H567/02 Section A - Levine et al. (2001): cross-cultural altruism
Learn how three everyday helping situations were compared across 23 cities, what community patterns emerged, and why a city's helping rate cannot define every person or prove a cultural cause.
Culture and the helping question
Levine, Norenzayan and Philbrick investigated whether spontaneous help to strangers differs across places and is consistent across situations.
Previous research often compared communities within one country. This study broadened the question to one large city in each of 23 countries. It measured non-emergency favours and examined links with population size, economic productivity, individualism–collectivism and pace of life. A separate comparison examined simpatia, a cultural tradition emphasising friendliness, politeness and concern for others, including strangers.
The researchers did not assume that collectivism necessarily means more help to strangers: concern for one's own group may not extend to outsiders. They also considered competing economic explanations. Prosperity might make help easier, or a focus on individual goals might make strangers' needs less salient. These are hypotheses to test, not stereotypes to accept.
The prescribed theme is Responses to people in need and the area is social psychology. The research asks how social and cultural context relates to helping. Its title's reference to altruism should not be taken as a direct measurement of selfless motives: observing help tells us what happened, not whether the helper expected any personal benefit.
Cities, sample and design
The study combined standardised field helping tasks with correlational comparisons of community characteristics.
The 23 large cities spanned the Americas, Europe, Africa and Asia, with populations above about 230,000. They included Rio de Janeiro, New York, Vienna, Shanghai, Calcutta, Lilongwe and Kuala Lumpur. Selection sought breadth but depended partly on where travelling students and colleagues could collect data. Countries and cities were therefore not randomly sampled, and one city cannot represent every community in its country.
Helping trials occurred in at least two downtown locations, during business hours on clear summer days, in one or more years from 1992 to 1997. For the pen and hurt-leg tasks, experimenters approached solitary pedestrians, usually the second eligible person crossing a predetermined line. Children under 17 and people considered unable or unlikely to assist, such as very old people, people with visible disabilities or heavy packages, were excluded. Across cities, 424 people were approached in the pen task (214 men, 210 women), and 493 in the hurt-leg task (253 men, 240 women); there were 281 blind-person trials.
The staged tasks are field experiments in the broad sense used for the prescribed study: experimenters create a need in a natural setting. However, culture, national wealth and city size were not manipulated or randomly assigned. Comparisons between cities are quasi-experimental, and associations with community variables are correlational. Calling the whole project a field experiment does not turn a wealth–helping correlation into causal evidence. Community analyses used city/country summaries, not all individual pedestrians as independent national data points.
Three helping procedures
Three observable tasks operationalised spontaneous, direct assistance without requiring a spoken request.
| Task and materials | Standardised action | What counted as help |
|---|---|---|
| Dropped pen | Walking at a practised moderate pace, the experimenter dropped a pen 10–15 feet from a solitary approaching pedestrian and continued without appearing to notice | Calling attention to the pen and/or returning it |
| Hurt leg | Wearing a visible brace and limping heavily, the experimenter dropped magazines and struggled to reach them within 20 feet of a pedestrian | Offering assistance and/or beginning to help |
| Blind person at a crossing | Wearing dark glasses and carrying a white cane, the experimenter waited at a suitable signal-controlled crossing just before the light turned green | At minimum, informing the experimenter that the light was green |
The crossing trial ended after 60 seconds or when the light turned red, whichever came first. Help therefore did not always mean physically escorting the person across. The pen task required very little effort; the other situations signalled greater need, but none recreated an apparent collapse like Piliavin's emergency.
Experimenters were college-aged men, neatly and casually dressed; in most cities one local person collected all data. Written instructions, field training and joint practice supported consistent acting, selection and scoring. Avoiding required verbal communication reduced language demands. It did not eliminate differences in how clothes, disability cues or a stranger's actions were understood locally.
Community measures were external to the helping trials: population figures, per-capita GDP adjusted for purchasing power, expert ratings of individualism–collectivism, and walking times over a flat unobstructed 60-foot stretch. Walking time indicated pace of life; longer time means slower walking. Spain, Brazil, Costa Rica, El Salvador and Mexico formed the study's simpatia group.
City rates and community findings
Helping varied markedly across the sampled cities, with some consistency across the three situations.
The table shows original published percentages rounded to two decimal places where needed. The average percentage is a convenient summary of the three task rates; the paper's formal overall helping index instead averaged their standardised z-scores.
Continue this lesson.
You’re previewing NovaLearn without an account. Create a free account to read the rest.
Create free account