The parable of the Good Samaritan is the West’s standard test of character, and the test it sets is a narrow one. It asks what a person does when they come upon someone in trouble and nothing but the encounter is pressing on them. A priest and a Levite pass a beaten man lying by the road. A Samaritan stops.
John Darley and Daniel Batson built a version of that road at Princeton and scored what happened on it. Over three days in December 1970 they sent seminary students one at a time from one building to another to record a short talk, some of them on the parable itself. On the way, every one of them passed a man slumped in a doorway, head down, eyes closed, who coughed twice and groaned as they drew level.
Of the five factors the experimenters tested together in the same regression models, one predicted whether a student helped: how much of a hurry the student had just been put in. The topic of the talk did not reach significance. Neither did any of the three measures of religiosity.
Princeton Theological Seminary supplied the subjects
Sixty-seven students there had already filled in questionnaires on types of religiosity. Forty-seven of them, the ones who could be reached by telephone, were booked for the experiment. Seven were dropped from the analysis, three because their procedures were contaminated and four because they were suspicious of the setup, which left forty. Each was paid a dollar for the questionnaire session and a dollar fifty for the experiment itself.
Forty students across six conditions is a small study, and it stays small at every point below. The paper also records a randomisation error that pushed too many students into the intermediate hurry conditions; the authors say it worked against the hurry prediction, and that it made no difference to the message tests. Everything that follows comes from one seminary, one alley, one victim and three December days in 1970.
The experiment ran on 14, 15 and 16 December, between ten in the morning and four in the afternoon. The schedule was kept tight in an attempt to hold the weather and the light reasonably consistent. According to the New York Times, temperatures during those hours moved by no more than five degrees Fahrenheit across the three days. No rain or snow fell, and the first two days were sunny while the third was cloudy.
The cover story was a study of the vocational careers of seminary students. Each student read a passage, was told to prepare a three to five minute talk on it with no notes allowed, and was handed a hand-drawn map to an office in the building next door. The route ran through an alley.
The assistant delivered a line on the way out
The passage was the first variable. Some students received a text about what jobs seminary graduates end up doing and whether it is possible to minister through an establishment job at all. The others received the parable of the Good Samaritan, printed in full.
The second variable was what the assistant said as the student left. In the high hurry condition he looked at his watch first: “Oh, you’re late. They were expecting you a few minutes ago. We’d better get moving. The assistant should be waiting for you so you’d better hurry. It shouldn’t take but just a minute.” In the intermediate condition: “The assistant is ready for you, so please go right over.” In the low hurry condition: “It’ll be a few minutes before they’re ready for you, but you might as well head on over. If you have to wait over there, it shouldn’t be long.”
The whole hurry manipulation was those few seconds as the student left, delivered once.
The victim scored them as they passed
He sat slumped in a doorway, head down, eyes closed, not moving. As each student passed, he coughed twice and groaned. Anyone who stopped got a scripted answer: a respiratory condition, pills already taken, a few minutes of rest and he would be fine, and thank you very much for stopping.
He also graded every student on a six point scale, running from failing to notice him at all up to refusing to leave, and he was blind to the questionnaire scores and to the condition each student was in. In some cases two of the boundaries between categories needed a second source to draw. The postexperimental questionnaire distinguished students who never registered the man from students who registered him and walked on, and the assistant’s report distinguished those who walked on from those who reported him on arrival.
Sixteen of the forty offered some form of help, either directly or by telling someone in the other building. Twenty-four did not. Broken out by hurry, sixty-three percent of the low hurry group offered help, forty-five percent of the intermediate group, and ten percent of the high hurry group. In the analysis of variance, hurry was significantly related to helping and the talk topic was not. A stepwise multiple regression across all five variables agreed. Hurry was the only significant predictor of whether a student helped, and the only significant predictor of how much help was given.
On several occasions, the authors record, a seminary student on the way to give the talk on the parable stepped over the man in the doorway to get there.
The sermon result is the one that travels, and forty students cannot settle it. Fifty-three percent of those carrying the Good Samaritan passage offered help, against twenty-nine percent of those carrying the vocations passage, and that gap did not reach significance. It runs in the direction the norm salience account predicts, and a sample split two ways from forty is not enough to show whether it is real. The authors were not neutral about which way it should fall: no effect was their own first hypothesis, drawn from the parable, and they report the data as congruent with it. Their summary is that the salience of helping norms is a weaker determinant in this situation than many, including themselves, would expect.
Almost all the hurried students mentioned him afterwards
Darley and Batson did not conclude that the students who kept going were indifferent. Almost all of the seminarians who had been put in a hurry mentioned the man afterwards, in the post-experiment interview, as someone who on reflection had possibly needed help. For some of those students the authors’ reading is that hurry deferred the interpretation, an effect they compare to Tolman’s narrowing of the cognitive map.
For others, they write, it seems more accurate to say the students decided not to stop, and those subjects appeared aroused and anxious afterwards. Those students were hurrying because the experimenter was depending on them to be somewhere. Conflict, in the paper’s own words, rather than callousness, can explain the failure to stop.
The religion scores turned up in the sixteen who helped
Among the students who did help, one of the three religiosity scores tracked the kind of help offered. Those scoring high on what the authors call religion as quest, a religiosity emerging out of a search for meaning in one’s personal and social world, tended to offer help that was more tentative and less complete. Those scoring high on a separate questionnaire scale, doctrinal orthodoxy, went the other way, toward a category added to the scoring sheet after the pilot run. It covered students who, after stopping, refused to leave the victim, or insisted on taking him somewhere outside the experiment for coffee or to the infirmary.
The authors first labelled that behaviour rigid, then argued with themselves about it in print. All that can safely be said, they write, is that one style of helping was directed at the victim’s presumed underlying needs and was little modified by what he said about them. Their second thought is that calling it inappropriate may say more about the experimenters than about the helper. Sixteen students is too thin a base for a personality finding, and those correlations come from that subgroup rather than from the forty.
The scale the study ran on had six steps. The last of them existed because the man in the doorway asked for it.