| Abstract: | Science attitude scales can be expected to predict science-related behavior. But A-B correspondence can seldom be expected to approach the r value of 1.00. Attitude and behavior covary to the degree that valid measurement and mediating variables are considered: Individual differences of subjects (e.g., high versus low self-monitoring); the social situation (e.g., direct experience); cognitive factors (e.g., attitude accessibility in memory). Attitude and behavior are reciprocal; that is, attitude can follow behavior. Variables other than attitude, e.g., behavioral intention, previous behavior, and habit, may under some conditions better predict behavior. Also, a viable alternative to attitude testing may be direct appraisal of some science classroom behaviors through ethnomethodology. |
