Week One - Experimental Design & Inferential Statistics Flashcards
5 ways to control for secondary/extraneous variables?
ELIMINATION: eg medication, history
CONSTANCY: always done at a particular, time, light level etc
SECONDARY VARIABLE CHANGED TO IV
RANDOMISATION: random allocation to groups
STATISTICAL ADJUSTMENT
What are the sources of invalidity in experiments? (10)
PROACTIVE HISTORY: minimised through elim, constancy and/or RA
RETROACTIVE HISTORY: something that happens to people during the experiment
REPEATED TESTING: learning or sensitisation from pre-test
STATISTICAL REGRESSION: regression toward the mean (extreme scores will move closer to the mean when re-tested). Control groups help this.
MATURATION: more so when long time between experiments (CG help).
LOSS OF SUBJECTS
INTERACTION EFFECTS: where a prev treatment affects a subsequent treatment (in repeated measures).
ERROR IN DV MEASURE
EXPERIMENTER BIAS: use double or single-blind measures
ERROR FROM STATS