Further explorations of medical decisions for individuals and for groups.

BACKGROUND: Important discrepancies between clinical practice and health policy may be related to the ways in which physicians and others make decisions about individuals and groups. Previous research has found that physicians and laypersons asked to consider an individual patient generally make different decisions than those asked to consider a group of comparable patients, but this discrepancy has not been observed in more recent studies. This study was designed to explore possible reasons for these findings. METHODS: Prospective jurors (N = 1,013) each made a recommendation regarding a risky treatment for an incurable blood condition. Perspective (individual vs group) was crossed with uncertainty frame (probability vs frequency) and response wording (original vs revised) in a 2 x 2 x 2 between-participants design. RESULTS: When the strength of participants' recommendations was considered, the effects of perspective, uncertainty frame, and response wording were not statistically significant. When recommendations were dichotomized, participants in the revised-response-wording conditions were more likely to recommend treatment to the group than to the individual. CONCLUSIONS: These results conflict with previous findings for this scenario and suggest that reported differences between decisions for individuals and decisions for groups are not robust.

Public preferences for prevention versus cure: what if an ounce of prevention is worth only an ounce of cure?

BACKGROUND: The belief that small preventive efforts bring large benefits may explain why many people say they value prevention above all other types of health care. However, it often takes a great deal of preventive medicine to prevent a bad outcome. This study explores whether people value prevention or cure more when each brings the same magnitude of benefit and examines whether preferences for prevention or cure vary according to the severity of the disability of the patients who can receive the preventive or curative intervention. METHODS: 289 prospective jurors were presented with a policy dilemma involving how best to allocate funds to benefit people with varying levels of disability. Each project was said to influence the functional ability of 100 nursing home residents, either by improving their level of function or by preventing their level of function from declining. RESULTS: When given a choice between preventive and curative interventions, more subjects preferred the preventive intervention (37% vs 21%, p=0.002). However, when the strength of people's preferences was taken into account, the preference for preventive interventions was not statistically significant (p=0.135). With both preventive and curative interventions, the subjects preferred helping patients with more severe disabilities (p < 0.005 for both comparisons). This preference for helping more severely disabled patients did not differ for prevention and cure (p=0.663). CONCLUSION: When the magnitude of benefit was held constant, the subjects slightly preferred prevention over cure. In addition, they preferred directing limited resources toward those with greater disabilities, regardless of whether those resources were targeted toward prevention or cure. These findings suggest that previously stated preferences for prevention over cure may result from a belief that small efforts at prevention will be repaid by large reductions in the later need for cure.

The actor-observer effect revisited: effects of individual differences and repeated social interactions on actor and observer attributions.

This research examined several factors hypothesized to influence the actor-observer effect (AOE). Participants engaged in 3 successive dyadic interactions: after each interaction, they rated the importance of 4 causal factors in influencing their behavior and that of their partner. The AOE held for 1 external factor, interaction partner, and 1 internal factor, personality, but not for situation or mood. Actor and observer attributions changed in predicted ways across the 3 interactions: Actors increasingly emphasized the importance of their partner, whereas observers increasingly emphasized personality: both actors and observers substantially lowered their attributions to the situation. We found consistent individual differences in attributional tendencies that allowed us to predict who showed the AOE. Together, the findings demonstrate that A-O differences depend on: (a) the specific causal factor invoked, (b) the individual's history in the situation, and (c) individual differences among attributors. Discussion focuses on the limited generality of the AOE and the need for a more complex formulation of A-O differences in attribution.