EFFECT OF SOCIAL DESIRABILITY BIAS AND MEMORY RECALL ON REPORTED CONTRIBUTIONS TO A WILDLIFE INCOME-TAX CHECKOFF PROGRAM

In this study of contributors and noncontributors to New, York's ''Ret urn a Gift To Wildlife'' tax check off program, mail survey responses reporting contributions to the program could be validated with tax rec ords. Evidence was found of memory-recall errors and social desirabili ty bias. The degree of bias was estimated to exceed twice the current contribution rate. Sociodemographic characteristics and participation in wildlife- related activities affected the degree of social desirabi lity bias. A model was proposed that used these variables to distingui sh those who misreported contributing from actual contributors. A ques tion eliciting future intention to perform a socially desirable act (i .e., contributing in the coming year) apparently results in an even st ronger social desirability bias than factual statements of past behavi or. Thus, using validated information yields a more realistic estimate of future contribution potential.