PS295 Chapter Notes - Chapter 13: Seat Belt, Internal Validity
CHAPTER 13: QUASI EXPERIMENTAL DESIGNS
-used because many real world questions can not be addressed within the narrow strictures
of experimentation
-ex do seat belt legislation reduce traffic fatalities?
-if researcher lacks control over assignment of participants to conditions and/or doesn't
manipulate the causal variable of interest, the design is quasi experimental
-experimental designs generally have high internal validity unlike quasi experimental
designs
-cannot have independent variable; rather a quasi-independent variable
Pretest-Posttest Designs
-measuring participants before introduction of DV and again after to see if QI has varied
O1 X O2
O2 = pretest measure
X = intro of the QI (quasi independent variable)
OS = posttest measure of drug use 1 year later
O stands for observation
-one group pretest-posttest design is a very poor research strategy bc it fails to eliminate
Document Summary
Used because many real world questions can not be addressed within the narrow strictures of experimentation. If researcher lacks control over assignment of participants to conditions and/or doesn"t manipulate the causal variable of interest, the design is quasi experimental. Experimental designs generally have high internal validity unlike quasi experimental designs. Measuring participants before introduction of dv and again after to see if qi has varied. X = intro of the qi (quasi independent variable) Os = posttest measure of drug use 1 year later. One group pretest-posttest design is a very poor research strategy bc it fails to eliminate most threats to internal validity like maturation effects, history effects etc. First measurement of drug use (o1) may have started students thinking about drugs. This resulted in lower drug use independently of the educational progam (testing effect) May also be threatened by regression to the mean.