Dependent Samples Matched Samples Experimental Designs BetweenSubjects Design

Dependent Samples Matched Samples Experimental Designs

Between-Subjects Design l l l Different groups of participants receive different levels of the IV Each participant serves in only one condition Independent samples may be used in each condition • Participants are selected without regard to who is in the other condition

Random Assignment: l A method of assigning research participants to groups so that each participant has an equal chance of being in any group.

Random Assignment The Logic of Random Assignment: l Individual differences (or differences in subject characteristics) should theoretically be equally distributed between groups

Random Assignment Potential Problems with Random Assignment: 1. May not actually balance participant’s characteristics/individual differences across groups 2. Small sample size decreases likelihood of balancing participant’s characteristics across conditions

Matched-Samples Design

An Example Ball Toss on One or Two Feet (samples matched by height)

Pair 1 foot 2 feet 1 2 3 4 5 6 7 8 9 10 3 2 2 2 0 1 2 2 1 2 1 1 3 2 1 1

An Experiment IV ? l Balance Levels of IV? l l 1 foot 2 feet DV? l # of baskets made

An Experiment Hypotheses? l HA: Participants who stand on one foot will differ in the number of baskets scored from those who stand on two feet when shooting baskets. l H 0: Participants who stand on one foot will not differ in the number of baskets scored from those who stand on two feet when shooting baskets.

Comparison of Between-Subjects Designs: Independent and Matched Samples Independent Samples Design Matched Samples Design l Different groups of participants l Different groups of receive different levels of the participants receive different IV levels of the IV l Each participant serves in only l Each participant serves in one condition only one condition l Independent samples are l Dependent samples are used in each condition l Participants are selected l Participants are matched without regard to who is in to someone in the other condition on variable(s) correlated with the DV

What Should We Match On? The characteristics on which matching takes place must be correlated with the DV Examples: l Gender l Age l Race l SES l IQ l Physiological variables (e. g. , height, weight, resting heart rate, blood pressure) l

Matched-Groups Design Matching: l A technique for controlling potential confounds l Involves systematically matching participant characteristics across groups

Matched-Groups Design How? l Exact match OR l Rank order participants by specified variable l Randomly assign the two individuals who scored the highest to different groups l Repeat for #3 and #4, and for #5 and #6, etc.

Matched-Groups Design Advantages of matching: 1. Increases internal validity 2. Decreases variability between groups 3. Increases power (IF the characteristics on which matching takes place are really correlated with the DV) 4. May decrease the sample size you need to find a significant effect

Matched-Groups Design Disadvantages of matching: 1. May increase # of participants you have to recruit and could decrease sample size if participant pool is limited • • May have to pretest a large # of participants in order to match on a specific variable (many may not be matched) May require an increase in total sample size to get an equal number of participants of each type for each group 2. Decreases power (IF the characteristics on which matching takes place are NOT correlated with the DV) 3. May increase demand characteristics

Matched-Groups Design Natural Pairs: l Sometimes there are obvious, natural pairings. Examples: l Husbands and wives l Siblings l Twins l Roommates