PSYC 203 Lecture Notes - Lecture 8: Internal Validity, External Validity, Statistical Conclusion Validity
Document Summary
Measure the impact on some behaviour of interest. Method of agreement: if x, then y, x is sufficient/necessary for y to occur, may be causal, analogous to experimental group. Method of difference: if not x, then not y, analogous to control group. Manipulated independent variable controlled by experimenter: situational some aspect of the environment, task, instructional. Don"t necessarily need an absence of the independent variable. Results could be due to independent variable or to confounding variable. Try to avoid: ceiling effects task is too easy, all scores very high, floor effects task is too difficult, all scores very low. Research example behaviour: study examining the influence of age, education, and peer pressure on helping, dependent variable helping behaviour, three independent variables: With a manipulated independent variable: assuming no confounds, conclude that independent variable causes the dependent variable. With a subject independent variable: no causal conclusions just conclude that the group differ, groups may differ in several ways.