Term
|
Definition
-random assignment of subjects to treatment conditions
-active manipulation of IV
-use of control group |
|
|
Term
| Quasi-experimental design |
|
Definition
-active manipulation of IV
-use of control group
-subjects NOT randomly assigned to groups |
|
|
Term
|
Definition
-only done when nothing better
-limited to describing outcomes |
|
|
Term
| Notation for types of studies: X, O, R |
|
Definition
X = treatment (IV)
O = observation or measurement (DV)
R = random assignment |
|
|
Term
| between-subject design (true experimental) |
|
Definition
-separate groups of subjects measured and compared
-different groups of subjects exposed to different treatments or levels of IV |
|
|
Term
| List 5 between-subjects designs |
|
Definition
-pretest-posttest control group
-posttest only control group
-solomon four-group
-randomized controlled trials |
|
|
Term
| Pretest-posttest control group design |
|
Definition
RO1 X 02
RO3 O4
-one group receives treatment, the other doesn't
-compare groups at pre- and post-test |
|
|
Term
| Pretest-posttest control group strengths and weaknesses |
|
Definition
strength: increases chance that subjects between groups relatively equal on all variables except IV (good for internal validity)
-weakness: threat to external validity is pretest sensitization |
|
|
Term
| Posttest only control group design |
|
Definition
R X O1
R O2
-random assignment to groups
-posttest measure only
-pretest unnecessary due to random assignment |
|
|
Term
| Posttest only control group strengths and weaknesses |
|
Definition
-strength: pretesting might sensitize subjects to experimental treatment or bias posttest scores due to repeated measures
-strength:most threats to IV controlled for, but...
-weakness: not widely used because no guarantee random assignment has controlled for selection bias |
|
|
Term
| Solomon Four-Group Design |
|
Definition
RO1 X O2
RO3 O4
R X O5
R O6
-basically = pretest posttest control group design + posttest only control group design |
|
|
Term
| Solomon Four-Group strength |
|
Definition
| -controls for interactions between pretest and IV that combine to affect DV |
|
|
Term
|
Definition
| compare 3+ groups (e.g., efficacy of 3+ therapies) |
|
|
Term
|
Definition
| 2+ IVs (e.g., interaction of 3 therapies given at 3 levels on clinical outcome. in this case, could look at treatment effect, frequency effect, and treatment X frequency interaction) |
|
|
Term
| types of randomization: completely randomized design |
|
Definition
| all people meeting criteria and volunteering to participate randomly assigned to groups |
|
|
Term
| types of randomization: randomized block design |
|
Definition
| all people meeting criteria and volunteering to participate are assigned to different blocks (e.g., mild disorder, moderate disorder, and severe disorder), and randomly assigned to tx conditions within those blocks (e.g., all "mild" subjects randomly assigned to Tx A or Tx B) |
|
|
Term
| within-subjects (repeated measures) designs |
|
Definition
-subjects exposed to all levels of IV (all treatments), randomly assigned sequence
-tx effects compared within same person |
|
|
Term
| within-subjects design strengths and weaknesses |
|
Definition
-strength: no need to equate groups of subjects because only one group studied
-strength: fewer subjects needed
-strength: more economical in terms of time and effort
-weakness: order & carryover effects |
|
|
Term
| control for carryover effects |
|
Definition
-counterbalancing - assign subjects to diff. sequences of tx (for order)
-washout period - period of time between tx to allow effects to dissipate (for carryover)
-period effects (for carryover) |
|
|
Term
| balanced latin square design purpose |
|
Definition
| to control for order effects |
|
|
Term
|
Definition
-has 2+ IVs
-one IV is between-subjects
-one IV is within-subjects |
|
|
Term
| types of quasi-experimental designs |
|
Definition
-nonequivalent comparison group design
-single time-series design
-multiple time-series design |
|
|
