Darley and Latane’s experiment: the intercom study (1968)
Aim
The aim of this experiment is to show how the number of people around an emergency would make
us more or less likely to intervene and help. The bystander effect would state that the more
witnesses to a predicament, it makes people less likely to offer assistance based on the assumption
that someone else would.
Participant demographic and procedure
The experiment was done on a group of college students, 59 females and 13 males, as part of a class
requirement and they were told to talk to each other through an intercom about the stresses of
living in an urban setting. The single participant taking part in each trial is tricked into thinking they
were talking to another real person but in reality, it is only a voice recording. The experiment was
split into three different types of trials, the first one with only one other “person” they were talking
to, the next with another confederate, and the last with 4 confederates. The person on the voice
recording would claim to have epilepsy and have a seizure, prompting an emergency to see if the
participant would intervene believing someone was actually in trouble; conclusions are then drawn
on the correlation between the number of confederates in the trial vs the likelihood and speed at
which the participant would help the “victim” in need.
Findings
The participants were all stressed by the simulated situation, not knowing that it was not a real
victim that needed medical attention; this was observed by their trembling hands and sweating
palms. The statistics also showed that participants were more likely to respond when they are
informed there are less bystanders, and they also responded faster. If it was only a trial between the
participant and victim, 85% people responded and it took an average time of 52 seconds. If there
was another confederate (3 people in total), 62% responded at an average 93 seconds. Lastly, with 4
confederates, only 31% responded at an average of 166 seconds. This would be the case because
they believe that other people would help so their assistance was unnecessary; but in small groups
they felt more responsible.
Explanation
• Some people may not have recognised that it was a seizure and were confused by the
situation
• Some may have felt conflicted as they didn’t know the ‘victim’ personally and so didn’t want
to intrude upon them at a vulnerable time to respect their privacy, but also want to help
them
• Constraints: this experiment was performed on a small demographic of people as they were
in the same career stage, age, and were all studying in the same college. The intercom also
played a part in making the experiment more artificial, as in daily life emergency situations,
we would be able to observe other bystanders.
The bystander effect & the diffusion of responsibility
The diffusion of responsibility is when we are placed in a situation with others present, we deny
personal responsibility and believe that someone else will do what is necessary in the given scenario.
The more people around, the less likely someone will accept the responsibility; therefore, it is
, actually more likely for someone to offer assistance when an emergency occurs in places where
there are fewer bystanders, as they would feel responsible to help.
Additionally, we may feel the need to fit in in social situations, consequently if others don’t help it
inhibits us to do so as well even if we think it would be the right thing to do. Finally, we may also
believe that someone else in that same space would be better qualified to assist the one in need,
therefore it would be wrong to insert ourselves. However, in some cases (face-to-face and enclosed
situations) people are more likely to act as they are in direct presence of the “victim”.
Piliavin et al. (1969) - the arousal cost-reward model
Aim
To investigate bystanderism and the reasons behind people displaying helping behaviour through
Piliavin’s arousal cost reward model of helping. Observing why and how we are motivated to help
others, whether through empathy or to reduce one’s own negative feelings. We then balance the
costs and rewards of helping or reduce our own distress by stepping away from the situation (also
depends if we deduced that the situation required assistance).
Participants
In this experiment the participants were not controlled because it was conducted as a field
experiment in the subway, therefore the ‘participants’ of the experiment were unaware that they
were part of a psychological study. This way, Piliavin could guarantee a higher rate of ecological
validity. The experiment was recorded by two female researchers, and the people who played
‘victims’ were two men that dressed and acted identically, both aged 25-35. Lastly, the experiment
was conducted in New York between 11am-3pm (non-rush hour) during the same 7.5-minute
journey; there was also a ‘model-helper’ who assisted the ‘victims’ after 70 seconds if nobody else
intervened.
Procedure
The two people playing victims posed as either a drunk man carrying a bottle, who smelled of
alcohol, and the other was a sober victim walking with a cane. In total, 103 trials were performed: 38
with the drunk victim and 65 with the sober one. The independent variable was the type of victim
that acted in that train, which affected the dependent variable: the reaction of the people around
them, seeing whether they intervened or not, and how long it took them.
Results
Results showed that in 78% of the trials someone offered help spontaneously, which included 62 of
the 65 trials with the sober victim, and 19 of the 38 trials with the drunk victim. In 60% of the trials in
which someone helped, others also offered assistance. For the sober victim, the average response
time was 5 seconds, whereas it took a median 109 seconds to respond to the drunk man if help was
offered at all. Additionally, 90% of helpers were male. However, contrary to the findings of Darley
and Latané on the diffusion of responsibility; the more people on the train, the more quickly help
was offered.
Explanation
Firstly, this study is high in ecological validity because it has demonstrated the model in a real-life
situation within the field experiment, performed with participants who were unaware of the staging,
therefore less likely to show demand characteristics to please/displease the study. However, it is
