The aims

1. The study asked what makes bystanders help a stranger who collapses, and how quickly that help arrives.
2. The researchers varied whether the victim appeared ill or drunk, the race of the victim and whether a model helped first, while the number of bystanders present was recorded as it happened rather than set by them.

How the field experiment was run

1. The setting was a seven and a half minute non-stop run on the New York subway, and there were 103 trials with about 4450 passengers present across them.
2. Each trial used four confederates, meaning people secretly working for the researchers, with two women recording the data, one man playing the victim and one man playing a model helper.
3. The victim collapsed 70 seconds into the journey and stayed on the floor until he was helped.
4. In the ill condition he carried a cane, and in the drunk condition he carried a bottle and smelled of alcohol.

The results

1. Help arrived before the 70 second mark in 52 of the 57 trials with an ill white victim and in 7 of the 8 trials with an ill black victim.
2. On the same measure the drunk victim was helped in only 5 of 24 trials when white and 4 of 14 when black.
3. Of the 81 people who helped spontaneously, meaning within 70 seconds, 90% were men.
4. There was no strong correlation between the number of bystanders present and whether or how quickly help came.

The conclusions

1. Piliavin et al concluded that seeing the collapse creates unpleasant arousal, and that a bystander acts to reduce that arousal in whichever way costs least.
2. Helping an ill man costs little, whereas helping a drunk man risks disgust or trouble, which is why far fewer people helped him.
3. Because helping did not fall away as the number of bystanders rose, the results challenge a simple diffusion of responsibility account of the bystander effect.

Strengths and weaknesses

1. A strength is that the passengers were on an ordinary journey and did not know they were in a study, so the helping recorded was genuine.
2. A second strength is that with 103 trials and about 4450 people present, the patterns rest on far more observations than a single laboratory session could gather.
3. A weakness is that the passengers gave no informed consent, were deceived and could not really withdraw from a moving train.
4. A second weakness is that the researchers could not control how many passengers were in the carriage or whether any had medical training, so extraneous variables may have shaped who helped.

Turning this into marks

1. Name the study as a field experiment on the New York subway and quote the 103 trials and the 70 second collapse.
2. Give the contrast rather than the general point, so quote 52 of 57 ill trials against 5 of 24 drunk trials.
3. Weigh the realism of the setting against the ethical cost rather than retelling the study.