PSYB01H3 Chapter Notes - Chapter 8: Selection Bias, Ex Post Facto Law, Time Series
Document Summary
When we cannot directly manipulate an independent variable. Subject variables: type of an independent variable; particular individual characteristics that cannot be manipulated and instead must be selected; ex. Natural treatment: when exposures to events situations or setting that emanate from the real world define how participants are selected; levels are exposure and non exposure; ex historical events, natural disasters, policy changes. Other quasiexperimental designs similar to this: interrupted-time-series design, multiple-group before-and-after design, repeated-measures-panel designs: there are more kinds as well. Natural treatments as quasi-independent variables: ex post facto: analysis, after the fact analysis of its affect on particular dependent variable. Subject variables as quasi-independent variables: can combine quasi-independent variables and manipulated independent variables, mixed factorial design: combines both between- and within-subjects factors. Nonequivalent-control-group designs: as its not random assignment control group cannot be considered equivalent to experimental group, have high external validity and ecological validity in particular. Confounding factor for interpreting results of studies that match groups on pretest measures.
