PSY 2174 Lecture Notes - Lecture 7: Confounding, Internal Validity, Language Development
Document Summary
A well-designed experiment will allow you to say that your manipulation caused a change in your dependent variable, and that nothing else could have caused that change. Critical to demonstrate that any change occurs because of the manipulation. Without it, there is nothing to test. Measures desired sources of variance to evaluate the effects of the iv and the. Due to chance factors and individual differences. Make sure that there are real differences between the groups. Make sure the groups are similar as possible at the outset of the study (rule out confounding variables) A specific test of whether the independent variable manipulation actually worked the way it was intended. Testing hypothesis that females, but not males, tend to turn anger inward rather than express it externally. Expose both genders to a frustrating task, then measure hostility at test. Looks as though females were not necessarily angered by the frustration manipulated.