-
Systematic variation could be due to the IV
- this is what we test when we use F ratios
- F = t2Compares within-groups variance and between groups variance
- if treatment has no effect, ratio = 1
-
Systematic variation could be due to Confounds
uncontrolled systematic variance: systematically biases the results of research
-
Sources of variation due to chance, error, or nonsystematic variation
- individual differences
- transient states
- environmental factors
- differential treatment
- measurement error
-
Individual differences
preexisting differences among participants: cognitive, social, emotional, behavioral differences
-
Transient States
- healthy/ill
- tired/rested
- happy/sad
- enthusiastic/resentful
- (moods, attitudes, physical condition)
-
Environmental factors
external noise, weather, light, ect...
-
Differential treatment
- experimenters' moods and health also change
- cannot be friendlier toward some participants than others
-
Internal validity
the degree to which a researcher draws accurate conclusions about the effects of the IV
-
Threats to Internal Validity
- history
- maturation testing/pretest sensitization
- instrumentation
- statistical regression
- selection or biased assignment of p's to condition
- differential attrition or mortality
- selection-maturation interaction
- diffusion or imitation of treatments
|
|