J3•M cd4-1c V>
ac (Ucda> •H£ o
O<3
a)oacd* d•HO>cde CO v O OVi •H • •cd Q m c om
acd e l"» o•H •H • •i d CO 1"* C Ma>££oi— i0)PQ
r-m
v O
in
c m
<DOacd
' O
o>cd6Ma)Watd•H
0.1£0)>o,o<
CO
e•HCO
<N <* CM* * • •vO CO vO rH
00 uo• • • •m co m H
|X Q IX Q5ft Cft
So ViM O«H 4-J
4-J H cdCJ •H U
4J a) X> 4-1O ■X3 cd O0) •H a) exCM <u MCM CM O iH M_j <Dw o <; hH O P-i
51
0)o 0 ts VO M t0 •H o • •cd 4-» CM CM•o O
•H •H03 o t>09 > »P CO CMa) cd OO • •
a 0 co CM•H M td»d cd4-1 tdno a& cd o S sr ©Va> •H 4J o •
0 •a cd *-l H OCO fl) M
a 4-JPQ a>
& p vO *—I
0 o M 0 0 • •*H tH 0 ) *H r—4 r —44J 0 ) P 4 tdU PQ
•H>*4-1O 0 )
O03 id 0 CO oo0 cd •H o • •
O •d 4-> t - i CO CM•H •H O4-1 o *H2 > >
cd ®ci CO CM*H 0 0 0 • •
U u *H r—4 v—l4J cd Ed4J Ed<u
0P cdcd •H O ts CM vO<u •d 4-> o • »
a 0) cd ►4 •-4 oS U
4-Jd) 0)> P . f i m CMo M 00 • •
,Q <U CM »—4
td
i>c «to
52
and negates the similarity manipulation.When asked "Did the accident affect the builder’s
(Victim) reward?", a Sim-Dis by Harmavoidance interaction was obtained, F(l, 64) = 5.81, p.<.02, with subjects who were either Sim-Above Median or Dis-Below Median minimizing the consequences of the accident, As in the previous measure, the presumably most threatening and the presumably least threatening conditions provide similar responses. The attributions of the least threatening group remain uninterpretable, unless they are construed to be an uninvolved group whose responses might differ as a function of their distance from the motivational issues confronting the other groups.
To the question "How likeable a person is the reader (Perpetrator), a Sim-Dis by Harmavoidance interaction was obtained, F(l, 64) = 5.37 , p < . 0 2 , with the Perpetrator being most negatively evaluated by subjects who are either Sim-Above Median or who are Dis-Below. Again, assuming that the Dis-Below group represents an uninvolved, "rational" condition, it is possible to see in the Sim-Above negative evaluation a denial of liking that can be interpreted as a response to the threat — a sort of denial of simi
larity .To the question "How blameworthy is the builder ( V i c
tim) for what happened at the end of the tape?", a HSP-LSP by Perpetrator-Victim by Harmavoidance interaction was ob-
53
tained, F(l, 64) = 4.82, £ < . 0 3 . When subjects believed that they would not likely be in a similar situation, those who identified with the Victim blamed him more than those who identified with the perpetrator. When subjects believed that they would be involved in a similar situation, both the nature of the future task and their personal dispositions affected the blame assigned to the Victim; the greatest blame was assigned by future perpetrators who were below the median. Once more the results of experimental conditions vary as a function of the personal disposition of Harmavoidance.
GENERAL DISCUSSION
The purpose of this study was to test hypotheses generated by Shaver's defensive attribution formulation as opposed to those of Lerner's just-world hypothesis. In doing so, it was expected that a better understanding would be obtained of the relevant variables in the motivated attribution of responsibility phenomenon. Furthermore, methodological Improvements were attempted in order to facilitate the interpretation and usefulness of research data.
In regard to the primary hypothesis that persons would attribute less responsibility to the perpetrator of an accident if they feel that they may be in a similar situation in the future and if they feel that they are similar to the perpetrator, Shaver’s defensive attribution model received partial support. The perceived personal similarity to the perpetrator did not affect the attributions of fate-similar subjects. Nevertheless, these subjects attributed less foreseeability and intentionality of the accident to the perpetrator and also tended to attribute more responsibility to chance and to the experimental setup. These latter attributions are not in accord with Lerner’s just-world prediction that the subjects would reject a belief in a "capricious world".
54
55
Of particular importance was the clarification of responses of responsibility attributions by means of several different measures which tapped different meanings of the term responsibility. This technique allowed the investigator rather than the subject to control the response interpretation.
Both Shaver’s and Lerner’s attribution models rely on the negative consequences of the event to evoke the particular manifestations of defensive attributions which were being studied in this experiment. As noted above, the negative consequences of the accident as measured by the dependent variables "Did what happened at the end of the tape affect the Builder’s reward?" and "How much did the Builder make as a reward for his performance?" were not similarly perceived. For instance, the Similar-Above Median Harmavoidance subjects claimed the accident to have had a lesser effect on the Builder’s reward than Similar-Below Median Harmavoidance subjects. Furthermore, as stated earlier, the mean rating of 4.2 (which equates to approximately $3.50) as the Builder’s reward seemed rather high for a performance that was apparently negated by the accident. An inspection of the responses to the dependent measure "How much did the Builder make as a reward for his performance?" showed a bimodal distribution of reward ratings as presented in Figure 1. Approximately 34% of the subjects estimated the Builder’s reward to be $0.00, while another 33% esti-
56
Figure 1.Bimodal Distribution of Builderfs Reward Ratings
Frequency of Score
302520151050
1 2 3 4 5 6 7 8 9Score
57
mated the Builder’s reward to be between $5.25 and $6.00.The remaining reward estimations ranged from $0.75 to $6.75. The Builder was apparently seen as receiving no reward or as being compensated to the pre-accident level of performance in spite of the accident. Comments such as "There was a videotape record of the Builder’s performance", from subjects during the debriefing might explain the ease with which the negative consequences of the accident could be denied. The weakening of as important a factor as the negative consequences of the witnessed event may present a dampening influence on the production of defensive attributions. One would suppose that the effect of defensive attributions would have been greater had the subjects been presented with a stimulus which was not ambiguous in regard to the negative outcome of the accident. In future, studies, one might include as subjects only those who through manipulation checks are identified as having perceived the experimental stimulus as was intended by the researcher. Otherwise, an undetermined and unexpected amount of "noise" may be included with the "signal" of results.
The second prediction being tested, namely, that the subjects who felt that they would be in a similar situation as the victim of the accident, and who also felt similar to the victim, would not derogate the victim was not supported. Rather, the High Situational Possibility Victims appeared to be responding in accordance with the just-world
58,
hypothesis by finding the Victim more blameworthy than was found by other subjects. However, in the additional analysis including Harmavoidance, those High Situational Possibility Victims who were also Above Median Harmavoiders declared the Victim less blameworthy. This supports Shaver1s defensive attribution prediction and further stresses the importance he places on the interaction of relevant variables in attribution research, specifically those variables of situational possibility, personal similarity, and personal dispositions.
59
Appendix A Computer Print-Out Questionnaire
(modified)
CORRELATIONS BETWEEN PERSONALITY TRAITS AND SPATIAL IMAGERY Fill in identifying information, then begin questionnaire. Name :A g e :Date :Circle One: Future Builder Future Reader
Observe Builder Observe Reader Circle One: Similar to subjects in the film
Dissimilar to subjects in the film Respond to each statement by placing a slash mark on the line where it corresponds to the position of your feelings concerning the statement.
The task discriminates subjects with good spatial imagery from those without.Disagree _______ /_____ /_________ £_______ /_ /________ / AgreeThe tape clearly portrays the duties of the subjects.
Agree _______ /______/________ /_______ /_______ / / Disagree
Personality traits correlate with a person’s spatial imagery. Agree _______ /_____ /_________ /_______ /_______ /_______ / Disagree
60
I understand what "transforming verbal descriptions into
3-D figures1’ means.
Disagree / I I I / / Agree
The builder’s task is difficult.Agree / / I I / / Disagree
The builder’s performance was poor.Disagree I I / / / / AgreeThe reader’s performance was poor.
Agree / / I I / / DisagreeI would be able to perform the builder ’s task
Agree / / / / / / DisagreeI would rather be the builder in the task.
Agree f i l l / / DisagreeThere is a x>isk for the builder in not having a guaran-
teed reward.Disagree I I I / / / AgreeThe builder’s monetary reward is adequate.Disagree / / / / / / Agree
A person’s experiences could effect his building ability.Agree _______ /_______/______ /______ /______ /_____ / Disagree
The PRF test is a valid measuring device for personalities. Disagree _______ / /______ /______ /_____ _/_____ / Agree
AI hereby give permission for my responses to be used
as data in this study.Signature __________________________
61
Appendix B Relevant, Second Questionnaire
PLACE A SLASH MARK ON THE LIN AT A POINT WHICH REPRESENTS YOUR FEELINGS CONCERNING THE QUESTION.Did the builder cause what happened at the end of the tape?completely not at all
caused______________________________________________causedDid the reader cause what happened at the end of the tape?not at all completely
cans ed _____ ____________________________________causedDo you think the reader foresaw what happened?not at all completely
fores aw »__ ________________________________________ f ores awDo you think the builder foresaw what happened?completely not at all
foresaw______ ______________________ ________________ _f ore sawDo you think the reader intended to push the table?completely not at all
intended______________________________________________ intendedHow responsible is the reader for what happened?completely not at allresponsible________________ responsibleHow responsible is the builder for what happened?not at all completelyresponsible________ responsible
62How conscientious a person is the reader?completely not at allcons c lent io us_____ conscientiousHow likeable a person is the reader?completely not at all
likeable_____________ ._______________ likeableHow likeable a person is the builder?not at all likeable
likeable__________________________________________ completelyHow much did the builder make as a reward for his performance?
$ 0 . 0 0 _ __________________________________________ $6.7 5How similar are the builder’s beliefs and attitudes to yours?completely not at all
similar ______________________________________ similarHow similar are the reader’s beliefs and attitudes to yours?completely not at all
s i m i l a r ________________________________________similarDid what happened at the end of the tape affect the builde r ’s reward?completely not at all
a f f e c t_______________________________ ______________ a f f e c tHow similar are the reader’s personality traits to yours?completely not at all
similar______________________________________________ similarHow similar are the builder’s personality traits to yours?not at all completely
similar similar
63
ReferencesApsler, R. & Friedman, H. Chance outcomes and the just
world: A comparison of observers and recipients.Journal of Personality and Social Psychology, 19 75,3 1 , 887-894.
Bradley, G. W. Self-serving biases in the attribution process: A reexamination of the fact or fiction question. Journal of Personality and Social Psychology, 1978,3 6 , 56-71.
Brewer, M. B. An information-processing approach to attribution of responsibility. Journal of Experimental Social Psychology, 1977, 13, 58-69.
Buss, A. R. Causes and reasons in attribution theory: Aconceptual critique. Journal of Personality and Social Psychology, 1978, 3 6 , 1311-1321.
Byrne, D. Interpersonal attraction and attitude similarity. Journal of Abnormal and Social Psychology, 1961, 6 2 , 713-715.
Chaiken, A, L., & Darley, J. M. Victim or perpetrator: Defensive attribution of responsibility and the need for order and justice. Journal of Personality and Social Psychology, 1973, 2 5 , 268-275.
Cialdini, R. B., Braver, S. L., & Lewis, S. K. Attributions! bias and the easily persuaded other. Journal of
64
Personality and Social Psychology, 1974, 30 , 631-637 .Duval, S., & Wicklund, R. A. Effects of objective self-
awareness on attribution of causality. Journal of Experimental Social Psychology, 197 3, 9̂ , 17-31.
Feather, N. T. Attribution of responsibility and valence of success and failure in relation to initial confidence and task performance. Journal of Personality and Social Psychology, 1969, 129-144.
Feather, N. T., & Simon, J. G. Attribution of responsibility and valence of outcome in relation to initial confidence and success and failure of self and others. Journal of Personality and Social Psychology, 1971a,18, 173-188.
Fishbein, M . , & Ajzen, I. Attribution of responsibility:A theoretical note. Journal of Experimental Social Psychology, 1973 , _9, 148-153.
Harvey, J. H . , Harris, B., & Barnes, R. D. Actor-observer differences in perceptions of responsibility and freedom. Journal of Personality and Social Psychology,1975, 22-28.
Heider, F. The psychology of interpersonal relations.New York: Wiley, 1958.
Izard, C. E. Personality similarity, positive affect,and interpersonal attraction. Journal of Abnormal and Social Psychology, 1960, 6 1 , 484-485.
Jackson, D. N. Personality research form. New York: Research Psychologist Press, 1968.
Jones, E. E., Bell, L., & Aronson, E. The reciprocation of attraction from similar and dissimilar others. In C. C. McClintock (Ed.), Experimental social psychology. New York: Holt, Rinehart, and Winston, 1973.
Jones, E. E., & Davis, K. E. From acts to dispositions:The attribution process in person perception. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 2). New York: Academic Press, 1965.
Jones, E. E., & Harris, V. A. The attribution of attitudes Journal of Experimental Social Psychology, 1967, _3, 1-24.
Jones, E. E., Kanouse, D., Kelley, H., Nisbett, R . ,%Valins, S., & Weiner, B. Attribution: Perceiving the
causes of behavior. Morristown, N. J.: General Learning Press, 1972.
Jones, E. E., & Nisbett, R. E. The actor and the observer: Divergent perceptions of the causes of behavior Morristown, N. J.: General Learning Press, 1971.
Jones, E. E., Rock, L., Shaver, K. G., Goethals, G. R.,& Ward, L. M. Patterns of performance and ability attribution: An unexpected primacy effect. Journal of Personality and Social Psychology, 196 8, JjO, 317-340.
Kelley, H. H. Causal schemata and the attribution process. Morristown, N. J.: General Learning Press, 1972.
^ K e l l ey, H. H. Attribution theory in social psychology.In D. Levine (Ed.), Nebraska symposium on motivation, 1967, Vol. 15. Lincoln, Nebraska: University of Nebraska Press, 1967.
Langer, E. J . , & Roth, J. Heads I win, tails i t ’s chance The illusion of control as a function of the sequence of outcomes in a purely chance task. Journal of Personality and Social Psychology, 1975 , 32 , 951-955 .
Lerner, M. J. Observer’s evaluation of a victim: Justice guilt, and veridical perception. Journal of Personality and Social Psychology, 1971 , 2 0 , 127-135 .
Lerner, M. J., & Matthews, G. Reactions to the suffering of others under conditions of indirect responsibility Journal of Personality and Social Psychology, 19 6 7,5, 319-3*25,
Lerner, M. J,, & Miller, I). T. Just world research and the attribution process: Looking back and ahead. Psychological Bulletin, 1978, 8 5 , 1030-1051.
Lerner, M. J., & Simmons, C. H. Observer’s reaction to the "innocent victim” : Compassion or rejection? Journal of Personality and Social Psychology, 1966,
203-210.McKillip, J,, & Posovac, E, J, Attribution of responsi
bility for an accident: Effects of similarity to the victim and severity of consequences. Proceedings of the 80th Annual Convention of the American Psycho-
67
logical Association, 1972, _7, 181-182.Miller, D. T. Ego-involvement and attributions of success
and failure. ' Journal of Personality and Social Psychology, 1976, 34, 901-906.
Miller, D. T. What constitutes a self-serving attribu-tional bias? A reply to Bradley. Journal of Personality and Social Psychology, 1978 , J36̂ 1221-1223.
Miller, D. T., & Norman, S. A. Actor-observer differences in perceptions of effective control. Journal of Personality and Social Psychology, 1975, _31̂ 503-515.
Miller, D. T., & Ross, M. Self-serving bias in the attribution of causality: Fact or fiction? PsychologicalBulletin, 1975, 82̂ , 213-225.
Monson, T. C., & Snyder, M. Actors, observers, and the *
attribution process. Journal of Experimental Psychology, 1977, L3, 89-111.
Nie, M. H., Hull, C. H . , Jenkins, J. G., Steinbrenner,K . , & Bent, D. H. Statistical package for the social sciences. New York: McGraw Hill, 1975.
Nisbett, R. E., Caputo, G., Legant, P., & Maracek. J.Behavior as seen by the actor and as seen by the observer. Journal of Personality and Social Psychology,
1973, AZ> 154-164.Nisbett, R. E., & Valins, S. Perceiving the causes of
o n e ’s own behavior. Morristown, N. J.: General Learning Press, 19 71.
68
Phares, E. J., Wilson, K. G., & Klyver, N. W. Internal- external control and reaction to threat. Journal of Personality and Social Psychologyv 1971, , 2 85-288.
Reader, N . , & English, H. B. Personality factors inadolescent female friendships. Journal of Consulting Psychology, 1947, 11, 212-220.
Regan, D. T., Straus, E., & Fazio, R. Liking and theattribution process. Journal of Experimental Social Psychology, 1974, ^0, 385-397.
Regan, D. T . , & Totten, J. Empathy and attribution: Turning observer into actor. Journal of Personality and Social Psychology, 1975, 3_2, 850-856.
Rotter, J. B. Generalized expectancies for internal versus external control of reinforcement. Psycho logical Monographs, 1966, j$_0 (1, Whole No. 609).
Schroeder, D. A., & Linder, D. E. Effects of actor’scausal role, outcome severity, and knowledge of prior accidents upon attribution of responsibility.Journal of Experimental Social Psychology, 19 7 6,
accident. Journal of Personality and Social Psychology, 1970, 14_, 101-113.
Shaver, K. G. Defensive attribution: Effects of situational possibility and personal similarity on assignments of responsibility. (Mimeo) Reported in K. G.
12, 340-356.K. G. Defensive attribution: Effects of severity
and relevance on the responsibility assigned for an
Shaver, Defensive attribution of responsibility, Washingtons National Institute of Mental Health, 1971.
Shaver, K. G. Intentional ambiguity in the attribution of responsibility: A reply to Fishbein and Ajzen. Unpublished manuscript, College of William and Mary, 1973.
1/ Shaver, K. G. An introduction to attribution processes. Cambridge, Mass.: Winthrop, 1975.
Shaw, M. E., & Sulzer, J. L. An empirical test of Heider's levels in attribution of responsibility. Journal of Abnormal and Social Psychology, 19 6 4, 69 , 39-46 .
Sicoly, F., & Ross, M. Facilitation of ego-biased attributions by means of self-serving observer feedback. Journal of Personality and Social Psychology, 19 7 7 ,3 5 , 734-741.
%Sosis, R. H. Internal-external control and the perception
of responsibility of another for an accident. Journal of Personality and Social Psychology, 19 74 , 3 0 ,
395-399.Sorrentino, R. M . , & Boutiiier, R. G. Evaluation of a
victim as a function of fate similarity-dissimilarity, Journal of Experimental Social Psychology, 19 7 4 , 1 0 ,
,84-93 .Storms, M. B« Videotape and the attribution process:
Reversing actor's and observer’s points of view.Journal of Personality and Social Psychology, 19 7 3,
70
Streufert, S., & Streufert, S. C. Effects of conceptual structure, failure, and success on attribution of causality and interpersonal attitudes. Journal of Personality and Social Psychology, 19 69 , _1_1, 138-147.
Sulzer, J. L. Heider's "levels model" of responsibility attributions. Paper presented at the Symposium on Attribution of Responsibility Research, Williamsburg,Va., July, 1971.
Taylor, S. E., & Fiske, S. T. Point of view and perceptions of causality. Journal of Personality and Social Psycho
logy, 1975, 32., 439-445.Vidmar, N . , & Grinklaw, L. D. Attributing responsibility for
an accident: A methodological and conceptual critique. Canadian Journal of Behavioral Science, 1974, j6,112-130.
^ Walster, E. Assignment of responsibility for an accident. Journal of Personality and Social Psychology, 1966,2, 73-79.
Walster, E. "Second-guessing" important events. Human Relations, 1967, 2£, 239-250.
Weiner, B., Frieze, I., Kukla, A., Reed, L., Rest, S., & Rosenbaum, R. Perceiving the causes of success and failure. New York: General Learning Press, 1971.
Weiner, B., & Kukla, A. An attributional analysis ofachievement motivation. Journal of Personality and Social Psychology, 19 7 0, 15, 1-20.
Winer, B. J. Statistical principles in experimental de sign. New York: McGraw Hill, 1971.
Wolfson, M. R . , & Salancik, G. R. Observer orientation and actor-observer differences in attributions for failure. Journal of Experimental Social Psychology 1977, 13., 441-451.
72
VITA
Glenn Paule-Carres
Born in Chicago, Illinois, Hay 14, 1949. Graduated from Lane Technical High School, Chicago, Illinois, 1967; B.A., Knox College, Galesburg, Illinois, 1972; M.A. candidate in Department of Psychology, College of William and Mary, Williamsburg, Virginia, 1974-1976, degree conferred August, 1979. In September, 1976, the author entered the Doctoral Program in Clinical Psychology at Loyola University of Chicago; proposed date of graduation-'- February, 1980.
The author is married to Pattimae Paule-Carres and has one child, Larissa. He plans on settling in Fairfax County, Virginia and pursuing a career as a psychotherapist,
