A NOTE ON THE UNBIASEDNESS TEST OF RATIONALITY USING SURVEY DATA

This paper examines the performance of the unbiasedness test of ration ality using survey data by accounting for measurement errors associate d with the survey expectations, and also the generated regressor probl em arising from the quantitative expectations having been generated fr om an auxiliary equation. It is shown that ignoring the generated regr essor problem can yield incorrect coefficient estimates and markedly l arger standard error estimates, which could lead to an incorrect accep tance of a false null hypothesis.