Experimental Research Hanser and Wheeler
Principles Independent Variable Dependent Variable
Control –Random assignment to groups –Setting –Time of day –Therapist –Etc.
Generalizability –Random selection/random sample (match population of interest) –Size of sample –Normal distribution –External validity
Experimental Design –Way to structure conditions in scientific inquiry –Test hypotheses by controlling certain variables and allowing others to change
Hypotheses –Prediction –Null hypothesis –Reject or accept
Experimental Designs Pre-Experimental True Experimental: Independent Groups True Experimental: Related Groups Quasi Experimental
Independent Groups Experimental group – Control group Random assignment to conditions Pretest-posttest, posttest only Three or more groups – multiple levels Factorial design – interaction effects
Related Groups Subjects act as their own controls Counteracts individual variety within participant groups Counterbalancing – order of conditions Matched pairs – participants matched on age, gender, SES, intelligence, pretest scores, etc.
Quasi Experimental Do not determine cause-effect Point to relationships that exist Convenience sample Ex post facto (after the fact) – researcher does not have control over independent variables because they have already occurred or are not manipulatable
Correlational Researcher examines two or more data sets to note the degree of relationship Researcher does not determine group membership or arrange/manipulate events; rather she or he studies existing relationships Used to make predictions
Causal-Comparative Concerned with probable causation Identify variables or factors and then explain why these factors occurred at a given point in time Wheeler!