This experiment expands the social psychological framework for the understanding of altruistic helping intentions. The results indicated that the evolutionary adaptations of reciprocal altruism and kin selection have significant effects on intentions for making a risky rescue attempt. It is especially striking that a construct related to reciprocal altruism was the strongest predictor of the likelihood of helping. The effect of the expectancy for the target to help if positions in the scenario were reversed was three times as strong as the effect of oneness, the egoistic construct touted by Cialdini’s (1997) research team, and three and a half times as strong as the effect of empathic concern, the construct Batson’s (1997) research team promotes as the mediator of altruistic helping. The effect of kinship had a complex relationship to helping intentions, with a negative effect as mediated by the construct of oneness and a positive effect as mediated by the construct of expectancy. Although kin targets provoked lower mean level of oneness than friend targets, the level of oneness made a greater impact on the likelihood of helping for kin targets than for friend targets. Kinship also had a significant unique effect that was not accounted for by the previously established psychological predictors of helping.

Kinship and the Psychological Predictors of Helping

Most predictions were supported by the results of data analysis, however one of the findings was the opposite of prediction. Kinship did not increase mean levels of the psychological predictors of helping, empathy and oneness, disconfirming the first hypothesis. In fact, kin targets actually produced lower mean level of oneness than friend targets. This may be due to differences in the quality of relationships to siblings and friends. A comparison of undergraduates’ closest friend relationship and closest sibling relationship revealed that participants felt closer to their friends than their siblings and felt more differentiation with their siblings than with their friends (Pulakos, 1989).

There may be an evolutionary basis for this differentiation. Trivers (1974) applied Hamilton’s (1964b) theory of inclusive fitness to within-family conflict. Because offspring have an r of .5 with their parents and siblings, their strategies for maximizing inclusive fitness will differ from those maximizing the inclusive fitness of their parents or siblings. Individual children will want to receive more than their “fair share,” at the expense of other children, in order to maximize their own fitness. In fact, children may even try to gain more resources than the parents can afford to give. Parents are equally related to their offspring, so they will tend to resist these demands. Hence, sibling rivalries will occur as a result of competition for the attention and resources of parents. Children may try to differentiate themselves from their siblings as a strategy to gain a greater portion of resources.

In contrast to relationships with siblings, individuals can choose who they want as friends. Individuals tend to spend more time with their friends than they do with siblings (Pulakos, 1989). Although children may be biologically closer to their siblings, they may be socially closer to their friends. The post-hoc analysis of the current data indicated that participants liked their friends more than their siblings. In Cialdini et al.’s (1997) study with four levels of relationship, a Tukey post-hoc comparison found that oneness was higher for close family members than for good friends in one of five scenarios, no differences were found for the other four scenarios. This contradiction in results should be resolved through replication with additional studies.

It is also striking that kinship did not have a significant effect on levels of empathy. This finding fails to replicate the results of Rushton’s (1991) re-analysis, where 71% of the variance in experienced empathy was accounted for by the genetic relationship to the target. Cialdini et al. (1997) found that as relationship closeness increased so did empathic concern for a needy other, using four levels of relationship; a near-stranger, an acquaintance, a good friend, and a family member. Cialdini et al.’s (1997) categories confound social proximity with kinship. Although kin may benefit from kin selecting effects, participants may not necessarily feel as socially close to their family members as they do to a close friend. It is possible that this effect was significant due to the variance resulting from the interpersonal closeness of the relationship to the three non-kin targets. A Tukey post-hoc comparison did not find significant differences between the level of empathy experienced for good friends and close family members in any of five scenarios (Cialdini et al., 1997). This may indicate that influences stemming from kin selection and social proximity cancelled each other out in terms of their effects on empathy.

As evident in the confirmatory factor analysis, empathic concern significantly covaried with oneness. In regards to the social psychological debate between altruist and egoist advocates, results supported Baston et. al’s (1997) hypothesis that empathic concern does make a unique contribution to the variance in helping. At the same time, the effect of empathic concern, explaining 11% of the variance in the likelihood of helping, was significantly weaker than the effect of oneness, accounting for 13% of the variance in the likelihood of helping. Hypothesis 2 predicted that once the effects of kinship, reciprocal altruism and oneness were accounted for, sadness, aversive arousal, and empathic concern would not make a statistically significant unique contribution to the variance in the likelihood of helping.

As predicted, aversive arousal and sadness did not make a statistically significant unique contribution to the variance in the likelihood of helping. This result was consistent with Cialdini et al.’s (1997) analysis, and Batson et al.’s (1989) results discounting the effect of sadness. One conclusion that may be reached from this replication of results is that the relief of sadness and aversive arousal experienced when one helps a person in a state of need does not substantially influence the likelihood of helping behaviors. In this case, measures of these constructs may be removed from future studies without arousing concern. On the other hand, one might object that these measures did not sufficiently capture the experience of sadness and aversive arousal.

To respond to this objection, one can examine the results of exploratory factor analysis of these constructs found in Fultz et al. (1988) and the confirmatory factor analysis in the current study, presented in Table 1. Fultz et al. (1988) found that “feeling low” loaded between .78 and .80, “low-spirited” loaded between .74 and .86, and “heavy hearted” loaded between .67 and .79 on the factor interpreted as sadness. None of these items loaded above .38 on factors interpreted as empathy and distress. In the current study, “Feeling low,” “Low-spirited,” and “Heavy-hearted” had correlations of .90, .93, and .71 respectively with the factor interpreted as sadness. The Cronbach alpha for the sadness scale was .87. Fultz et al. (1988) found that “disturbed” loaded between .78 and .87, “troubled” loaded between .78 and .80, and “uneasy” loaded between .69 and .84 on the factor interpreted as distress. None of these items loaded above .39 on factors interpreted as empathy and sadness. In the present study, "Disturbed," "Troubled," and "Uneasy" had correlations of .80, .92, and .89 respectively with the factor interpreted as aversive arousal. The Cronbach alpha for the aversive arousal scale was .90.

These results support the criterion validity and convergent validity of the sadness items. One can also argue that these items have face validity. Results from Fultz et al. (1988) also support the divergent validity of these items. One possibility for future research would be to combine the measures used in this experiment with alternative measures. A confirmatory factor analysis could determine whether the items target the same construct as alternative measures. Path analysis or structural equation modeling could be used to determine if the alternative measures are better predictors of helping behaviors than the current measures.

As a whole, these results indicate that kin selecting influences are not mediated by the currently touted psychological predictors of altruistic helping intentions. Possible mediators of kin selecting effects will be discussed in the next section. Evidently, kinship has a complex relationship to helping intentions. Kinship had a negative effect as mediated by oneness, and a positive effect as mediated by expectancy. The summation of these two indirect effects resulted in a nonsigniticant net mediated effect of kinship on helping intentions.

The egoistic influence of oneness was moderated by kinship. The degree to which one saw one’s kin as part of oneself was a stronger predictor of helping intentions than the degree to which one saw one’s friend as part of oneself. This result illustrates the importance of considering kinship as an important factor in social psychology, and opens the door to other possible moderating influences on the effects of oneness. The degree of empathic concern felt towards the target was not affected by kinship, and kinship did not moderate the impact of empathic concern on helping intentions. This result distinguishes empathy from oneness, in that empathy appears to be a psychological experience independent of kinship, and is a non-moderated predictor of altruistic helping intentions.