Attribute framing and goal framing effects in health decisions

Levin, Schneider, and Gaeth (LSG, 1998) have distinguished among three type s of framing-risky choice, attribute, and goal framing-to reconcile conflic ting findings in the literature. In the research reported here, we focus on attribute and goal framing. LSG propose that positive frames should be mor e effective than negative frames in the context of attribute framing, and n egative frames should be more effective than positive frames in the context of goal framing. We test this framework by manipulating frame valence (pos itive vs negative) and frame type (attribute vs goal) in a unified context with common procedures. We also argue that the nature of effects in a goal- framing context may depend on the extent to which the research topic has "i ntrinsic self-relevance" to the population. In the context of medical decis ion making, we operationalize low intrinsic self-relevance by using student subjects and high intrinsic self-relevance by using patients. As expected, we find complete support for the LSG framework under low intrinsic self-re levance and modified support for the LSG framework under high intrinsic sel f-relevance. Overall, our research appears to confirm and extend the LSG fr amework. (C) 2001 Academic Press.