Individual and situational determinants of altruism and helping behaviour

Abstract

Studies which have attempted to assess the relative contributions of individual and situational determinants of helping behaviour and altruism have yielded diverse and contradictory data. The present investigation attempts to reconcile some of these inconsistent findings by examining the interaction of individual factors such as personality and sex with situational influences such as group size and task involvement. Explanations of helping behaviour and empirical studies are critically reviewed and methodological problems considered. The first four experiments take place in the laboratory and employ a simulated emergency representing a person falling off a ladder. Results indicate that individual factors may be more meaningful predictors of helping behaviour when examined in light of relevant situational effects. Differences in the helpfulness of low and high Machiavellians, as measured by Christie's Mach IV Inventory, seem to emerge only when face-to-face interaction, latitude for improvisation, and emotional involvement are possible. In such situations, groups of low Machs appear more likely than groups of high Machs to aid a distressed victim. However, when subjects are alone, not communicating, or in mixed-Mach groups, no differences in helping are found. The findings also demonstrate that group size effects on helping may be enhanced when communication channels are blocked, ambiguous emergencies are employed, and passive confederates pose as bystanders. The last two experiments pool previous findings and more naturalistic data to explore the consistency of the helpfulness of individual subjects across several situations. The results cast doubt on the existence of general helping dispositions. The findings suggest that seemingly irrelevant factors often ignored by researchers may play an important role in studies of helping behaviour. Contradictory results of previous research may be at least partly due to a number of inappropriate assumptions about the strength of single variables, and between-study comparisons may thus be unjustified.