KINE 2049 Chapter Notes - Chapter 5: Dependent And Independent Variables, Internal Validity
Document Summary
Selecting an appropriate experimental design: e. g. using a two-group design rather than a single group with a pre- test and post-test. Carefully conducting the experiment: this reduces threats e. g. instrumentation errors, local history events, differential selection of subjects, etc. Systematic elimination of potential confounding variables (possibly the most important factor in protecting the internal validity of the study) Best recommendations are: (1) randomly assign groups, (2) use control groups, and (3) be as careful as possible in making measurements. Improves internal validity by assuring groups are equivalent. Either way, number of required subjects is increased and more time + resources required. Best to rely on proper random sampling than pretest to ensure. Expectancy (self-fulfilling prophecy) initial equivalence: refers to u(cid:374)i(cid:374)te(cid:374)tio(cid:374)al effe(cid:272)ts o(cid:374) study (cid:272)aused (cid:271)y resear(cid:272)her"s prior beliefs of outcome of study, a. k. a. self-fulfilling prophecy, to remove this threat, use double-blind or automated procedures.