Journal of Personality and Social Psychology

1974, Vol. 30, No. 1, 57-63

AWARENESS OF INTERPERSONAL CONSEQUENCES,

RESPSONSIBILITY DENIAL, AND VOLUNTEERING x
SHALOM H. SCHWARTZ 2
Department of Sociology, University of Wisconsin—Madison

The effects on helping of two factors presumed to influence awareness of

consequences for others were tested. We hypothesized that helping increases
as (a) the salience of consequences in the situation increases, and—when con-
sequences are not salient—(6) the individual tendency to think spontaneously
of consequences for others increases. The tendency to accept versus deny
responsibility for the consequences of one's own behavior was also hypothesized
to increase helping, but no effect for internal-external control was anticipated.
Seventy coeds, who had completed relevant personality scales, received tele-
phone appeals to help raise money for Head Start, in which salience of the
consequences of volunteering was manipulated. All hypotheses were confirmed.
A high refusal rate among those high on awareness of consequences who were
exposed to high salience is discussed in terms of reactance thresholds. The
internal-external control results are viewed as suggesting conditions under
which internal-external control influences helping.

One precondition for the activation of was designed to clarify the impact on helping
moral norms postulated by Schwartz (1968a, of both salience of consequences and disposi-
1968b, 1970b) is that an individual be aware tional awareness of consequences.
that his potential actions have some conse- Tilker (1970) studied helping responses in
quences for the welfare of another. The par- the form of intervening by observers to stop
ticular consequences that enter his awareness the shocks being administered by another to
determine which specific norms he will see as a "learner" victim. He found that observers
pertinent. The probability that an individual who received both auditory and visual feed-
will become aware of particular consequences back from the victim regarding his suffering
for others may be a function of how promi- intervened more quickly and more directly
nent or conspicuous these consequences are in than did those receiving only auditory feed-
the situation, a variable we call "salience of back; and the latter intervened more than
consequences." Awareness of consequences did those receiving no feedback. The serious-
may also depend on a dispositional tendency ness of consequences of action or inaction
to perceive and think in terms of the conse- made known by verbal statements and by
quences of action for others. markings of intensity on the shock generator
It follows from the above that in situa- was constant across conditions. Variations in
tions governed by widely accepted helping feedback increased the salience of these con-
norms, factors conducive to awareness of sequences. Thus it was salience rather than
interpersonal consequences promote activa- degree of consequences or dependency that
tion of these norms, and hence foster helping varied.
behavior through normative pressure. Evi- Results of a second study manipulating
dence supporting this hypothesis is meager salience of consequences were less clear.
and somewhat problematic. The current study Schwartz (1970a) varied the salience of the
1
This research was supported by Grants GS-2048
consequences for a dying woman of locating
and GS-3422SX from the National Science Founda- a matching bone-marrow-transplant donor in
tion and by the Wisconsin Alumni Research Fund. a direct appeal for volunteers. When the odds
Thanks are due to Connie Kolpin and David Cray, for finding a matching donor were described
who coded the awareness of consequences protocols, as very small (1/1,000), volunteering rose
and to Kristi Olah, who served as experimenter.
2
Requests for reprints should be sent to Shalom across three levels of increasing salience.
Schwartz, Department of Sociology, University of When the odds were described as greater
Wisconsin, Madison, Wisconsin 53706. ( 1 / 2 5 ) , however, volunteering increased from
57
58 SHALOM H. SCHWARTZ

low to moderate salience but dropped to its (f = .44) among those judged high in spon-
lowest rate under high salience of conse- taneous attention to interpersonal conse-
quences. A similar boomerang effect, for vol- quences on the basis of their awareness of
unteering to be on call as a marrow donor in consequences scores.
a nationwide pool of potential donors, was The original awareness of consequences
found in the high salience of consequences scale was designed for use with males, and
condition in a later study (Schwartz, 1973).8 only half of the projective stories it con-
The boomerang effect in these studies may tained were deemed valid (Schwartz, 1968a).
be explained by suggesting that direct ap- A new version was therefore developed for
peals arouse psychological reactance. This use in later research. This version was ad-
reactance is expressed overtly as refusal whenministered to subjects who participated in a
the pressure produced by vividly portraying study of bystander reactions to an emergency
consequences is experienced as illegitimate (Schwartz & Clausen, 1970). Awareness of
(e.g., as exploitative manipulation of emo- consequences was found to be unrelated either
tions). In contrast, when potential helpers doto speed or type of help offered by bystand-
not attribute the salience of consequences to ers, whether male or female. Nor did aware-
a source possibly actuated by a desire to in- ness of consequences interact with any of the
fluence or manipulate them—presumably as experimental conditions to influence helping.
in Tilker (1970)—high levels of salience are The failure of awareness of consequences
not perceived as illegitimate. In the current in the second study may be due to invalid
study, two levels of salience were chosen. measurement: The scale was administered
These levels were intended to be sufficiently while subjects were in a state of high arousal
weak that no unwarranted pressure would be and concern over their reactions to others
experienced even in a direct appeal. We hy- following exposure to the emergency. Of more
pothesized that a simple positive relationshiptheoretical interest, the presence of awareness
between salience of consequences and helping of consequences effects in the first study and
could be found under such conditions of a their absence in the second may be attributa-
direct appeal. ble to the much greater salience of conse-
Support for the existence of relatively quences in the emergency study. There, the
stable individual differences in the tendency victim directly cried out for help while ap-
to think in terms of consequences for the parently choking and in pain. We may infer
welfare of others is forthcoming from only that individual differences in spontaneous
one study. Schwartz (1968a) found that awareness of consequences were swamped by
scores on a projective test designed to tap the salience of consequences in the situation.
spontaneous awareness of consequences were Everyone was made aware that the victim
significantly, though not highly (f — .27), might suffer serious damage unless he was
related to peer ratings of helpfulness, consid-
helped. Hence, differences in awareness of
erateness, and reliability in daily contacts in
consequences could have no additional im-
a sample of college males. In this study, per-pact on norm activation and thereby on help-
sonally accepted norms pertaining to the ing.
behavior were also measured. These data per- Based on the above inference, we hypothe-
mitted a more direct test of the postulated sized the following interaction between sali-
process of awareness leading to norm activa- ence of consequences and awareness of con-
tion leading to behavior. Consistent with this
sequences in the current study: Awareness of
process, no association between personal consequences is positively related to helping
norms and peer-rated behavior was found for under low salience of consequences but is
those scoring low on awareness of conse- unrelated to helping under high salience of
quences, but there were substantial correla- consequences.
tions between personal norms and behavior While this research was designed to exam-
8
S. Schwartz. Unpublished data, July 1972. ine variables presumed to influence awareness
(Available from the author.) of consequences for others, two other varia-
 AWARENESS OF INTERPERSONAL CONSEQUENCES AND VOLUNTEERING 59

bles were also measured. The first, ascription Awareness oj Consequences Scale
of responsibility, refers to the disposition to This is a projective scale described as "a test of
accept or deny one's own responsibility for your understanding of the way people go about
the interpersonal consequences of his actions. making the choices they do." It presents six inci-
Assuming norms have been activated through dents in which the central character faces a decision
that has potential consequences for the welfare
awareness of consequences and focusing of of others. Subjects are asked what thoughts and
responsibility in an appeal, these norms may feelings might be running through the central char-
yet be neutralized and lose their potential acters' minds as they decide what to do. Evaluation
impact on behavior if an individual denies of responses assumes that individuals introduce into
his personal responsibility for the conse- their descriptions of how others make decisions pri-
marily those thoughts that might enter their own
quences (see Schwartz, 1970b, for a fuller minds in similar circumstances.
analysis). Thus variations in ascription of Responses are coded on a 5-point scale for "the
responsibility might obscure the impact of extent to which the actor is aware of the potential
the consequences variables. To control for this consequences of his behavior for the welfare of oth-
ers," regardless of actual decisions reached and
obscuring effect, ascription of responsibility consequences mentioned for self. A score of 0 indi-
was treated as a covariate. We hypothesized cates no awareness; a score of 4 indicates awareness
a positive correlation between ascription of of long-range consequences, adopting the perspective
responsibility and helping, because higher of others and reflecting on consequences from their
ascription of responsibility scores indicate a viewpoint. The intercoder reliability for total scores
of two independent coders was r = .85, with agree-
tendency to accept rather than deny responsi- ment within 1 scale point on 95% of the responses.
bility. Scores for all six stories were positively intercorre-
The final variable measured was internal- lated, and they all loaded positively on a single
external control (Rotter, 1966). Findings of factor.6
several studies have been interpreted as dem- Procedure
onstrating that helping is associated with
Subjects received the following telephone contact
attributing the control of an individual's fate during the evening hours from a female experimenter
to internal sources. We contend that the cor- who was blind to their status on awareness of conse-
relations of internal control with helping in quences, ascription of responsibility, and internal-
these studies can better be understood as external control.
evidence that internals tend to take active Hello. May I speak to _? My name is
steps to improve their own lot (Gore & Rot- Mrs. Ramlow, and I'm a UW [University of
ter, 1963; Strickland, 1965) or to control Wisconsin] student interested in the Madison
Head Start program. Head Start is trying to raise
their own outcomes and demonstrate their funds to make up for its low budget. To help
competence (Midlarsky, 1971)." In the cur- raise money, a bake sale has been planned. We've
rent study, subjects were unlikely to perceive arranged to set up a booth on the Square for
helping as a means to improve their own lot two weeks beginning Monday and continuing
or to demonstrate competence or control. We until Saturday of each week. We need about 30
volunteers. Mothers of the Head Start children
therefore expected no association between have agreed to do the baking.
internal-external control and helping.
Salience of consequences was now manipulated by
METHOD spelling out the implications of helping for the
mothers, their children, and the project. The conse-
The awareness of consequences, ascription of re- quences mentioned in the fo'g^-salience condition
sponsibility, and internal-external control scales had could easily be inferred in the Zow-saliencc condition
been completed by 58 female sophomores or juniors by subjects who thought in terms of the conse-
10 to 14 months earlier in conjunction with another quences themselves.
study (Schwartz & Clausen, 1970). An additional
4
12 coeds completed the scales 1 month before the E. Midlarsky. Some determinants of aiding
current study in response to an offer of $2 to "help under experimentally induced stress. Unpublished
develop norms for Wisconsin students to compare manuscript, 1971. (Available from E. Midlarsky,
with those at other colleges." The 70 subjects were Department of Psychology, University of Denver,
stratified into those above and below the median on Colorado 80210)
5
awareness of consequences and then were randomly For a more complete discussion of the character-
assigned to one of two levels of salience of conse- istics of the awareness of consequences scale, see
quences. Schwartz (1968a).
60 SHALOM H. SCHWARTZ

TABLE 1 thanked and given an address to contact for infor-

INTERCORRELATIONS, MEANS, AND STANDARD mation about other things they might do for Head
DEVIATIONS OF VARIABLES Start. Several girls were referred to Head Start in
this way, but this further expression of helpfulness
Intercorrelations could not be examined as a function of the manipu-
lations because the latter dealt specifically with the
Variable
Units
M SV bake sale. The index of helping behavior used in the
AC AR IE volun- analysis was the number of half-hour units volun-
teered teered.

Salience of RESULTS
consequences -.01 .00 -.09 .26*
AC -.08 .04 .01 8.04 2.87
AR -.14 .48** 19.94 4.82 The means, standard deviations, and inter-
IE -.06 11.90 4.61
Units correlations among the variables are presented
volunteered 3.00 3.43
in Table 1. The randomization of awareness
Note. AC *=> awareness of consequences, AR = ascription of consequences, ascription of responsibility,
of responsibility, IE = internal-external control.
* p < .025.
and internal-external control across levels of
**p < .001. salience of consequences is attested to by the
near-zero correlations in the first row.
[High] But they can't do the selling, too, The intercorrelations also indicate that the
because then they will have to be away from
their children much of the day. However, mothers three personality variables were relatively
will be able to stay home if college girls do the independent of one another.
selling. So I've just started calling girls whose Overall, 53% of the coeds volunteered
names I picked out of the directory to see if we time (a minimum of 1 hour and a maximum
can get enough volunteers so we can carry the
project off and help the Head Start kids.
of 12 hours) to man the bake-sale booth.
[Low] The plans call for the booth to stay open Correlations with helping, indexed by the
from morning to evening each day throughout number of half-hour units volunteered,
the week. College girls will man the booth and do showed the predicted relationships for sa-
the selling. So I've just started calling girls whose lience of consequences and ascription of re-
names I picked out of the directory to see who
would like to volunteer to help out with this. sponsibility and the expected lack of asso-
ciation for internal-external control. The
The manipulation was followed by the detailed importance of denial of responsibility for
request for help in both conditions:
interpersonal consequences as a determinant
Could you help us with the bake sale during the of behavior in this setting, reflected in the
two weeks starting next Monday, . ? [Pause. ascription of responsibility correlation, af-
If yes] The booth will be open from about 9 a.m.
to about 6:30 p.m. Monday through Saturday. firms the value of including ascription of re-
We're signing people up for i-hour periods. How sponsibility as a covariate when examining
much time would you be willing to give? [Pause. the effects of the variables presumed to in-
If time specified] I'm really counting on the girls fluence initial awareness of consequences.
who volunteer; so if there's any doubt in your
mind I would rather not count on you. [Pause.
Ascription of responsibility accounted for
If they reiterate interest] We'll send you a card to almost 25% of the variance in volunteering.
remind you and give details later this week. In order to permit a detailed analysis of
Note that to maximize the validity of volunteering
the predicted Salience of Consequences X
as an index of helping, respondents were urged not Awareness of Consequences interaction, the
to volunteer unless they were certain they would sample was divided as nearly equally as pos-
actually come. It was implied that refusing would sible into five levels of awareness of conse-
be more helpful than agreeing and then failing to quences. Results of the analysis of covariance
show up. Furthermore, the experimenter paused
three times, permitting subjects to withdraw unless for units volunteered are reported in Table 2.°
they initiated three separate committing responses. A least-squares solution was employed be-
In every case the experimenter managed to deliver cause the unequal cell frequencies reflected
the entire manipulation prior to eliciting a positive 0
or negative response. Although the modal response was refusal (i.e., 0
Subjects who volunteered were sent a card three units), the data met the various assumptions under-
days later stating that due to unanticipated prob- lying the analysis of variance; hence no transforma-
lems, the bake sale had been postponed. They were tion was employed.
 AWARENESS OF INTERPERSONAL CONSEQUENCES AND VOLUNTEERING 61

the actual distribution of awareness of conse- TABLE 2

quences scores in the sample. The F ratio and ANALYSIS OF COVARIANCE FOR UNITS VOLUNTEERED
group means were adjusted for the linear
Source df MS F
effects of ascription of responsibility.
As hypothesized, subjects exposed to high Salience of consequences (A) 1 54.89 7.12**
salience of consequences volunteered more Awareness of consequences
grouped (B) 4 1.40 <1
time than those exposed to low salience (an AXB 4 27.22 3.53*
average of 3.86 versus 2.09 units), accounting Within 59 7.71
for 6% of the variance. In the high-salience
*p <.02.
condition, 67% of the subjects volunteered ** p < .01.
at least one hour compared with 38% in
the low-salience condition. A clear under- tously among those presumed most likely
standing of this effect, however, requires ex- to attend to interpersonal consequences on
amination of the Salience of Consequences X their own.
Awareness of Consequences interaction.
As indicated in Table 2, the predicted inter- DISCUSSION
action was significant. It accounted for 9.5%
of the variance in volunteering. Examination The predicted effect of salience of conse-
of the adjusted means portrayed in Figure 1 quences on helping was based on a postulated
reveals that the form of the interaction was process of norm activation. We hypothesized
only partly as hypothesized. Under low sa- that self-accepted helping norms relevant to
lience of consequences there was a linear in- specific consequences are more likely to be
crease in volunteering as awareness of conse- activated when varied and intense conse-
quences increased. This effect was reflected in quences for others are brought to awareness.
a correlation of .39 (p < .025) between Salience of consequences was further tied to
awareness of consequences and volunteering. norm activation by the supposition that men-
When consequences were only weakly high- tioning consequences explicitly in an appeal
lighted in the appeal, volunteering varied gives them a social validation that renders it
positively with the subjects' dispositional more difficult subsequently to deny their
tendencies to become aware of the conse- reality or meaning. As a result, the norms
quences for others. that are activated when the consequences are
By contrast, the correlation between aware-
ness of consequences and volunteering was
insignificant under high salience of conse-
quences (r = —.27, p > .10). Although this
absence of a relationship conformed to the
hypothesis, the hint of a negative effect in
the correlation and in Figure 1 came as a
surprise. A post hoc comparison among means
revealed that under high salience of conse-
quences, subjects with the highest level of
awareness of consequences volunteered signifi-
cantly less than did those with the first three
levels of awareness of consequences (p < .05,
Duncan's multiple-range test). While the 2-fj 6-7 6-y 10-11 12-15

latter showed virtually the same level of AWAftKNUSS OP CONSEQUENCHS XOHE (AC)

volunteering, a level higher than that for FIGURE 1. Mean number of half-hour units volun-
any of the groups exposed to low salience of teered for bake sale as a function of awareness of
consequences and salience of consequences. (Numbers
consequences, volunteering under high sa- in parentheses represent size of sample on which
lience began to drop at the fourth level of each mean is based. Means are adjusted for the
awareness of consequences and fell precipi- linear effects of ascription of responsibility.)
62 SHALOM H. SCHWARTZ

brought to awareness cannot as easily be the view that awareness of consequences taps
neutralized by denial of consequences. spontaneous awareness, a view derived from
Salience of consequences in an appeal might considering the item format and coding sys-
also increase helping through two alternative tem. If awareness of consequences tapped a
processes: by raising the potential helper's denial tendency, it would relate positively to
level of arousal (Piliavin, Rodin, & Piliavin, helping, especially under high salience, since
1969) or by strengthening his empathic ex- awareness of consequences would influence
perience (Aronfreed, 1970). These alternative denial of the consequences highlighted in the
processes seem unlikely to have been impor- appeal. The findings implied, instead, that
tant here, however. In the high-salience condi- awareness of consequences was swamped
tion, the information about consequences was under high salience and had either no effect
provided matter-of-factly, without any vivid, or a negative one on helping. Under low sa-
emotionally arousing content. The conse- lience, however, the positive relation of aware-
quences mentioned were such that they might ness of consequences to helping, expected for
have been noted by subjects on their own a spontaneous tendency, appeared.
with a little thought. Moreover, the potential Why did subjects showing the highest level
helpers—white coeds—were socially dissimi- of awareness of consequences volunteer so
lar and psychologically detached from the little under high salience of consequences?
disadvantaged mothers and children they were The expectation of a relatively high level of
asked to help. volunteering under high salience was con-
The findings for salience of consequences firmed for those low or moderate on aware-
paralleled those for low and moderate levels ness of consequences; but those highest on
of salience in Schwartz (1970a). Together awareness of consequences apparently re-
these studies indicate that, even in an appeal, sponded negatively to having their attention
raising salience of consequences to a moder- drawn to interpersonal consequences. Perhaps
ate level can increase help. The questions of individuals very high on awareness of conse-
how high salience can become before eliciting quences have a lower threshold for experi-
a boomerang of refusal and what variables encing reactance to an appeal for help. It
influence the reactance threshold must still be seems logical that those who recognize the
investigated. One hint regarding the latter interpersonal consequences in a situation may
question is contained in the awareness of view emphases on consequences in an appeal
consequences results. as more blatant, unwarranted, and manipula-
The confirmation of the hypothesized tive, and hence be moved to refuse. This
Awareness of Consequences X Salience of reasoning could account for the unexpected
Consequences interaction offered some valida- finding, but it requires systematic investiga-
tion both for the concept of a disposition tion. We admit to uneasiness at turning to a
spontaneously to become aware of conse- reactance explanation here because the pres-
quences for others and for the scale used to sure exerted to persuade respondents in the
operationalize this concept. The stability of high-salience condition appear to us to have
this disposition was indirectly supported by been rather weak.
the fact that the hypothesis was confirmed The correlation of .48 between ascription
even though more than 80% of the subjects of responsibility and volunteering is unusually
completed the awareness of consequences scale high for predicting a specific behavior with a
10 or more months prior to the experimental general disposition, especially after almost a
contact. year. The ascription of responsibility correla-
In an earlier study (Schwartz, 1968a) it tion here was stronger than that for predicting
was suggested that the awareness of conse- helping reactions in an emergency (r = .18;
quences scale might tap either spontaneous Schwartz & Clausen, 1970) and in everyday
awareness or the tendency to deny interper- peer contacts (f = .28; Schwartz, 1968b). In
sonal consequences following their entry into the earlier studies ascription of responsibility
awareness. Confirmation of the interaction served primarily as a modifier of the impact
hypothesis in the current study reinforced of norms and situational variables on helping.
 AWARENESS OF INTERPERSONAL CONSEQUENCES AND VOLUNTEERING 63

Comparing the performance of ascription of mental analyses. In J. Macauley & L. Berkowitz

responsibility in the various studies suggests (Eds.), Altruism and helping behavior. New York:
Academic Press, 1970.
that it has a particularly strong direct impact
GORE, P., & ROTTER, J. A personality correlate of
in situations such as the present one, where social action. Journal of Personality, 1963, 31,
many rationales for denying personal respon- 58-64.
sibility to help are available and admissible MIDLARSKY, E. Aiding under stress: The effects of
(e.g., there are lots of other girls; there's no competence, dependency, visibility, and fatalism.
special need for me; I have too much school Journal of Personality, 1971, 39, 132-149.
PILIAVIN, L, RODIN, J., & PILIAVIN, J. Good samari-
work). Ascription of responsibility seems to tanism: An underground phenomenon? Journal of
have less direct impact when others are likely Personality and Social Psychology, 1969, 13, 289-
to sanction failure to help regardless of the 299.
actor's denial of responsibility, as in an ROTTER, J. Generalized expectancies for internal
emergency or when a dependent peer has versus external control of reinforcement. Psycho-
logical Monographs, 1966, 80(1, Whole No. 609).
power to respond.
SCHWARTZ, S. Awareness of consequences and the
As expected, internal-external control was influence of moral norms on interpersonal behav-
unrelated to volunteering. Nor did it explain ior. Sociometry, 1968, 31, 35S-369. (a)
additional variance when included in the SCHWARTZ, S. Words, deeds, and the perception of
regression equation for predicting volunteer- consequences and responsibility in action situa-
ing that already contained the other variables. tions. Journal of Personality and Social Psychol-
ogy, 1968, 10, 232-242. (b)
At least one other failure of internal-external SCHWARTZ, S. Elicitation of moral obligation and
control to predict helping has been reported self-sacrificing behavior: An experimental study of
(Schwartz & Clausen, 1970); and one volunteering to be a bone marrow donor. Journal
wonders if other insignificant results have not of Personality and Social Psychology, 1970, 15,
gone unpublished. If we define helping as 283-293. (a)
SCHWARTZ, S. Moral decision making and behavior.
acting with the intention to improve another's In J. Macauley & L. Berkowitz (Eds.), Altruism
status, the findings of the helping studies in and helping behavior. New York: Academic Press,
which internal-external control has been used 1970. (b)
give little support to the claim that internal- SCHWARTZ, S. Normative explanations of helping be-
external control influences helping. If we de- havior: A critique, proposal, and empirical test.
fine helping in terms of the outcome of action Journal of Experimental Social Psychology, 1973,
9, 349-364.
for the other, internal-external control ap- SCHWARTZ, S., & CLAUSEN, G. Responsibility, norms,
pears related to helping when the activity and helping in an emergency. Journal of Person-
that helps coincidentally enables the actor to ality and Social Psychology, 1970, 16, 299-310.
demonstrate his competence or achieve his STRICKLAND, B. The prediction of social action from
own goals through controlling his environ- a dimension of internal-external control. Journal
of Social Psychology, 1965, 66, 405-412.
ment. Future research will test the validity
TILKER, H. Socially responsible behavior as a func-
of these assertions. tion of observer responsibility and victim feed-
back. Journal of Personality and Social Psychol-
REFERENCES
ogy, 1970, 14, 95-100.
ARONFREED, J. The socialization of altruistic and
sympathetic behavior: Some theoretical and experi- (Received November 6, 1972)