Subject variables. Address the difficulty of studying subject variables. To illustrate this, have them suppose that they did a study in which women did better than men on an intelligence test. Before saying that women are smarter than men, the students would have to establish that
- their sample of men and their sample of women were
representative of the population at large;
- the IQ test had construct validity; and
- the test had equal construct validity for men as for women.
Furthermore, even if all three points were established
- we would not know why women did better than men (heredity, environment), and
- the r-squared (coefficient of determination) would probably be small, revealing that very little variability was accounted for by gender.
Back to Chapter 7 Main Menu