Nancy: Research Class Assignment

Due by Sunday afternoon….please make sure run the assignment in “Turn It In” and make sure there isnt not
even 1% of plagarism.
This is for a Research Methods in Psychology class. This summary must not be a re-wording of the study’s abstract.  You are to read the study and summarize the main points


Our team is doing Bystander and Diffusion of Responsibility.


There are three articles to review. Each article must have a the questions answered from below:

1) brief background for the study (context for the study)

2) the hypothesis tested

3) who the participants in the study were 

6) general method

7) a brief summary of the results

8) how the authors interpreted their findings. If there is more than one experimental study reported in the article you should select the study that best demonstrates the topic the authors are testing and summarize that study. 

9) Finally you should give an analysis of how the study you reviewed relates to the study you and your team are considering (my team are doing Bystander and Diffusion of Responsibility).

10) Good grammar, punction, error free etc.


In addition, please make sure not to exceed more than one page per article
. This needs to be in APA format so double space with 12 point arial font.


If you have any questions please contact me.


The 3 articles are attached.


Bystander Response to Arterial Bleeding: Helping Skills, the
Decision-Making Process, and Differentiating

the Helping Response

R. Lance Shetland and William D. Heinold
Pennsylvania State University

Using a mock injury involving arterial bleeding, we explored the impact of
variables on two different decisions in the decision-making process leading to
help. Expertise (Red Cross Training), ambiguity, and number of bystanders were
manipulated in a 2 X 2 X 2 design. We observed the following responses: (a)
nonhelp, (b) ineffective direct help, (c) indirect help, and (d) direct help. The
decision to help or not was affected by ambiguity, sex, and the presence or
absence of other bystanders. Greater ambiguity led to less help; women helped
less than men; and fewer people helped when other bystanders were present.
Expertise affected the decision leading to the type of help used but not the
decision to help. Although training did not raise the intervention rate, it did
dramatically change the effectiveness of the help used and could yield as many as
28 more saved lives out of 80 such incidents. The number of bystanders affected
both the decision to help and the type of help used. Ineffective direct help
occurred most frequently when the bystander was alone. The presence of other
bystanders also affected the type of indirect help that was used.

The end of the second decade of research
on the bystander response to emergencies is
approaching. Surprisingly, after this length of
time, little is known in three important areas
about helping. First, there is only sparse
knowledge concerning the impact of helping
skills on the behavior of bystanders to emer-
gencies. This is so despite a research literature
on “competency,” which one might expect to
provide such an answer. Second, despite a
number of studies (see Piliavin, Dovidio,
Gaertner, & Clark, 1981, pp. 110-111) in
which researchers have found bystanders re-
sponding in both direct and indirect fashions,
little is known about the variety of helping
responses and the conditions that produce
these different responses. Third, there has

William D. Heinold was an Instructor, Health Educa-
tion, at the Pennsylvania State University during the
time of the study and is now a Visiting Lecturer at
Huaibei Teachers College, Huaibei, Anhui Province, Peo-
ple’s Republic of China.

Requests for reprints should be sent to R. Lance
Shetland, Department of Psychology, 543 Moore Building,
Pennsylvania State University, University Park, Pennsyl-
vania 16802,

been little or no research that indicates the
role that different variables play on each
decision that leads to a helping response.
These are the areas that we have investigated.
We begin our discussion of the literature by
examining studies in which the “competence”
of bystanders on helping has been investigated.
We then discuss the literature pertinent to
the decision-making process that leads to a
helping (or nonhelping) response.

Form and Nosow (1958), reporting on the
aftermath of a major disaster, suggested that
the expertise of bystanders was an important
determinant of helping. Since that time, sev-
eral investigators have looked at the perception
of expertise by a bystander and its relation
to helping, A number of investigators have
used a “false feedback” strategy to induce
“competence” and then tested the impact of
this variable on helping, using experimental
laboratory tasks. Midlarsky (1971) and Mid-
larsky and Midlarsky (1973) induced subjects
to believe that they could withstand electric
shock better or worse than the average subject
and then placed these subjects in a position
in which thev could receive an electric shock

Journal of Personality and Social Psychology, 1985, Vol. 49, No. 2,


Copyrighl 1985 by ihe Amerkan Psychological Association, Inc. 0022-3514/B5/S00.7




for someone else. Schwartz and Ben David
(1976) manipulated subjects’ perceptions of
their ability to handle rats. Subjects then
found themselves in a situation in which they
could help by capturing an escaped rat. Ka7-
din and Bryan (1971) manipulated in subjects
relevant (physical factors) or irrelevant (cre-
ativity) competence pertaining to the act of
donating blood. Those authors used a behav-
ioral measure of the propensity to donate
blood as a dependent measure. Ashton and
Severy (1976) manipulated competence by
informing subjects that they scored at the
90th or 50th percentile on generalized com-
petence, and then placed subjects in a position
so that they could respond to an accident.

The results of these investigations generally
suggest that competence either produced more
or quickened the rate of helping. An exception
to this generalization was the study by Ashton
and Severy (1976), which did not show a
significant effect of competence on helping.
In addition, Midlarsky and Midlarsky (1973)
found that highly competent subjects helped
more only when costs were low. Interestingly,
Kazdin and Bryan (1971) discovered that
both relevant and irrelevant competence pro-
duced greater helping. It then seems that an
alternative interpretation of this literature
may be that these manipulations resulted in
more help by producing positive mood states
rather than or in conjunction with perceptions
of efficacy.

Two studies provided subjects with knowl-
edge pertinent to the helping situation. Bea-
man, Barnes, Klentz, and McQuirk (1978)
gave subjects information concerning psycho-
logical factors that normally inhibit helping
in emergencies. They found that those who
had been exposed to these facts helped more
frequently. Pantin and Carver (1982) showed
subjects films that detailed various first-aid
procedures. They found that subjects who
had not seen the films responded more slowly
in large groups than in small ones.


who had seen the films responded more
quickly regardless of group size.

There are several other studies in which
the authors examined the competency vari-
able but in either a nonexperimental or a
correlational fashion. Clark and Word (1974)
found that people with knowledge of electric-
ity tended to help more frequently during an

electrical emergency. Shetland and Stebbins
(1980) and Huston, Ruggierio, Conner, and
Geis (1981) found that people who believed
themselves to be physically able or competent
were more likely to help directly in crime
situations. In addition, Huston et al. found
that these people were more likely to have
been involved in a number of different formal
training programs.

The medical literature includes several
studies in which researchers implemented
compound manipulations in order to raise
the survival rate of’Victims of acute myocar-
dial infarction by creating bystander and
victim awareness of early warning signs and
upgrading other areas of emergency care.
Unfortunately no attempt is usually made to
evaluate the effectiveness of individual com-
ponents of these manipulations (e.g., Cramp-
ton, Aldrich, Gascho, Miles, & Stillerman,
1975). Eisenberg, Bergner, and Hallstrom
(1979) evaluated a program in Kings County,
Washington, that trained large numbers of
bystanders and instituted other changes. Those
authors did attempt to evaluate the effective-
ness of bystander training apart from other
changes, using logistic regression procedures.
They stated that the initiation of cardiopul-
monary resuscitation (CPR) was an important
factor and suggested that bystander-initiated
CPR can be substituted for this factor. The
conclusion that can be drawn is that by-
stander-initiated CPR is an important predic-
tor of survival rates for myocardial infarctions.
Because CPR is a technique requiring train-
ing, one can infer that training is correlated
with the reduction of death rates. However,
from the study’s design, one cannot infer that
the relation is a causal one.

From the review of the literature, it is clear
that little is known about the impact of
learned skills on the intervention rate of
bystanders. Our purpose was twofold. First,
we wished to test the impact of learned skills
on both the quantitative (amount of help)
and qualitative (type of intervention) aspects
of bystander intervention. For example, Shot-
land and Stebbins (1983) proposed in their
“limited altruism” model that the lack of
learned skills is one cause of indirect help.
Furthermore, we intended to evaluate this
impact in a causal manner.

A second but related purpose is to explore


how the independent variables affect specific
decisions that lead to a helping response.
Latane and Darley (1969) suggested that the
decision-making process consisted of five se-
quential stages: (a) noticing the event, (b)
interpreting the event as an emergency, (c)
deciding that the bystander has a responsibility
to act, (d) deciding on the form of assistance
to implement, and (e) deciding how to im-
plement the assistance. Schwartz (1977) pre-
sented a decision-making model consisting of
six sequential stages. The two models are
similar in many respects, but Schwartz’s
model provides somewhat more detail. For
example, Latane and Darley’s first two steps
appear to be combined in Schwartz’s first
step: awareness of a state of need. Another
point of correspondence is Latane and Dar-
ley’s third step of personal responsibility and
Schwartz’s fifth step of activation of “moral
responsibility.” However, between Latane and
Darley’s second and third steps, Schwartz
placed two additional steps. The authors all
agreed that the last step or steps involve
decisions concerning how to help. We are
interested in how the independent variables
affect (a) the feeling of responsibility and the
decision of whether to act and (b) the decision
and implementation of how to help.

The bystander literature suggests several
important factors that have been shown to
affect helping. These variables may interact
with or possibly inhibit any effect of skills
training and may affect the decision-making
process at different stages. One such factor is
ambiguity (e.g., Clark & Word, 1972; Yaki-
movich & Saltz, 1971). In ambiguous situa-
tions, trained people may be more likely to
investigate the potential problem because they
do know how to solve the problem, whereas
untrained individuals may reinterpret it as a
nonemergency because they cannot provide
assistance. A second important factor in the
helping area is the impact of the number of
bystanders potentially able to help. Training
may inhibit the use of social cues in the
interpretation of the incident and diminish
the diffusion of responsibility (Latane & Dar-
ley, 1969). This occurs because training may
cause people to have greater confidence in
their ability to meet the emergency; hence
trained bystanders may feel less need to rely
on others either to interpret or to share the

responsibility to help. Therefore, our inde-
pendent variables consist of training, ambi-
guity, and the number of bystanders.

These independent variables should affect
different stages of the decision-making pro-
cess. We expected training to affect how
people help and possibly whether they help
by instilling confidence. We expected ambi-
guity to affect the rate of helping, but not the
mode of help delivered. We made this predic-
tion because ambiguity should affect only
whether the situation is interpreted correctly;
after this judgment, it should have no impact
on the decision-making process. We expected
that the number of bystanders should affect
the rate of helping and possibly the mode of
helping. Other bystanders serve as a resource
and could be asked to provide assistance.

Our dependent variable of helping was
examined in a different fashion than is cus-
tomary. Several authors have suggested that
the act of intervention can be carried out in
different ways, namely, directly intervening
or responding in some indirect fashion, such
as calling for help (Piliavin & Piliavin, 1972;
Shetland, 1976). In fact, as we mentioned
previously, in a dozen studies (see Piliavin et
al., 1981) three types of helping, direct, in-
direct, and nonhelp are categorized. Shetland
and Stebbins (1980) found that helping could
be separated into four categories: (a) direct,
(b) “passive” or ineffective, (c) indirect, and
(d) nonresponding. Clearly, in a situation that
requires training to solve, direct help can be
both effective and ineffective.

Helping categories can be ranked according
to their utility to the bystander. One might
expect people to resort to direct appropriate
helping if they know how. If they lack the
knowledge, one might expect bystanders to
resort to indirect helping (i.e., finding others
who can help appropriately). Alternatively,
they can attempt to help directly, but to do
so inadequately, or simply not respond to the
emergency. We now describe the outcome of
skills training on these different types of


Training Program

Each term the Health Education Department of Penn-

sylvania State University offers a two-credit course covering
a wide variety of trauma topics. On successfully com-


pleting the course, a student obtains American Red Cross
certification in Advanced First Aid and Emergency Care.
The usual course enrollment is approximately 250 students
and is instructed by 15 Red Cross certified first-aid

instructors. The types of trauma dealt with in the course
include severe bleeding, poisoning, skeletal injuries, pul-
monary and cardiac, arrest, and so on. One of the course
requirements has been participation in out-of-class re-
search projects. One of the experimental manipulations
was an aspect of the training received in this course.


The emergency chosen as a test was an injury involving
serious arterial bleeding. It was chosen because it is an
injury that is clearly defined as an emergency once the
injury is seen (Shetland & Huston, 1979). Severe arterial
bleeding can be considered a life-threatening emergency,
and death could occur from such a wound in a period
of 4-10 min if approximately a quart of blood were lost
by an adult (Hafen, 1981). Second, on sight the problem
is obvious because blood rhythmically spurts from the

wound. Third, there are standard accepted intervention
procedures that are easily observed and recorded and

would not injure a confederate. The proper intervention
for external bleeding is simply to apply direct pressure
to the wound and solicit outside help (Hafen, 1981).
Other emergency procedures such as CPR or the Heimlich

maneuver could lead to confederate injuries.

Independent Variables

Three variables were manipulated in a 2 X 2 X 2
(Ambiguity X Number of Perceived Bystanders X Level
of Training) design. Because training in control of bleeding
procedures was not presented until the beginning of the
fourth week of instruction, training was manipulated in
a pretest/posttest manner. Following the model of Yaki-
movich and Saltz (1971), we manipulated ambiguity so

that in the low-ambiguity condition the victim screamed
for help, and in the high-ambiguity condition there was

only the sound of the accident that produced the arterial
bleeding. The perceived number of bystanders was ma-
nipulated such that subjects were either alone or in a
group (one male and one female confederate in all cases).

Confederates were trained to show no visible reactions
to the accident stimuli, but to react normally to any
directions that subjects might issue to them during the
course of the emergency.


Students taking the training course who had had
previous first-aid training were eliminated from the subject
population. Eligible students were randomly assigned to
one of the eight cells of the design. Thus pretraining
subjects were randomly assigned to Weeks 1, 2, or 3,
whereas posttraining subjects were randomly assigned to
Weeks 8, 9; or 10 of a 10-week term. Subjects then self-
selected a time of the week from available appointments
in which to report to an ostensible interview concerning
student health services, conducted jointly by the College
of Health Planning and Administration and the U.S.
Department of Health and Human Services. One hundred
ten subjects were assigned to each of the pretraining and

posttraining groups; 105 and 104 prospective subjects,
respectively, signed up. Eighty-three usable subjects re-
sulted from the pretraining group and 80 from the
posttraining group. Subjects were lost because of missed
appointments, arriving with friends, coming up the back
steps and thereby not passing the location where the
accident was to occur, and so on. The usable subjects
were 100 female and 63 male students.


As subjects arrived at the campus ROTC building (at

45-min intervals) for their interviews, they necessarily
passed a worker in the foyer adjacent to the hall leading

to the interview room on the third floor. The worker was
standing atop a 7-foot steptadder, ostensibly repairing an
overhead glass lighting fixture. From the time subjects
entered the building until the “accident,” the worker
made working sounds using an electric drill, hammer,
and screwdriver. Subjects walked by the workman, through
the propped-open door of a set of double doors, in order
to enter a hail approximately 30 m long that led to the
interview room. The second door was forced closed by
work equipment. This arrangement allowed the worker

to observe subjects unobtrusively in the hall, but provided
concealment for the set-up and execution of the forth-
coming accident. After subjects walked through the double

doors they could not observe the workman in the foyer

When subjects arrived at the interview room, approx-
imately 10 m from the foyer, they saw a sign posted to
the door indicating that the interviewer had been called
downstairs and would return immediately. They were
instructed that they were meanwhile to begin filling out
one of the interview forms provided. If the subject was

to be with confederate fellow bystanders, the female
confederate was seated and working on the form and the
male confederate arrived shortly thereafter: When the
worker saw that Uie subjects (and confederates) were

seated in the hall, he moved behind the closed door out
of view, and prepared for the accident.

He first changed into a second pair of white painters’
pants which were hidden at the work site. These trousers
were identical to the pair he was wearing except equipped

with special devices for simulating an arterially bleeding
wound below the right knee. A realistic imitation wound,

plastic tubing, a hand pump, and theatrical blood were
used to create the pulsating bleeding emanating from his
leg.’ The “wound” was glued under a jagged tear in the
right pants leg just below the knee. The tubing was
strapped inside the trousers and ran from the wound to
the hand pump, which the worker could operate to
pulsate blood to the wound.

The worker made a pool of “blood” approximately 6
in. in diameter in the location where his right knee was
going to be and placed shards of broken glass around
the puddle. After the “stage” was set, the workman
toppled the ladder, causing a loud crash as ladder and
tools (hammer, saw, screwdriver, wood slats, and metal
screws, etc.) hit the linoleum-tiled floor. In the high-

1 All these components are commonly found in moulage
kits used to imitate wounds and bleeding and are some-
times used in the filmmaking industry.



ambiguity condition it was clear that an accident of some
type had occurred, but not whether help was needed. In

the low-ambiguity condition, the worker yelled, “Help”
immediately after the accident sound, thereby providing

that information. The exact nature of the accident was
not known to subjects unless they reentered the foyer, as
the incident could not be seen from the hallway. Lying
motionless and turned on his left side, the “victim”
pumped between 5- and 1-in. streams of “blood” at

approximately 1-s intervals with his right hand, which
was laid across his chest. This position completely con-
cealed the apparatus.

From the time of the end of the sound of the accident,
subjects were given 1 min to rise from their chairs. Those

subjects who remained seated after the minute had
elapsed were considered nonresponders and their trials
were terminated. If subjects who initiated a response
passed through the double doors to the foyer, they were
given an additional minute to complete a response. The
trial was terminated before this additional minute was
over if the subject made an effort to leave the floor and
came into contact with our “containment” men. These
men were stationed on the landing below each of the two
stairways that provided access to the top floor, where the
emergency took place. Subjects who initiated contact
with the containment person and described the emergency

or asked for help were counted as indirect interveners.
The trial was also terminated if subjects applied direct
pressure to the “wound”

Subjects were under constant surveillance through

hidden video cameras and their responses were coded
from the video monitors. Coding was cross-checked for
accuracy in a later viewing of the videotapes. No coding

errors were detected because the behavioral incidents of
interest were easily observed and were qualitative in
character All procedures were cleared through the Uni-
versity Behavioral Science Ethics Committee.


The scene was believable. Of 163 subjects,
only 6 showed any sign of suspicion. All six
became suspicious only after making a direct
pressure response. One of the suspicious sub-
jects was a surgical nurse who had slipped
through the subject screening and had applied
direct pressure to the wound before realizing
that the blood did not “smell right.” The
other subjects who became suspicious felt the
wound shift because of the pressure applied
to it.

As hypothesized, four different clusters of
responses were observed and coded. The first
category consisted of nonhelpers, who either
never explored the situation or explored but
offered no help. They generally remained
seated and did not exhibit much arousal. The
three remaining categories were composed of
people who wanted to help and explored the
situation after the sound of the emergency.
These subjects appeared to exhibit extreme

arousal, characterized by such behaviors as
trembling and the initiation, but did not
complete a number of different acts such as
approaching the victim, starting to leave the
scene, and so on. These categories were as
follows: ineffective helpers, who tried to help
directly and touched the victim without of-
fering appropriate first aid; indirect helpers,
who offered no appropriate first aid but sought
outside intervention; direct interveners, who
offered appropriate first aid consisting of di-
rect pressure. Although subjects are classified
by their response into only one category, they
may have exhibited other types of behavior
as well. Hence bystanders were categorized
by their most helpful response. Direct inter-
veners may have started “helping” with an
ineffective response (e.g., checking for a pulse)
then switched to direct pressure and shouted
for help. Given that the most immediate need
was stopping the blood, the person was clas-
sified as a direct responder. Indirect responders
frequently began with an ineffective response,
then made the second best response (the best
being to apply direct pressure themselves),
which was to try to find someone else who
could handle the situation. Ineffective helpers
attempted only ineffective help. Nonrespond-
ers produced no behavior that could be con-
strued as an attempt to help. The number of
subjects in each of the four response categories
by experimental condition is presented in
Table 1.

Statistical Testing Procedures

The statistical tests consisted of three sepa-
rate multiway contingency table analyses
(Bishop, Fienberg, & Holland, 1975). In the
three analyses, we made use of the same 2 X
2 X 2 X 2 (Training X Number of Confererate
Bystanders X Ambiguity X Sex) factorial de-
sign.2 The dependent variables differed in the
three analyses. The first enabled us to analyze
all four helping categories and was used to
determine which independent variables caused
differences in the modes of responding. This
analysis provided the only accurate portrayal
of the fit of a complete model in which one
can use all of the response categories.

The second analysis was designed to answer

2 Sex was not incorporated into the original design,
but was later included in the analysis.


Table 1

Percentages and Frequency Distribution of Subjects by Response Type for Training Level, Ambiguity
Level, and Number of Perceived Bystanders (Collapsed Across Sex)

Untrained group Trained group

Ambiguous Unambiguous Ambiguous Unambiguous

Response type


Ineffective help

Indirect help

Direct help




























the question “What factors determine whether
bystanders try to help?” This is the classical
helping behavior analysis. All three helping
categories were collapsed into a single helping
category so that there were two categories:
helpers and nonhelpers.3 The third analysis
was used to determine which factors caused
bystanders to choose a specific helping strat-
egy. Therefore the three different types of
help (direct, indirect, and ineffective) were
the three dependent categories in this analysis.

Overall model. The “overall analysis,” or
the analysis in which we used all four response
categories, yielded a model of best fit con-
taining the following four main effects: (a)
the presence of confederate bystanders,4 (b)
amibiguity, (c) sex, and (d) training: x2(33,
N = 163) = 25.38, p > .83.5 An alternative
which was similar to the accepted model but
differed because it contained the interaction
of ambiguity and sex was rejected. The inter-
action did not significantly improve the fit of
the model, G2(3) = 7.41, and the inclusion
of the term would have created a less parsi-
monious model. The accepted model provided
an excellent fit to the data.

Helper versus nonhelper model. The anal-
ysis used to differentiate people who made
an attempt to help from those who did not
yielded a model of best fit that contained the
main effects of (a) ambiguity, (b) sex, and (c)
the presence of bystanders x2(12, N =
163) = 17.87, p > .12. Although the model
is acceptable, the fit is poor, indicating the

involvement of other unspecified factors in
the model (e.g., individual differences). Again,
an interaction of sex and ambiguity did not
improve the fit of the model over the main

Helper model. The analysis that was used
to differentiate the three different attempts to
help contained only one variable: training.

The accepted model provided an excellent fit
to the data, x2(28, N = 98) = 17.91, p > .92.

3 The coding system that is most frequently used in
bystander intervention research is a binary system; the
person is coded as either helping or not helping. In
general, when data is analyzed and reported, it is the
percentage of people that help that is presented. To
analyze and then report the rate of nonhelping would
provide totally redundant information and therefore is
not done. In the coding system used in this study, the
normal helping response is divided into three categories.
Results that are equivalent to a standard binary analysis
are presented in the discussion of variables that affect
the rate of nonresponding. Hypothesis-testing statistics
should be equivalent, and one can obtain the rate of
responding by subtracting the proportion of nonresponders
from 1.0.

4 The partial association tables indicated that the num-
ber of bystanders present was only marginally significant
and could be included or excluded from either the overall
or helper-versus-nonhelper models. Given the over-
whelming support in the literature for the presence of
the effect (see Latane, Nida, & Wilson, 1981), we included
it in both models.

‘ All chi-squares associated with the multiway contin-
gency table analyses are likelihood ratio chi-squares. G2

is a likelihood ratio chi-square and is an indication of
the relative fit of the model.


We now provide a description of the effects
of the independent variables on each of the
four response categories in turn. Any discrep-
ancies between the results derived from each
model are described and discussed.


Nonhelping response. Approximately 40%
of the subjects did not respond to the emer-
gency. Of these individuals, 97% did not
explore the reasons for the sound (two non-
responders had explored).

The helpers-versus-nonhelpers analysis in-
dicated that the level of ambiguity in the
emergency was important in determining the
rate of intervention. More nonresponders were
in the ambiguous situation (58%) than were
in the nonambiguous situation (22%; X =
4.65, p < .001). Female subjects (50%) were more frequent nonresponders (50%) than were male subjects (24%; X = 3.39, 'p < .001).

The presence of other bystanders tended
to increase the rate of nonresponding. Subjects
with confererate bystanders were nonrespon-
sive 48% of the time in comparison with lone
subjects’ rate of 33% (X = 1.73, p < .09). These results are consistent with the eifects found in the overall analysis, the major dif- ference being that the trend associated with lone bystanders reached statistical significance (X = 2.00, p < .05).

Ineffective direct response. Approximately
8% of the subjects responded by approaching
and touching the victim, but not making a
correct first-aid response and not seeking
other help. Of the subjects who were alone,
13% responded with ineffective direct help in
comparison with 2.5% of subjects who were
with other bystanders. This effect is estimated
to be only marginally significant as estimated
from both the analysis of helpers (X = 1.53,
p < ,13) and the overall analysis (X = 1.84, p < .07). In addition, as we mentioned pre- viously, the variable did not add significantly to the fit of the model that differentiated among the three helping categories. However, there is more support for an effect of the number of bystanders on the mode of help; we present evidence of this support in the Discussion section.

Indirect response. Approximately 30% of
subjects either yelled for help or searched for

outside intervention without providing ade-
quate first aid. These two responses occurred
with approximately equal frequency. These
individuals were apparently searching for
someone who might be able to help. To
confederate bystanders, they frequently re-
sponded, “Call an ambulance!”, “Get some
help!”, “This guy’s hurt!”, and so on.

The only variable that distinguished indi-
rect responders from the other helping re-
sponders was training. Indirect responses oc-
curred more frequently among untrained
(42%) rather than trained subjects (18%; X =
3.37, p< .001).

A discrepancy between the overall analysis
and the analysis of the helping responses was
that the overall analysis indicated that sex
and ambiguity were significantly related to
indirect helping. These effects are artifacts of
the statistical dependency between the four
response categories. If people in the ambig-
uous condition and women are frequently
among the nonresponders, there are fewer
women helpers and helpers in the ambiguous
condition left to respond indirectly or in any
other fashion. When one drops nonhelpers
from the analysis and reconditionalizes the
data, ambiguity and sex do not differentiate
among direct, indirect, or ineffective helpers.

Direct pressure response. Twenty-two per-
cent of the subjects responded with a direct
pressure response, the prescribed method of
controlling arterial bleeding. The only variable
that separated direct pressure responders from
other helpers was training. Only 5% of the
subjects in the untrained condition responded
in this fashion, whereas 39% of subjects in
the trained group responded with direct pres-
sure (X = 4.31, p < .001, helper analysis). Over 70% of the subjects who applied direct pressure also called for help, the most desir- able response and the response taught in training.


The four helping responses just illustrated
are not unique to the literature. Three of
these helping responses were observed in
previous investigations, and a similar pattern
of four responses (if “passive helping” is
equivalent to “ineffective helping”) was found
with a simulated rape (cf. Shetland & Steb-


bins, 1980). What is the difference between
the research in which two categories of helping
are found and the research in which either
three or four are found? On the basis of the
“Limited Altruism” model of Shetland and
Stebbins (1983, p. 109) and the results of this
study, we conclude that the lack of knowledge
of how to solve the problem directly is one
cause of indirect helping. As Piliavin et al.
(1981) and Shetland and Stebbins (1980)
suggested, studies that obtain indirect helping
tend to involve more severe emergencies than
do the ones that obtain two responses. Shot-
land and Huston (1979) found that both a
“rape in progress” and “a cut artery” were
rated by subjects as “definite” emergencies.
In the studies cited by Piliavin et al., seizures,
unconscious victims, physical attacks, thefts,
and so on, were portrayed. In most laboratory
simulations yielding two responses, research-
ers use situations that are not as severe. The
more severe the emergency is, the more likely
it is that bystanders do not know or are not
confident enough to use needed specialized
knowledge, and thus they resort to indirect
helping. An ineffective direct response may
also occur partly for the same reason. How-
ever, as we more fully describe later, this
response also occurs because some avenues
of indirect helping are also closed off to these

Obviously, the complexity of the dependent
variable(s) also depends on when the investi-
gator terminates the bystanders’ responses. If
subjects are stopped before they can try to
help, valuable data are lost. It seems that if
helping behavior researchers explore the full
range of bystander responses, more could be
learned about both the dynamics of bystander
behavior, the decision-making process, and
the individual differences that lead to differ-
ential responding.

One of the most striking results of this
study is the efficaciousness of training. Put
in its starkest context, with a real injury of
the magnitude of the simulation, the victim
could have died without first aid in 4-10
min. In the untrained group, only 4 of 83
victims had a reasonable chance of survival.
In the trained group, 32 of 80 victims had
that chance. Thus the effect of training had
a net value of saving as many as 28 lives.
Training appears to be an important practical

Twenty-eight more subjects in the trained
group than in the untrained group used
direct pressure in their response. Before
training, what type of response would these
subjects have made? Clearly, most of the
direct-pressure responders before training
would have been indirect helpers, the other
response type affected by training. Because
training caused 28 more direct-pressure re-
sponses, it also caused 21 fewer indirect re-

In other words, skills training seems to
significantly change not the rate of responding
but rather the types of responses of people
who try to help. Thus skills training affects
the decision concerning how to help rather
than the decision of whether to help. One
might ask why is it that indirect responders
are affected by training.

Perhaps indirect responders felt, in Schwartz’s
(1977) terms, a “moral obligation” to help.
Perhaps this moral obligation was lacking
among many of the nonresponders, who
showed fewer signs of arousal and concern.
Apparently, nonresponding does not primarily
result from a lack of knowledge of the correct
course of action, but indirect helping does.
Thus training raises the rate of appropriate
behavior by teaching skills to people who are
motivated to help. Training for an appropriate
response does not seem to significantly change
the rate of responding of those not motivated
to help.

The rate of nonresponding was affected by
three main effects: sex, ambiguity, and the
number of bystanders. The sound of the
emergency indicated that the workman could
have been seriously injured, although alter-
native explanations were still possible, for the
emergency was heard but not seen, at least
initially. The cry for help provided further
evidence that the worker needed help; thus
more subjects chose to explore or try to help.
Women may have helped less frequently be-
cause they felt less competent than men to
face a victim with a serious injury. Perhaps
women feel less adept at moving heavy fallen
objects, or are socialized to take less initiative
unless the situation is unambiguous (see
Deaux, 1976). The explanations for the im-
pact of the number of bystanders on the rate
of helping are well known (cf. Latane &
Darley, 1969; Latane, Nida, & Wilson, 1981)
and do not need further discussion.


The results suggest that lone bystanders
were more likely to be ineffective direct help-
ers than were bystanders who were in the
presence of others. Why should the presence
or absence of other bystanders affect the
methods of help?

The easiest explanation is that when by-
standers are alone, they are less likely to
think that others who possess greater first aid
knowledge are immediately available. There-
fore, the responsibility is theirs to do the best
they can without the aid of others; that is,
no diffusion of responsibility can take place
(Latane & Darley, 1969). In this case the best
they could do was not good enough because
either they had no training or they did not
remember their training.

If other bystanders had been present, we
would have expected many more ineffective
helpers to have resorted to indirect helping
because these ineffective helpers would have
known that calling for help would have at
least been heard. In fact, some of the indirect
helpers did, before calling for help, produce
some direct ineffective responses. Additional
support for this expectation was provided
from 10 direct-pressure helpers who did not
include a call for help, which they were
trained to do. This lapse is perfectly reason-
able once it is understood that 9 of these 10
bystanders were alone. Moreover, as we pre-
viously mentioned, approximately half of the
indirect helpers shouted for help, whereas the
other half searched for help. As one would
expect, 70% of those who shouted were in
groups, whereas 78% of those who searched
for others to help were alone (Z = 3.84, p < .001).6 In summary, a bystander who is alone is less likely to consider at least one form of indirect responding, namely, calling out for help. Instead, bystanders may do the best they can by themselves. Thus in a certain percentage of the time, ineffective direct help- ing occurs.

Our data supports the conclusion of
Schwartz and Gottlieb (1970, 1980) that the
number of bystanders affects the interpreta-
tion of the event as well as the decision
concerning whether to help. In addition, our
data would support the proposition that if a
bystander is alone, it affects the decision
concerning how to help as well.

We conclude therefore, that the ambiguity
and sex variables affect the decision of whether

to help, but not how to help. (With other
types of emergencies, it seems reasonable that
sex may have different effects on whether and
how one helps; cf. Deaux, 1976.) Skills train-
ing affects the decision of how to help but
not that of whether to help, whereas the
presence of other bystanders affects both the
decision concerning whether to help and the
decision concerning how to help.

We also note that 98% of the subjects who
entered the foyer attempted to help the victim.
This is an indication that “the approach,”
the ubiquitous dependent variable in by-
stander research, is a valid measure of the
intent to help. However, our data would
suggest that the intent to help is not synon-
ymous with having actually provided aid. If
our victim required help within 4 to 10 min,
only 37% of the victims in which bystanders
had tried to help would have had a good
chance at survival because of the use of a
direct-pressure response. Indirect helping may
or may not have been successful, depending
on the time required to obtain a solution.

We would be remiss to end without dis-
cussing the potential threats to the validity
of our conclusions concerning the training
variable. As should be clear from our descrip-
tion of the method, the training variable was
tested in a pretest/posttest manner without a
control group. Do problems with this design
affect the validity of the conclusions drawn
from the training variable? As Campbell and
Stanley (1963) and Cook and Campbell
(1979) point out, this type of design is open
to alternative explanations because of threats
to internal validity. Threats that are due to
mortality and history appear to be relevant
to this study. Mortality does not seem to be
a threat because there was no differential loss
of subjects in comparing the attrition rates
from the pre- and posttraining groups in
terms of subjects either signing up for or
showing up for appointments. In addition,
each of the 163 subjects with usable data had
completed the course. Therefore, it appears
that the randomization of subjects was not
disturbed because of subject attrition.

History potentially poses a problem in that
there was an assassination attempt on Presi-

6 The statistical comparison was between the 78% of
those who were alone and searched for another helper
and the 30% of those in groups who chose that strategy.


dent Reagen in the period between the run-
ning of the pre- and posttraining groups.
However, the impact of this event does not
affect the general overall rate of responding
(pretraining 57%, posttraining 64%), the basic
type of response one might expect to be
affected by such an event. In addition, no
decrease resulted in the rate of nonresponding,
either as a main effect of training or as an
interaction between training and some other
variable. In fact, the basic changes we did
discover were changes in the type of reaction
responders took. Nothing surrounding the
assassination attempt or any other historical
event appears to have the capability to have
taught the proper direct pressure response or
a decrease in indirect helping. The same
evidence can be used to indicate that knowl-
edge of the experiment did not spread through
the class. We would have expected both the
rate and the effectiveness of the response to
have improved if information had been dif-
fused. We uncovered no indications that
knowledge about the experiment might have
been diffused.


Ashton, N. L., & Severy, L. J. (1976). Arousal and costs
in bystander intervention. Personality and Social Psy-
chology Bulletin, 2, 268-272.

Beaman, A. L., Barnes, P. J., Klentz B., & McQuirk, B.
(1978). increasing helping rales through information
dissemination. Personality and Social Psychology Bul-
letin, 4, 406-411.

Bishop, Y. M., Fienberg, S. E., & Holland, P. W. (with
the collaboration of R. J. Light & F. Mosteller). (1975).
Discrete multivariale analysis: Theory and practice.

Cambridge, MA: MIT Press.

Campbell, D. T., & Stanley, J. C. (1963). Experimental
and quasi-experimental designs for research. Chicago:
Rand McNally.

Clark, R. D., Ill, & Word, L. E. (1972). Why don’t
bystanders help? Because of ambiguity? Journal of
Personality and Social Psychology, 24, 392-400.

Clark, R. D., Ill, & Word, L E. (1974). Where is the

apathetic bystander? Situational characteristics of the
emergency. Journal of Personality and Social Psychology,
29, 279-287.

Cook, T. D., & Campbell, D. T. (1979). Quasi-experi-
mentation design and analysis issues for field settings.

Chicago: Rand McNally.
Crampton, R. S., Aldrich, R. E, Gascho, J. A., Miles,

J. R., & Stillerman, R. (1975). Reduction of prehospital,
ambulance and community coronary death rates by
the community-wide emergency cardiac care system.
The American Journal of Medicine, 58, 151-165.

Deaux, K. (1976). The behavior of women and men.
Monterey, CA: Brooks/Cole.

Eisenberg, M. S., Bergner, L., & Hallstrom, A. (1979).
Cardiac resuscitation in tlte community: Importance
of rapid provision and implications for program plan-

ning. Journal of the American Medical Association,
241, 1905-1907.

Form, W. H., & Nosow, S. (1958). Community in disaster.
New York: Harper & Row.

Hafen, B. Q. (1981). First aid for health emergencies
(2nd ed.). SI. Paul, MN: West.

Huston, T., Ruggiero, M., Conner, E, R., & Geis, G.
(1981). Bystander intervention into crime: A study
based on naturally occurring episodes. Social Psychol-
ogy Quarterly, 44, 14-23.

Kazdin, A. E., & Bryan, J. H. (1971). Competence and
volunteering. Journal of Experimental Social Psychol-
ogy, 7, 87-97.

Latane, B., & Darley, J. M. (1969). Bystander “apathy.”
American Scientist, 57(2), 244-268.

Laiane, B., Nida, S., & Wilson, D. (1981), The effects of
group size on helping behavior. In J. P. Rushlon & R.
M. Sorrentino (Eds.). Altruism and helping behavior
(pp. 287-314). Hillsdale, NJ: Erlbaum.

Midlarsky, E. (1971). Aiding under stress: The effects of
competence, dependency, visibility, and fatalism. Jour-
nal of Personality, 39. 132-149.

Midlarsky, E., & Midlarsky, M. (1973). Some determinants
of aiding under experimentally induced stress. Journal
of Personality, 41, 305-327.

Pantin, H. M., & Carver, C. S. (1982). Induced compe-
tence and the bystander effect. Journal of Applied
Social Psychology, 12, 100-111.

Piliavin, J. A., & Piliavin, 1. M. (1972). Effect of blood
on reactions to a victim. Journal of Personality and
Social Psychology, 23, 353-361.

Piliavin, J. A., Dovidio, J-, Gaertner, S., & Clark, R. D.
(1981). Emergency intervention. New York: Academic

Schwartz, S. H. (1977). Normative influences on altruism.

In L. Berkowitz (Ed.), Advances in experimental social
psychology (NA 10, pp. 221-279). New York: Academic

Schwartz, S., & Ben David, A. (1976). Responsibility and
helping in an emergency: Effects of blame, ability and

denial of responsibility. Sociometry, 39, 406-415.
Schwartz, S., & Gottlieb, A. (1970). Bystander reactions

to a violent theft: Crime in Jerusalem. Journal of
Personality and Social Psychology, 34, 1188-1199.

Schwartz, S., & Gottlieb, A. (1980). Bystander anonymity
and reactions to emergencies. Journal of Personality

and Social Psychology, 39, 418-430,
Shotland, R. L. (1976). Spontaneous vigilantism: A by-

stander response to criminal behavior. In H. J. Rosen-
baum &. P. C. Sederberg (Eds.), Vigilante politics (pp.
30-44). Philadelphia: University of Pennsylvania Press.

Shotland, R. L., & Huston, T. L. (1979). Emergencies:
What are they and do they influence bystanders to
intervene? Journal of Personality and Social Psychology,
37, 1822-1834.

Shotland, R. L., & Stebbins, C. A. (1980). Bystander
response to rape: Can a victim attract help? Journal
of Applied Social Psychology, 10, 510-527.

Shotland, R. L., & Stebbins, C. A. (1983). Emergency
and cost as determinants of helping behavior and the

slow accumulation of social psychological knowledge.
Social Psychology Quarterly, 46, 36-46.

Yakimovich, D., & Saltz, E. (1971). Helping behavior:
The cry for help. Psychonomic Science, 23, 427-428.

Received June 22, 1983

Revision received September 4, 1984 •

Journal of Personality and Social Psychology
1968, Vol. 8, No. 4,






New York University


Columbia University

Ss overheard an epileptic seizure. They believed either that they alone heard
the emergency, or that 1 or 4 unseen others were also present. As predicted
the presence of other bystanders reduced the individual’s feelings of personal
responsibility and lowered his speed of reporting (p < .01). In groups of size 3, males reported no faster than females, and females reported no slower when the 1 other bystander was a male rather than a female. In general, personality and background measures were not predictive of helping. Bystander inaction in real-life emergencies is often explained by "apathy," "alienation," and "anomie." This experiment suggests that the explanation may lie more in the bystander's response to other observers than in his indifference to the victim.

Several years ago, a young woman was
stabbed to death in the middle of a street in
a residential section of New York City. Al-
though such murders are not entirely routine,
the incident received little public attention
until several weeks later when the New York
Times disclosed another side to the case: at
least 38 witnesses had observed the attack—
and none had even attempted to intervene.
Although the attacker took more than half
an hour to kill Kitty Genovese, not one of
the 38 people who watched from the safety
of their own apartments came out to assist
her. Not one even lifted the telephone to
call the police (Rosenthal, 1964).

Preachers, professors, and news commenta-
tors sought the reasons for such apparently
conscienceless and inhumane lack of interven-
tion. Their conclusions ranged from “moral
decay,” to “dehumanization produced by
the urban environment,” to “alienation,”
“anomie,” and “existential despair.” An anal-
ysis of the situation, however, suggests that
factors other than apathy and indifference
were involved.

A person witnessing an emergency situa-
tion, particularly such a frightening and

1 This research was supported in part by National
Science Foundation Grants GS1238 and GS1239.
Susan Darley contributed materially to the design
of the experiment and ran the subjects, and she and
Thomas Moriarty analyzed the data. Richard
Nisbett, Susan Millman, Andrew Gordon, and
Norma Neiman helped in preparing the tape

dangerous one as a stabbing, is in conflict.
There are obvious humanitarian norms about
helping the victim, but there are also rational
and irrational fears about what might happen
to a person who does intervene (Milgram &
Hollander, 1964). “I didn’t want to get
involved,” is a familiar comment, and behind
it lies fears of physical harm, public embar-
rassment, involvement with police procedures,
lost work days and jobs, and other unknown

In certain circumstances, the norms favor-
ing intervention may be weakened, leading
bystanders to resolve the conflict in the direc-
tion of nonintervention. One of these circum-
stances may be the presence of other on-
lookers. For example, in the case above, each
observer, by seeing lights and figures in other
apartment house windows, knew that others
were also watching. However, there was no
way to tell how the other observers were
reacting. These two facts provide several
reasons why any individual may have delayed
or failed to help. The responsibility for help-
ing was diffused among the observers; there
was also diffusion of any potential blame for
not taking action; and finally, it was possible
that somebody, unperceived, had already
initiated helping action.

When only one bystander is present in an
emergency, if help is to come, it must come
from him. Although he may choose to ignore
it (out of concern for his personal safety,
or desires “not to get involved”), any pres-



sure to intervene focuses uniquely on him.
When there are several observers present,
however, the pressures to intervene do not
focus on any one of the observers; instead
the responsibility for intervention is shared
among all the onlookers and is not unique
to any one. As a result, no one helps.

A second possibility is that potential blame
may be diffused. However much we may wish
to think that an individual’s moral behavior
is divorced from considerations of personal
punishment or reward, there is both theory
and evidence to the contrary (Aronfreed,
1964; Miller & Bollard, 1941, Whiting &
Child, 19S3). It is perfectly reasonable to
assume that, under circumstances of group
responsibility for a punishable act, the pun-
ishment or blame that accrues to any one
individual is often slight or nonexistent.

Finally, if others are known to be present,
but their behavior cannot be closely observed,
any one bystander can assume that one of
the other observers is already taking action
to end the emergency. Therefore, his own
intervention would be only redundant—per-
haps harmfully or confusingly so. Thus,
given the presence of other onlookers whose
behavior cannot be observed, any given by-
stander can rationalize his own inaction by
convincing himself that “somebody else must
be doing something.”

These considerations lead to the hypothesis
that the more bystanders to an emergency,
the less likely, or the more slowly, any one
bystander will intervene to provide aid. To
test this propostion it would be necessary to
create a situation in which a realistic
“emergency” could plausibly occur. Each sub-
ject should also be blocked from com-
municating with others to prevent his getting
information about their behavior during the
emergency. Finally, the experimental situa-
tion should allow for the assessment of the
speed and frequency of the subjects’ reaction
to the emergency. The experiment reported
below attempted to fulfill these conditions.


Overview. A college student arrived in the labora-
tory and was ushered into an individual room from
which a communication system would enable him
to talk to the other participants. It was explained
to him that he was to take part in a discussion

about personal problems associated with college life
and that the discussion would be held over the
intercom system, rather than face-to-face, in order
to avoid embarrassment by preserving the anonym-
ity of the subjects. During the course of the dis-
cussion, one of the other subjects underwent what
appeared to be a very serious nervous seizure simi-
lar to epilepsy. During the fit it was impossible for
the subject to talk to the other discussants or to
find out what, if anything, they were doing about
the emergency. The dependent variable was the
speed with which the subjects reported the emer-
gency to the experimenter. The major independent
variable was the number of people the subject
thought to be in the discussion group.

Subjects. Fifty-nine female and thirteen male stu-
dents in introductory psychology courses at New
York University were contacted to take part in an
unspecified experiment as part of a class requirement.

Method. Upon arriving for the experiment, the
subject found himself in a long corridor with doors
opening off it to several small rooms. An experi-
mental assistant met him, took him to one of the
rooms, and seated him at a table. After filling out a
background information form, the subject was given
a pair of headphones with an attached microphone
and was told to listen for instructions.

Over the intercom, the experimenter explained that
he was interested in learning about the kinds of
personal problems faced by normal college students
in a high pressure, urban environment. He said
that to avoid possible embarrassment about dis-
cussing personal problems with strangers several
precautions had been taken. First, subjects would
remain anonymous, which was why they had been
placed in individual rooms rather than face-to-face.
(The actual reason for this was to allow tape
recorder simulation of the other subjects and the
emergency.) Second, since the discussion might be
inhibited by the presence of outside listeners, the
experimenter would not listen to the initial discus-
sion, but would get the subject’s reactions later, by
questionnaire. (The real purpose of this was to
remove the obviously responsible experimenter from
the scene of the emergency.)

The subjects were told that since the experimenter
was not present, it was necessary to impose some
organization. Each person would talk in turn, pre-
senting his problems to the group. Next, each person
in turn would comment on what the others had said,
and finally, there would be a free discussion. A
mechanical switching device would regulate this dis-
cussion sequence and each subject’s microphone
would be on for about 2 minutes. While any micro-
phone was on, all other microphones would be off.
Only one subject, therefore, could be heard over
the network at any given time. The subjects were
thus led to realize when they later heard the seizure
that only the victim’s microphone was on and that
there was no way of determining what any of the
other witnesses were doing, nor of discussing the
event and its possible solution with the others.
When these instructions had been given, the discus-
sion began.


In the discussion, the f u t u r e victim spoke first,
saying that he found it difficult to get adjusted to
New York City and lo his studies. Very hesitantly,
and with obvious embarrassment, he mentioned that
he was prone to seizures, particularly when studying
hard or taking exams. The other people, including
the real subject, took their turns and discussed
similar problems (minus, of course, the proneness to
seizures). The naive subject talked last in the
series, after the last prerecorded voice was played.2

When it was again the victim’s turn to talk, he
made a few relatively calm comments, and then,
growing increasingly louder and incoherent, he

I-er-um-I think I-I necd-er-if-if could-er-er-some-
body er-er-er-er-er-er-er give me a liltle-er-give
me a little help here because-er-I-er-I’m-er-er-
h-h-having a-a-a real problcm-er-right now and
I-er-if somebody could help me out it would-it
would-er-er s-s-sure be-sure be good . . . because-
cr-there-er-cr-a cause I-er-I-uh-I’ve got a-a one of
the-er-sei er-cr-things coming on and-and-and
I could really-er-use some help so if somebody
would-er-give me a little h-help-uh-er-er-er-er-er
c-could somebody-er-er-help-er-uh-uh-uh (choking
sounds). . . . I’m gonna die-er-er-I’m . . . gonna
die-er-help-er-er-seizure-er-[chokes, then quiet].

The experimenter began timing the speed of the
real subject’s response at the beginning of the vic-
tim’s speech. Informed judges listening to the tape
have estimated that the victim’s increasingly louder
and more disconnected ramblings clearly repre-
sented a breakdown about 70 seconds after the
signal for the victim’s second speech. The victim’s
speech was abruptly cut off 125 seconds after this
signal, which could be interpreted by the subject
as indicating that the time allotted for that speaker
had elapsed and the switching circuits had switched
away from him. Times reported in the results are
measured from the start of the fit.

Group size variable. The major independent vari-
able of the study was the number of other people
that the subject believed also heard the fit. By the
assistant’s comments before the experiment, and also
by the number of voices heard to speak in the first
round of the group discussion, the subject was led
lo believe that the discussion group was one of three
sizes: either a two-person group (consisting of a
person who would later have a fit and the real
subject), a three-person group (consisting of the
victim, the real subject, and one confederate voice),
or a six-person group (consisting of the victim, the
real subject, and four confederate voices). All the
confederates’ voices were tape-recorded.

Variations in group composition. Varying the kind
as well as the number of bystanders present at an

2 To test whether the order in which the subjects
spoke in the first discussion round significantly af-
fected the subjects’ speed of report, the order in
which the subjects spoke was varied (in the six-
person group). This had no significant or noticeable
effect on the speed of the subjects’ reports.

emergency should also vary the amount of respon-
sibility felt by any single bystander. To test this,
several variations of the three-person group were
r u n . In one three-person condition, the taped by-
stander voice was that of a female, in another a
male, and in the third a male who said that he was
a premedical student who occasionally worked in
the emergency wards at Bellevue hospital.

In the above conditions, the subjects were female
college students. In a final condition males drawn
from the same introductory psychology subject pool
were tested in a three-person female-bystander

Time to help. The major dependent variable was
the time elapsed from the start of the victim’s fit
until the subject left her experimental cubicle. When
the subject left her room, she saw the experimental
assistant seated at the end of the hall, and invari-
ably went to the assistant. If 6 minutes elapsed
without the subject having emerged from her room,
the experiment was terminated.

As soon as the subject reported the emergency,
or after 6 minutes had elapsed, the experimental
assistant disclosed the true nature of the experi-
ment, and dealt with any emotions aroused in the
subject. Finally the subject filled out a questionnaire
concerning her thoughts and feelings during the
emergency, and completed scales of Machiavellian-
ism, anomie, and authoritarianism (Christie, 1964),
a social desirability scale (Crowne & Marlowe, 1964),
a social responsibility scale (Daniels & Berkowitz,
1964), and reported vital statistics and socioeconomic

Plausibility of Manipulation

Judging by the subjects’ nervousness when
they reported the fit to the experimenter, by
their surprise when they discovered that the
fit was simulated, and by comments they
made during the fit (when they thought
their microphones were off), one can con-
clude that almost all of the subjects perceived
the fit as real. There were two exceptions in
different experimental conditions, and the
data for these subjects were dropped from
the analysis.

Effect of Group Size on Helping

The number of bystanders that the sub-
ject perceived to be present had a major ef-
fect on the likelihood with which she would
report the emergency (Table 1). Eighty-five
percent of the subjects who thought they
alone knew of the victim’s plight reported
the seizure before the victim was cut off, only
31% of those who thought four other by-
standers were present did so.




Group size

2 (5 & victim)
3 (S, victim, & 1 other)
6 (.9, victim, & 4 others)



% responding
by end of fit


in sec.







Note.—p value of diffciences: x
2 = 7.91, p < .02; 7'' = 8.09,

p < .01, for speed scores.

Every one of the subjects in the two-
person groups, but only 62% of the subjects
in the six-person groups, ever reported the
emergency. The cumulative distributions of
response times for groups of different per-
ceived size (Figure 1) indicates that, by
any point in time, more subjects from the
two-person groups had responded than from
the three-person groups, and more from the
three-person groups than from the six-person

Ninety-five percent of all the subjects who
ever responded did so within the first half
of the time available to them. No subject
who had not reported within 3 minutes after
the fit ever did so. The shape of these dis-
tributions suggest that had the experiment
been allowed to run for a considerably longer
time, few additional subjects would have

Speed of Response

To achieve a more detailed analysis of the
results, each subject’s time score was trans-


12o 16O 2oo 24O 28O

Seconds from Beginning of Fit

FIG. 1. Cumulative distributions of helping responses.

formed into a “speed” score by taking the
reciprocal of the response time in seconds
and multiplying by 100. The effect of this
transformation was to deemphasize differ-
ences between longer time scores, thus re-
ducing the contribution to the results of the
arbitrary 6-minute limit on scores. A high
speed score indicates a fast response.

An analysis of variance indicates that the
effect of group size is highly significant
(/> < .01). Duncan multiple-range tests indi- cate that all but the two- and three-person groups differ significantly from one another ( # < . O S ) .

Victim’s Likelihood of Being Helped

An individual subject is less likely to re-
spond if he thinks that others are present.
But what of the victim? Is the inhibition
of the response of each individual strong
enough to counteract the fact that with five
onlookers there are five times as many people
available to help? From the data of this
experiment, it is possible mathematically to
create hypothetical groups with one, two, or
five observers.8 The calculations indicate that
the victim is about equally likely to get help
from one bystander as from two. The victim
is considerably more likely to have gotten
help from one or two observers than from
five during the first minute of the fit. For
instance, by 45 seconds after the start of the
fit, the victim’s chances of having been helped
by the single bystanders were about 50%,
compared to none in the five observer condi-
tion. After the first minute, the likelihood of
getting help from at least one person is high
in all three conditions.

Effect of Group Composition on Helping the

Several variations of the three-person group
were run. In one pair of variations, the
female subject thought the other bystander
was either male or female; in another, she
thought the other bystander was a premedical
student who worked in an emergency ward
at Bellevue hospital. As Table 2 shows, the

8 The formula for the probability that at least one
person will help by a given time is 1 — ( 1 — P ) ”
where n is the number of observers and P is the
probability of a single individual (who thinks he is
one of n observers) helping by that time.




Group composition

Female S, male other
Female S, female other
Female 5, male medic

Male S, female other




% responding
by end of fit



in sec.






» Three-person group, mule victim.

variations in sex and medical competence
of the other bystander had no important or
detectable affect on speed of response. Sub-
jects responded equally frequently and fast
whether the other bystander was female,
male, or medically experienced.

Sex of the Subject and Speed of Response

Coping with emergencies is often thought
to be the duty of males, especially when
females are present, but there was no evi-
dence that this was the case in this study.
Male subjects responded to the emergency
with almost exactly the same speed as did
females (Table 2 ) .

Reasons for Intervention or Nonintervention

After the debriefing at the end of the ex-
periment each subject was given a 15-item
checklist and asked to check those thoughts
which had “crossed your mind when you
heard Subject 1 calling for help.” Whatever
the condition, each subject checked very few
thoughts, and there were no significant dif-
ferences in number or kind of thoughts in the
different experimental groups. The only
thoughts checked by more than a few sub-
jects were “I didn’t know what to do” (18
out of 65 subjects), “I thought it must be
some sort of fake” (20 out of 65), and “I
didn’t know exactly what was happening”
(26 out of 65).

It is possible that subjects were ashamed
to report socially undesirable rationalizations,
or, since the subjects checked the list after
the true nature of the experiment had been
explained to them, their memories might
have been blurred. It is our impression, how-
ever, that most subjects checked few reasons
because they had few coherent thoughts
during the fit.

We asked all subjects whether the presence
or absence of other bystanders had entered
their minds during the time that they were
hearing the fit. Subjects in the three- and
six-person groups reported that they were
aware that other people were present, but
they felt that this made no difference to their
own behavior.

Individual Difference Correlates of Speed of

The correlations between speed of report
and various individual differences on the per-
sonality and background measures were
obtained by normalizing the distribution of
report speeds within each experimental condi-
tion and pooling these scores across all con-
ditions (« = 62-65). Personality measures
showed no important or significant correla-
tions with speed of reporting the emergency.
In fact, only one of the 16 individual dif-
ference measures, the size of the community
in which the subject grew up, correlated
(r = -.26, p < .05) with the speed of helping.


Subjects, whether or not they intervened,
believed the fit to be genuine and serious.
“My God, he’s having a fit,” many subjects
said to themselves (and were overheard via
their microphones) at the onset of the fit.
Others gasped or simply said “Oh.” Several
of the male subjects swore. One subject said
to herself, “It’s just my kind of luck, some-
thing has to happen to me!” Several subjects
spoke aloud of their confusion about what
course of action to take, “Oh God, what
should I do?”

When those subjects who intervened
stepped out of their rooms, they found the
experimental assistant down the hall. With
some uncertainty, but without panic, they
reported the situation. “Hey, I think Num-
ber 1 is very sick. He’s having a fit or
something.” After ostensibly checking on the
situation, the experimenter returned to report
that “everything is under control.” The sub-
jects accepted these assurances with obvious

Subjects who failed to report the emer-
gency showed few signs of the apathy and


indifference thought to characterize “unre-
sponsive bystanders.” When the experimenter
entered her room to terminate the situation,
the subject often asked if the victim was
“all right.” “Is he being taken care of?”
“He’s all right isn’t he?” Many of these sub-
jects showed physical signs of nervousness;
they often had trembling hands and sweating
palms. If anything, they seemed more emo-
tionally aroused than did the subjects who
reported the emergency.

Why, then, didn’t they respond? It is our
impression that nonintervening subjects had
not decided not to respond. Rather they were
still in a state of indecision and conflict con-
cerning whether to respond or not. The
emotional behavior of these nonresponding
subjects was a sign of their continuing
conflict, a conflict that other subjects resolved
by responding.

The fit created a conflict situation of the
avoidance-avoidance type. On the one hand,
subjects worried about the guilt and shame
they would feel if they did not help the
person in distress. On the other hand, they
were concerned not to make fools of them-
selves by overreacting, not to ruin the on-
going experiment by leaving their intercom,
and not to destroy the anonymous nature of
the situation which the experimenter had
earlier stressed as important. For subjects in
the two-person condition, the obvious distress
of the victim and his need for help were so
important that their conflict was easily re-
solved. For the subjects who knew there were
other bystanders present, the cost of not help-
ing was reduced and the conflict they were in
more acute. Caught between the two nega-
tive alternatives of letting the victim continue
to suffer or the costs of rushing in to help,
the nonresponding bystanders vacillated be-
tween them rather than choosing not to
respond. This distinction may be academic
for the victim, since he got no help in either
case, but it is an extremely important one
for arriving at an understanding of the
causes of bystanders’ failures to help.

Although the subjects experienced stress
and conflict during the experiment, their
general reactions to it were highly positive.
On a questionnaire administered after the
experimenter had discussed the nature and

purpose of the experiment, every single sub-
ject found the experiment either “interesting”
or “very interesting” and was willing to par-
ticipate in similar experiments in the future.
All subjects felt they understood what the
experiment was about and indicated that they
thought the deceptions were necessary and
justified. All but one felt they were better
informed about the nature of psychological
research in general.

Male subjects reported the emergency no
faster than did females. These results (or
lack of them) seem to conflict with the
Berkowitz, Klanderman, and Harris (1964)
finding that males tend to assume more re-
sponsibility and take more initiative than
females in giving help to dependent others.
Also, females reacted equally fast when the
other bystander was another female, a male,
or even a person practiced in dealing with
medical emergencies. The ineffectiveness of
these manipulations of group composition
cannot be explained by general insensitivity
of the speed measure, since the group-size
variable had a marked effect on report speed.

It might be helpful in understanding this
lack of difference to distinguish two general
classes of intervention in emergency situa-
tions: direct and reportorial. Direct interven-
tion (breaking up a fight, extinguishing a
fire, swimming out to save a drowner) often
requires skill, knowledge, or physical power.
It may involve danger. American cultural
norms and Berkowitz’s results seem to sug-
gest that males are more responsible than
females for this kind of direct intervention.

A second way of dealing with an emergency
is to report it to someone qualified to handle
it, such as the police. For this kind of inter-
vention, there seem to be no norms requiring
male action. In the present study, subjects
clearly intended to report the emergency
rather than take direct action. For such
indirect intervention, sex or medical com-
petence does not appear to affect one’s quali-
fications or responsibilities. Anybody, male
or female, medically trained or not, can find
the experimenter.

In this study, no subject was able to tell
how the other subjects reacted to the fit.
(Indeed, there were no other subjects actu-
ally present.) The effects of group size on


speed of helping, therefore, are due simply to
the perceived presence of others rather than
to the influence of their actions. This means
that the experimental situation is unlike
emergencies, such as a fire, in which by-
standers interact with each other. It is, how-
ever, similar to emergencies, such as the
Genovese murder, in which spectators knew
others were also watching but were pre-
vented by walls between them from com-
munication that might have counteracted the
diffusion of responsibility.

The present results create serious difficul-
ties for one class of commonly given explana-
tions for the failure of bystanders to inter-
vene in actual emergencies, those involving
apathy or indifference. These explanations
generally assert that people who fail to inter-
vene are somehow different in kind from the
rest of us, that they are “alienated by indus-
trialization,” “dehumanized by urbanization,”
“depersonalized by living in the cold society,”
or “psychopaths.” These explanations serve
a dual function for people who adopt them.
First, they explain (if only in a nominal
way) the puzzling and frightening problem
of why people watch others die. Second, they
give individuals reason to deny that they
too might fail to help in a similar situation.

The results of this experiment seem to
indicate that such personality variables may
not be as important as these explanations
suggest. Alienation, Machiavellianism, ac-
ceptance of social responsibility, need for
approval, and authoritarianism are often
cited in these explanations. Yet they did not
predict the speed or likelihood of help. In

sharp contrast, the perceived number of
bystanders did. The explanation of bystander
“apathy” may lie more in the bystander’s
response to other observers than in presumed
personality deficiencies of “apathetic” indi-
viduals. Although this realization may force
us to face the guilt-provoking possibility that
we too might fail to intervene, it also suggests
that individuals are not, of necessity, “non-
interveners” because of their personalities.
If people understand the situational forces
that can make them hesitate to intervene,
they may better overcome them.


ARONFREED, J. The origin of self-criticism. Psycho-
logical Review, 1964, 71, 193-219.

fects of experimenter awareness and sex of subject
on reactions to dependency relationships. Sociom-
etry, 1964, 27, 327-329.

CHRISTIE, R. The prevalence of machiavellian orien-
tations. Paper presented at the meeting of the
American Psychological Association, Los Angeles,

CROWNE, D., & MARLOWE, D. The approval motive.
Now York: Wiley, 1964.

DANIELS, L., & BERKOWITZ, L. Liking and response
to dependency relationships. Human Relations,
1963, 16, 141-148.

MILGRAM, S., & HOLLANDER, P. Murder they heard.
Nation, 1964, 198, 602-604.

MILLER, N., & BOLLARD, J. Social learning and
imitation. New Haven: Yale University Press,

ROSENTHAL, A. M. Thirty-eight witnesses. New
York: McGraw-Hill, 1964.

WHITING, J. W. M., & CHILD, I. Child training and
personality. New Haven: Yale University Press,

(Received July 8, 1967)

Journal oj Personality and Social Psychology
1969, Vol. 13, No. 4,







University of Pennsylvani



Columbia University



University of Pennsylvania

A field experiment was performed to investigate the effect of several variables
on helping behavior, using the express trains of the New York 8th Avenu


Independent Subway as a laboratory on wheels. Four teams of students, each
one made up of a victim, model, and two observers, staged standard col-
lapses in which type of victim (drunk or ill), race of victim (black or white)
and presence or absence of a model were varied. Data recorded by observers
included number and race of observers, latency of the helping response and
race of helper, number of helpers, movement out of the “critical area,” and
spontaneous comments. Major findings of the study were that (a) an appar-
ently ill person is more likely to receive aid than is one who appears to be
drunk, (b) race of victim has little effect on race of helper except when the
victim is drunk, (c) the longer the emergency continues without help being
offered, the more likely it is that someone will leave the area of the emergency,
and (d) the expected decrease in speed of responding as group size increases—
the “diffusion of responsibility effect” found by Darley and Latane—does not
occur in this situation. Implications of this difference between laboratory and
field results are discussed, and a brief model for the prediction of behavior
in emergency situations is presented.

Since the murder of Kitty Genovese in
Queens, a rapidly increasing number of social
scientists have turned their attentions to the
study of the good Samaritan’s act and an
associated phenomenon, the evaluation of
victims by bystanders and agents. Some of
the findings of this research have been pro-
vocative and nonobvious. For example, there
is evidence that agents, and even bystanders,
will sometimes derogate the character of the
victims of misfortune, instead of feeling com-
passion (Berscheid & Walster, 1967; Lerner
& Simmons, 1966). Furthermore, recent find-

1 This research was conducted while the first au-
thor was at Columbia University as a Special Na-
tional Institute of Mental Health Research Fellow
under Grant 1-F3-MH-36, 328-01. The study was
partially supported by funds supplied by this grant
and partially by funds from National Science
Foundation Grant GS-1901 to the third author. The
authors thank Virginia Joy for allowing the experi-
mental teams to be recruited f rom her class, and
Percy Tannenbaum for his reading of the manuscript
and his helpful comments.

2 Requests for reprints should be sent to Jane
Allyn Piliavin, Department of Psychology, University
of Pennsylvania, 3813-1S Walnut Street, Philadelphia,
Pennsylvania 19104.

ings indicate that under certain circumstances
there is not “safety in numbers,” but rather
“diffusion of responsibility.” Darley and
Latane (1968) have reported that among by-
standers hearing an epileptic seizure over
earphones, those who believed other witnesses
were present were less likely to seek as-
sistance for the victim than were bystanders
who believed they were alone. Subsequent
research by Latane and Rodin (1969) on
response to the victim of a fall confirmed this
finding and suggested further that assistance
from a group of bystanders was less likely
to come if the group members were strangers
than if they were prior acquaintances. The
field experiments of Bryan and Test (1967),
on the other hand, provide interesting findings
that fit common sense expectations; namely,
one is more likely to be a good Samaritan
if one has just observed another individual
performing a helpful act.

Much of the work on victimization to date
has been performed in the laboratory. It is
commonly argued that the ideal research
strategy over the long haul is to move back
and forth between the laboratory, with its



advantage of greater control, and the field,
with its advantage of greater reality. The
present study was designed to provide more
information from the latter setting.

The primary focus of the study was on the
effect of type of victim (drunk or ill) and
race of victim (black or white) on speed of
responding, frequency of responding, and the
race of the helper. On the basis of the large
body of research on similarity and liking as
well as that on race and social distance, it
was assumed that an individual would be
more inclined to help someone of his race
than a person of another race. The expecta-
tion regarding type of victim was that help
would be accorded more frequently and
rapidly to the apparently ill victim. This
expectation was derived from two considera-
tions. First, it was assumed that people who
are regarded as partly responsible for their
plight would receive less sympathy and conse-
quently less help than people seen as not
responsible for their circumstances (Schopler
& Matthews, 1965).

Secondly, it was assumed that whatever
sympathy individuals may experience when
they observe a drunk collapse, their inclina-
tion to help him will be dampened by the
realization that the victim may become dis-
gusting, embarrassing, and/or violent. This
realization may, in fact, not only constrain
helping but also lead observers to turn away
from the victim—that is, to leave the scene
of the emergency.

Aside from examining the effects of race
and type of victim, the present research
sought to investigate the impact of modeling
in emergency situations. Several investigators
have found that an individual’s actions in a
given situation lead others in that situation
to engage in similar actions. This modeling
phenomenon has been observed in a variety
of contexts including those involving good
Samaritanism (Bryan & Test, 1967). It was
expected that the phenomenon would be
observed as well in the present study. A final
concern of the study was to examine the
relationship between size of group and fre-
quency and latency of the helping response,
with a victim who was both seen and heard.
In previous laboratory studies (Darley &
Latane, 1968; Latane & Rodin, 1969) in-

creases in group size led to decreases in fre-
quency and increases in latency of responding.
In these studies, however, the emergency was
only heard, not seen. Since visual cues are
likely to make an emergency much more
arousing for the observer, it is not clear that,
given these cues, such considerations as
crowd size will be relevant determinants of
the observer’s response to the emergency.
Visual cues also provide clear information as
to whether anyone has yet helped the victim
or if he has been able to help himself. Thus,
in the laboratory studies, observers lacking
visual cues could rationalize not helping by
assuming assistance was no longer needed
when the victim ceased calling for help.
Staging emergencies in full view of ob-
servers eliminates the possibility of such

To conduct a field investigation of the
above questions under the desired conditions
required a setting which would allow the re-
peated staging of emergencies in the midst
of reasonably large groups which remained
fairly similar in composition from incident to
incident. It was also desirable that each
group retain the same composition over the
course of the incident and that a reasonable
amount of time be available after the emer-
gency occurred for good Samaritans to act.
To meet these requirements, the emergencies
were staged during the approximately 7-J-
minute express run between the 59th Street
and 125th Street stations of the Eighth Ave-
nue Independent (IND) branch of the New
York subways.


About 4,450 men and women who traveled on the
8th Avenue IND in New York City, weekdays be-
tween the hours of 11:00 A.M. and 3:00 P.M. during
the period from April 15 to June 26, 1968, were
the unsolicited participants in this study. The racial
composition of a typical train, which travels through
Harlem to the Bronx, was about 45% black and
55% white. The mean number of people per car
during these hours was 43; the mean number of
people in the “critical area,” in which the staged
incident took place, was 8.5.

Field situation. The A and D trains of the 8th
Avenue IND were selected because they make no
stops between 59th Street and 125th Street. Thus,
for about H minutes there was a captive audience
who, after the first 70 seconds of their ride, became


Adjacent Area Critical Ares


Observer 2 — –











\ Victim

Mot or –
man’ s


FIG, I. Layout of adjacent and critical areas of subway car.

bystanders to an emergency situation. A single trial
was a nonstop ride between 59th and 125th Streets,
going in either direction. All trials were run only
on the old New York subway cars which serviced
the 8th Avenue line since they had two-person
seats in group arrangement rather than extended
seats. The designated experimental or critical area
was that end section of any car whose doors led
to the next car. There are 13 seats and some
standing room in this area on all trains (see
Figure 1),


On each trial a team of four Columbia General
Studies students, two males and two females, boarded
the train using different doors. Four different teams,
whose members always worked together, were used
to collect data for 103 trials. Each team varied the
location of the experimental car from trial to trial.
The female confederates took seats outside the criti-
cal area and recorded data as unobtrusively as pos-
sible for the duration of the ride, while the male
model and victim remained standing. The victim
always stood next to a pole in the center of the
critical area (see Figure 1). As the train passed the
first station (approximately 70 seconds after depart-
ing) the victim staggered forward and collapsed.
Until receiving help, the victim remained supine on
the floor looking at the ceiling. If the victim re-
ceived no assistance by the time the train slowed
to a stop, the model helped him to his feet. At the
stop, the team disembarked and waited separately
until other riders had left the station. They then
proceeded to another platform to board a train
going in the opposite direction for the next trial.
From 6 to 8 trials were run on a given day. AH
trials on a given day were in the same “victim

Victim. The four victims (one from each team)
were males between the ages of 26 and 35. Three
were white and one was black. All were identically
dressed in Eisenhower jackets, old slacks, and no tie.
On 38 trials the victims smelled of liquor and carried
a liquor bottle wrapped tightly in a brown bag
(drunk condition), while on the remaining 65 trials

they appeared sober and carried a black cane (cane
condition). In all other aspects, victims dressed and
behaved identically in the two conditions. Each
victim participated in drunk and cane trials.


Model. Four white males between the ages of



and 29 assumed the roles of model in each team.
All models wore informal clothes, although they were
not identically attired. There were four different
model conditions used across both victim conditions
(drunk or cane).

1. Critical area—early. Model stood in critical area
and waited until passing fourth station to assist
victim (approximately 70 seconds after collapse).

2. Critical area—late. Model stood in critical area
and waited until passing sixth station to assist victim
(approximately 150 seconds after collapse).

3. Adjacent area—early. Model stood in middle of
car in area adjacent to critical area and waited until
passing fourth station.

4. Adjacent area—late. Model stood in adjacent
area and waited until passing sixth station.

When the model provided assistance, he raised the
victim to a sitting position and stayed with him
for the remainder of the trial. An equal number of
trials in the no-model condition and in each of the
four model conditions were preprogrammed by a
random number table and assigned to each team.

3 It will be noted later that not only were there
more cane trials than drunk trials, they were also
distributed unevenly across black and white victims.
The reason for this is easier to explain than to cor-
rect. Teams 1 and 2 (both white victims) started
the first day in the cane condition. Teams 3 (black)
and 4 (white) began in the drunk condition. Teams
were told to alternate the conditions across days.
They arranged their running days to fit their sched-
ules. On their fourth day, Team 2 violated the
instruction and ran cane trials when they should
have run drunk trials; the victim “didn’t like” play-
ing the drunk! Then the Columbia student strike
occurred, the teams disbanded, and the study of
necessity was over. At this point, Teams 1 and 3
had run on only 3 days each, while 2 and 4 had
run on 4 days each.





No model
Number of trials run

Model trials
Number of trials run

Total number of trials

White victims











Black victim








Note.—Distribution of model trials for the drunk was as
follows: critical area: early, •!; late, 4; adjacent area: early, 5;
late, 3. The three model trials completed for the cane victim
were all early, with 2 from the critical area and 1 from the
adjacent area.

Measures. On each trial one observer noted the
race, sex, and location of every rider seated or stand-
ing in the critical area. In addition, she counted the
total number of individuals in the car and the total
number of individuals who came to the victim’s
assistance. She also recorded the race, sex, and loca-
tion of every helper. A second observer coded the
race, sex, and location of all persons in the adjacent
area. She also recorded the latency of the first
helper’s arrival after the victim had fallen and on
appropriate trials, the latency of the first helper’s
arrival after the programmed model had arrived.
Both observers recorded comments spontaneously
made by nearby passengers and attempted to elicit
comments from a rider sitting next to them.


As can be seen in Table 1, the frequency
of help received by the victims was impres-
sive, at least as compared to earlier laboratory
results. The victim with the cane received
spontaneous help, that is, before the model
acted, on 62 of the 65 trials. Even the drunk
received spontaneous help on 19 of 38 trials.
The difference is not explicable on the basis
of gross differences in the numbers of poten-
tial helpers in the cars. (Mean number of
passengers in the car on cane trials was
45; on drunk trials, 40. Total range was 15-

On the basis of past research, relatively
long latencies of spontaneous helping were
expected; thus, it was assumed that models
would have time to help, and their effects
could be assessed. However, in all but three
of the cane trials planned to be model trials,
the victim received help before the model was
scheduled to offer assistance. This was less

likely to happen with the drunk victim. In
many cases, the early model was able to
intervene, and in a few, even the delayed
model could act (see Table 1 for frequencies).

A direct comparison between the latency
of response in the drunk and cane conditions
might be misleading, since on model trials
one does not know how long it might have
taken for a helper to arrive without the
stimulus of the model. Omitting the model
trials, however, would reduce the number of
drunk trials drastically. In order to get
around these problems the trials have been
dichotomized into a group in which someone
helped bejore 70 seconds (the time at which
the early model was programmed to help)
and a group in which no one had helped by
this time. The second group includes some
trials in which people helped the model and
a very few in which no one helped at all.4

It is quite clear from the first section of
Table 2 that there was more immediate,
spontaneous helping of the victim with the
cane than of the drunk. The effect seems to
be essentially the same for the black victim
and for the white victims.


What of the total number of people who
helped? On 60% of the 81 trials on which the
victim received help, he received it not from
one good Samaritan but from two, three, or
even more.0 There are no significant differ-

4 If a comparison of latencies is made between
cane and drunk nonmodel trials only, the median
latency for cane trials is 5 seconds and the median
for drunk trials is 109 seconds (assigning 400 seconds
as the latency for nonrespondents). The Mann-
Whitney U for this comparison is significant at
p < .0001.

5 Among the white victim teams, the data from
Team 2 differ to some extent from those for Teams
1 and 4. All of the cane-after 70 seconds trials are
accounted for by Team 2, as are 4 of the S drunk-
before 70 trials. Median latency for cane trials is
longer for Team 2 than for the other teams; for
drunk trials, shorter. This is the same team that vio-
lated the “alternate days” instruction. It would ap-
pear that this team is being rather less careful—that
the victim may be getting out of his role. The data
from this team have been included in the analysis
although they tend to reduce the relationships that
were found.

6 The data from the model trials are not included
in this analysis because the model was programmed
to behave rather differently from the way in which
most real helpers behaved. That is, his role was to
raise the victim to a sitting position and then appear



Trials on which
help was offered :

Before 70 sec.



After 70 sec.




Total number
of trials














% of trials on which 1 +
persons left critical areab











X2timo = 23.19

p < .001

X2cana-drunk = 11.71

p < .001

% of trials on which 1 -}-
comments were recorded5











X2timo = 31.45
p < .001

X2oano-drlmk = 37.95
p < .001

Mean number of














Note.—Percentage and means not calculated for n’s less than 4.
a Fisher’s exact test, estimate of two-tailed probability.b Black and white victims are combined for the analyses of these data.

ences between black and white victims, or
between cane and drunk victims, in the
number of helpers subsequent to the first
who came to his aid. Seemingly, then, the
presence of the first helper has important
implications which override whatever cogni-
tive and emotional differences were initially
engendered among observers by the charac-
teristics of the victim. It may be that the
victim’s uniformly passive response to the
individual trying to assist him reduced ob-
servers’ fear about possible unpleasantness in
the drunk conditions. Another possibility is
that the key factor in the decisions of second
and third helpers to offer assistance was the
first helper. That is, perhaps assistance was
being offered primarily to him rather than to
the victim. Unfortunately the data do not
permit adequate assessment of these or other
possible explanations.

Characteristics of Spontaneous First Helpers

Having discovered that people do, in fact,
help with rather high frequency, the next
question is, “Who helps?” The effect of two
variables, sex and race, can be examined. On

to need assistance. Most real helpers managed to
drag the victim to a seat or to a standing position
on their own. Thus the programmed model received
somewhat more help than did real first helpers.

the average, 60% of the people in the critical
area were males. Yet, of the 81 spontaneous
first helpers, 90% were males. In this
situation, then, men are considerably more
likely to help than are women (x2 = 30.63;
p < .001).

Turning now to the race variable, of the
81 first helpers, 64-% were white. This per-
centage does not differ significantly from the
expected percentage of 55% based on racial
distribution in the cars. Since both black and
white victims were used, it is also possible to
see whether blacks and whites are more likely
to help a member of their own race. On the
65 trials on which spontaneous help was of-
fered to the white victims, 68% of the helpers
were white. This proportion differs from the
expected 55% at the .05 level (xa – 4.23).
On the 16 trials on which spontaneous help
was offered to the black victim, half of the
first helpers were white. While this proportion
does not differ from chance expectation, we
again see a slight tendency toward “same-
race” helping.

When race of helper is examined separately
for cane and drunk victims, an interesting
although nonsignificant trend emerges (see
Table 3). With both the black and white cane
victims, the proportion of helpers of each race
was in accord with the expected



Same as victim
Different from victim

White victims







Black victim







All victims









Note.—Chi-squares are corrected for continuity. White victims, x~ = 2.11, p. =16 ; black victim, p = .16 (two-tailed esti-
mate from Fisher’s exact probabilities test); all victims, x2 = 3.26, p = .08.

split, With the drunk, on the other hand, it
was mainly members of his own race who
came to his aid.7

This interesting tendency toward same-race
helping only in the case of the drunk victim
may reflect more empathy, sympathy, and
trust toward victims of one’s own racial
group. In the case of an innocent victim
(e.g., the cane victim), when sympathy,
though differentially experienced, is relatively
uncomplicated by other emotions, assistance
can readily cut across group lines. In the case
of the drunk (and potentially dangerous)
victim, complications are present, probably
blarne, fear, and disgust. When the victim is
a member of one’s own group—when the
conditions for empathy and trust are more
favorable—-assistance is more likely to be
offered. As we have seen, however, this does
not happen without the passing of time to
think things over.

Recent findings of Black and Reiss (1967)
in a study of the behavior of white police
officers towards apprehended persons offer an
interesting parallel. Observers in this study
recorded very little evidence of prejudice
toward sober individuals, whether white or
black. There was a large increase in prejudice
expressed towards drunks of both races, but

7 It is unfortunate from a design standpoint that
there was only one black victim. He was the only
black student in the class from which our crews
were recruited. While it is tenuous to generalize from
a sample of one, the problems attendant upon at-
tributing results to his race rather than to his indi-
vidual personality characteristics are vitiated some-
what by the fact that response latencies and fre-
quencies of help to him in the cane condition fall
between responses to Teams 1 and 4 on the one
hand and Team 2 on the other.

the increase in prejudice towards blacks was
more than twice that towards whites.

Modeling Effects

No extensive analysis of the response to
the programmed model could be made, since
there were too few cases for analysis. Two
analyses were, however, performed on the
effects of adjacent area versus critical area
models and of early versus late models within
the drunk condition. The data are presented
in Table 4. While the area variable has no
effect, the early model elicited help signifi-
cantly more than did the late model.

Other Responses to the Incident

What other responses do observers make
to the incident? Do the passengers leave the
car, move out of the area, make comments
about the incident? No one left the car on
any of the trials. However, on 21 of the 103
trials, a total of 34 people did leave the
critical area. The second section of Table 2
presents the percentage of trials on which
someone left the critical area as a function
of three variables: type of victim, race of
victim, and time to receipt of help (before
or after 70 seconds). People left the area on
a higher proportion of trials with the drunk
than with the cane victim. They also were far
more likely to leave on trials on which help
was not offered by 70 seconds, as compared
to trials on which help was received before
that time.8 The frequencies are too small to

8 Individuals are also somewhat more likely to
leave the area with the black victim than with the
white victims (x2 = 3.24, £<.08). This race effect is most probably an artifact, since the black victim ran more drunk trials than cane trials, the white victims, vice versa.





Not received


Critical area









Adjacent area






Both areas








Note.—-Early versus late: p < .04 (two-tailed estimate from Fisher's exact test). All three cane-model trials were early model trials; two critical area, one adjacent. Help was received on all. Table includes drunk trials only.

make comparisons with each of the variables
held constant.

Each observer spoke to the person seated
next to her after the incident took place. She
also noted spontaneous comments and actions
by those around her. A content analysis of
these data was performed, with little in the
way of interesting findings, The distribution
of number of comments over different sorts
of trials, however, did prove interesting (see
Section 3 of Table 2). Far more comments
were obtained on drunk trials than on cane
trials. Similarly, most of the comments were
obtained on trials in which no one helped
until after 70 seconds. The discomfort ob-
servers felt in sitting inactive in the presence
of the victim may have led them to talk
about the incident, perhaps hoping others
would confirm the fact that inaction was

appropriate. Many women, for example, made
comments such as, “It’s for men to help him,”
or “I wish I could help him—I’m not strong
enough,” “I never saw this kind of thing
before—I don’t know where to look,” “You
feel so bad that you don’t know what to do,”

A Test of the Diffusion of Responsibility

In the Darley and Latane experiment it
was predicted and found that as the number
of bystanders increased, the likelihood that
any individual would help decreased and the
latency of response increased. Their study
involved bystanders who could not see each
other or the victim. In the Latane and Rodin
study, the effect was again found, with by-
standers who were face to face, but with
the victim still only heard. In the present



No. males in critical area



,. N1 and up

Kruskal-Wallis Test (H)































Note,—Means and medians in seconds. Model trials omitted; no response assigned 400 seconds.



, 6 Hypothetical 3-person groups
• • Natural 3-person groups
,,«,…» Hypothetical 7-person groups
e—,.__, Natural 7-person groups




_| 1 1 1 f,—! j , 1 r , 1 1 , r—| 1 j

1 2 3 4 5 6 1 8 9 10 11 12 13 14 15 16 17 18

Seconds from occurence of collapse to

helping response from one member of group

FIG. 2. Cumulative proportion of groups producing a helper over time (cane trials, white victims, male
helpers from inside critical area).

study, bystanders saw both the victim and
each other. Will the diffusion of responsibility
finding still occur in this situation?

In order to check this hypothesis, two
analyses were performed. First, all nonmodel
trials were separated into three groups ac-
cording to the number of males in the critical
area (the assumed reference group for spon-
taneous first helpers). Mean and median
latencies of response were then calculated for
each group, separately by type and race of
victim. The results are presented in Table 5.
There is no evidence in these data for dif-
fusion of responsiiblity; in fact, response
times, using either measure, are consistently

faster for the 7 or more groups compared to
the 1 to 3 groups.9

As Darley and Latane pointed out, how-
ever, different-size real groups cannot be
meaningfully compared to one another, since
as group size increases the likelihood that one
or more persons will help also increases. A
second analysis as similar as possible to that
used by those authors was therefore per-
formed, comparing latencies actually obtained

9 The total number of people in the car was
strongly related to the number of males in the
critical area. Similar results are obtained if latencies
are examined as a function of the total number of
people in the car.


for each size group with a base line of hypo-
thetical groups of the same size made up by
combining smaller groups. In order to have
as much control as possible the analysis was
confined to cane trials with white victims and
male first helpers coming from the critical
area. Within this set of trials, the most fre-
quently occurring natural groups (of males
in the critical area) were those of sizes 3
(n = 6) and 7 (n = 5). Hypothetical groups
of 3 (n = 4) and 7 (n= 25) were composed
of all combinations of smaller sized groups.
For example, to obtain the hypothetical
latencies for groups of 7, combinations were
made of (a) all real size 6 groups with all
real size 1 groups, plus (b) all real size S
groups with all real size 2 groups, etc. The
latency assigned to each of these hypothetical
groups was that recorded for the faster of the
two real groups of which it was composed.
Cumulative response curves for real and hypo-
thetical groups of 3 and 7 are presented in
Figure 2.

As can be seen in the figure, the cumula-
tive helping response curves for the hypo-
thetical groups of both sizes are lower than
those for the corresponding real groups. That
is, members of real groups responded more
rapidly than would be expected on the basis
of the faster of the two scores obtained from
the combined smaller groups. While these
results together with those summarized in
Table 5 do not necessarily contradict the
diffusion of responsibility hypothesis, they
do not follow the pattern of findings obtained
by Darley and Latane and are clearly at vari-
ance with the tentative conclusion of those
investigators that “a victim may be more
likely to receive help . . . the fewer people
there are to take action [Latane & Darley,
1968, p. 221].”

Two explanations can be suggested to ac-
count for the disparity between the findings
of Table 5 and Figure 2 and those of Darley
and Latane and Latane and Rodin. As indi-
cated earlier in this paper, the conditions of
the present study were quite different from
those in previous investigations. First, the
fact that observers in the present study could
see the victim may not only have constrained
observers’ abilities to conclude there was no
emergency, but may also have overwhelmed

with other considerations any tendency to
diffuse responsibility. Second, the present
findings may indicate that even if diffusion
of responsibility is experienced by people
who can actually see an emergency, when
groups are larger than two the increment in
deterrence to action resulting from increasing
the number of observers may be less than the
increase in probability that within a given
time interval at least one of the observers
will take action to assist the victim. Clearly,
more work is needed in both natural and
laboratory settings before an understanding
is reached of the conditions under which dif-
fusion of responsibility will or will not occur,

In this field study, a personal emergency

occurred in which escape for the bystander
was virtually impossible. It was a public,
face-to-face situation, and in this respect dif-
fered from previous lab studies. Moreover,
since generalizations from field studies to lab
research must be made with caution, few
comparisons will be drawn. However, several
conclusions may be put forth:

1. An individual who appears to be ill is
more likely to receive aid than is one who
appears to be drunk, even when the immediate
help needed is of the same kind.

2. Given mixed groups of men and women,
and a male victim, men are more likely to
help than are women.

3. Given mixed racial groups, there is some
tendency for same-race helping to be more
frequent. This tendency is increased when the
victim is drunk as compared to apparently ill.

4. There is no strong relationship between
number of bystanders and speed of helping;
the expected increased “diffusion of respon-
sibility” with a greater number of bystanders
was not obtained for groups of these sizes.
That is, help is not less frequent or slower
in coming from larger as compared to smaller
groups of bystanders; what effect there is, is
in the opposite direction.

5. The longer the emergency continues
without help being offered (a) the less impact
a model has on the helping behavior of ob-
servers; (b) the more likely it is that indi-
viduals will leave the immediate area; that


is, they appear to move purposive!}’ to an-
other area in order to avoid the situation;
(c) the more likely it is that observers will
discuss the incident and its implications for
their behavior.

A model of response to emergency situa-
tions consistent with the previous findings is
currently being developed by the authors. It
is briefly presented here as a possible heu-
ristic device. The model includes the following
assumptions: Observation of an emergency
creates an emotional arousal state in the
bystander. This state will be differently inter-
preted in different situations (Schachter,
1964) as fear, disgust, sympathy, etc., and
possibly a combination of these. This state
of arousal is higher (a) the more one can
empathize with the victim (i.e., the more one
can see oneself in his situation—Stotland,
1966), (b) the closer one is to the emergency,
and (c) the longer the state of emergency con-
tinues without the intervention of a helper.
It can be reduced by one of a number of
possible responses: (a) helping directly, (b)
going to get help, (c) leaving the scene of the
emergencv, and (d) rejecting the victim as
undeserving of help (Lerner & Simmons,
1966). The response that will be chosen is
a function of a cost-reward matrix that in-
cludes costs associated with helping (e.g.,
effort, embarrassment, possible disgusting or
distasteful experiences, possible physical
harm, etc.), costs associated with not helping
(mainly self-blame and perceived censure
from others), rewards associated with helping
(mainly praise from self, victim, and others),
and rewards associated with not helping
(mainly those stemming from continuation of
other activities). Note that the major motiva-
tion implied in the model is not a positive
“altruistic” one, but rather a selfish desire to
rid oneself of an unpleasant emotional state.

In terms of this model, the following after-
the-fact interpretations can be made of the
findings obtained:

1. The drunk is helped less because costs
for helping are higher (greater disgust) and
costs for not helping are lower (less self-
blame and censure because he is in part
responsible for his own victimization).

2. Women help less because costs for help-
ing are higher in this situation (effort,
mainly) and costs for not helping are lower
(less censure from others; it is not her role).

3. Same-race helping, particularly of the
drunk, can be explained by differential costs
for not helping (less censure if one is of op-
posite race) and, with the drunk, differential
costs for helping (more fear if of different

4. Diffusion of responsibility is not found
on cane trials because costs for helping in
general are low and costs for not helping are
high (more self-blame because of possible
severity of problem). That is, the suggestion
is made that the diffusion of responsibility
effect will increase as costs for helping in-
crease and costs for not helping decrease.
This interpretation is consistent with the
well-known public incidents, in which possible
bodily harm to a helper is almost always
involved, and thus costs for helping are very
high, and also with previous research done
with nonvisible victims in which either (a) it
was easy to assume someone had already
helped and thus costs for not helping were
reduced (Barley & Latane) or (b) it was
possible to think that the emergency was
minor, which also reduces the costs for not
helping (Latane & Rodin).

5. All of the effects of time are also con-
sistent with the model. The longer the emer-
genc}’ continues, the more likely it is that
observers will be aroused and therefore will
have chosen among the possible responses.
Thus, (a) a late model will elicit less help-
ing, since people have already reduced their
arousal by one of the other methods; (b)
unless arousal is reduced by other methods,
people will leave more as time goes on, be-
cause arousal is still increasing; and (c) ob-
servers will discuss the incident in an attempt
to reduce self-blame and arrive at the fourth
resolution, namely a justification for not
helping based on rejection of the victim.

Quite obviously, the model was derived
from these data, along with data of other
studies in the area. Needless to say, further
work is being planned b}’ the authors to test
the implications of the model systematically.



BERSCHEID, E., & WALSTEE, E. When does a harm-
doer compensate a victim ? Journal of Personality
and Social Psychology, 1967, 6, 43S-441.

BLACK, D. J., & REISS, A. J. Studies in crime and
law enforcement in major metropolitan areas.
(Report submitted to the President’s Commission
on Law Enforcement and Administration of
Justice) Washington, D. C.: United States Govern-
ment Printing Office, 1967.

BRYAN, J. H., & TEST, M. A. Models r.nd helping:
Naturalistic studies in aiding behavior. Journal of
Personality and Social Psychology, 1967, 6, 400-

DABLEY, J., & LATANE, B. Bystander intervention in
emergencies: Diffusion of responsibility. Journal of
Personality and Social Psychology, 1968, 8, 377-

LATANE, B., & DARLEY, J. Group inhibition of by-
stander intervention in emergencies. Journal of
Personality and Social Psychology, 1968, 10, 215-

LATANE, B., & RODIN, J. A lady in distress: Inhibit-
ing effects of friends and strangers on bystander
intervention. Journal of Experimental Social Psy-
chology, 1969, 5, 189-202.

LERNER, M. J., & SIMMONS, C. H. Observer’s reaction
to the “innocent victim”: Compassion or rejection?
Journal oj Personality and Social Psychology,
1966, 4, 203-210.

SCHACHTER, S. The interaction of cognitive and
physiological determinants of emotional state. In
L. Berkowitz (Ed.), Advances in experimental
social psychology. Vol. 1. New York: Academic
Press, 1964.

SCHOPLER, J., & MATTHEWS, M. W. The influence of
the perceived causal locus of partner’s dependence
on the use of interpersonal power. Journal of
Personality and Social Psychology, 1965, 4, 609-

STOTLAKD, E. A theory and experiments in empathy.
Paper presented at the meeting of the American
Psychological Association, New York, September

(Received October 21, 1968)

Are you stuck with your online class?
Get help from our team of writers!