PSY-B - Psychology PSY-B 311 Lecture Notes - Lecture 16: Interrupted Time Series, Time Series, Internal Validity
Document Summary
Cause and effect must co-vary. (changes in cause produce changes in effect) Experiments: manipulate a variable and measure its effect on another variable. Dependent variable (dv: ensure that subjects in experimental groups are initially equivalent before iv is introduced. Often researchers cannot randomly assign ps to experimental conditions. Sometimes researchers may be unable or unwilling to manipulate an iv. Occurs when all potential sources of confound variance are eliminated. Sources of confound variance: biased assignment to experimental conditions. Loss of subjects at different rate across conditions (survivors may differ from non-survivors: pretest sensitization. Giving a pretest before the experiment may alert ps to hypothesis of the study: history effects. Due to extraneous effects that occur outside of the research setting: maturation: normal changes that occur over time (associated with development) Not manipulated by the researcher; event that occurred for other reasons (antidrug program) Other explanations (confounds) for change in drug use between o1 and o2.