Term
| independent groups features |
|
Definition
-between subjects, independent samples
-manipulated IV
-random assignment (equivalent groups) |
|
|
Term
|
Definition
-between subjects, related samples
-manipulated IV
-matching (equivalent groups)
--to match: measure potential confound, pair Ps on scores, randomly assign one of each to either group |
|
|
Term
| nonequivalent groups features |
|
Definition
-between subjects, independent samples
-subject variable as IV |
|
|
Term
|
Definition
-within subjects, related samples design
-manipulated IV (all subjects participate in all levels)
-COUNTERBALANCE to avoid sequence effects |
|
|
Term
| analyses for one level design |
|
Definition
-t test
-assumptions: interval/ratio data, normal distribution, homogeneity of variance |
|
|
Term
| advantages of multilevel designs (2) |
|
Definition
1. discover nonlinear effects
2. address multiple questions and rule out alternative explanations (falsification) |
|
|
Term
| analyses for multilevel design |
|
Definition
-t test inappropriate, increased type II error
-one way ANOVA (H0: M1 = M2 = M3, H1: at least one is different)
-use post hoc to test for significance |
|
|
Term
| data presentation options (3) |
|
Definition
1. sentences (2-3 levels)
2. table (many data points, precise values)
3. graph (large differences, interactions)
-line: continuous variable ONLY
-bar: discrete variable, continuous possible |
|
|
Term
| control group designs (3) |
|
Definition
1. placebo (to yield baseline measure)
2. waiting list control group (ensure similarity of control and treatment groups - ethics)
3. yoked control group (each member of control group is "yoked" to an experimental group P)
-used when the experimental Ps participate for varying amounts of time/experience different events |
|
|
