-
Simple random assignment
- easiest way to be sure that the experimental groups are roughly equal to start
- each participant has equal probability of being assigned to any experimental condition
-
Matched random assignment
- increase similarity between groups
- using p's scores on a measure known to be relevant to outcome of experiment
- matched a pair of subjects- one randomly assigned to one condition, other to other condition
-
Repeated measure designs
- when different subjects are assigned to different conditions: randomized groups design or between subjects
- in some studies the same individuals are in each treatment condition: within-subjects or repeated measures design
-
Advantages of Within-Subjects
- more powerful: more likely to detect effects of IV - p's are identical
- require fewer p's
-
Disadvantaged of Within-Subjects
- order effects: each p gets all treatments, order in which they receive them may have an effect on behavior
- guard against order effects: counterbalancing- presenting the levels of IV in different orders to different participants
- even with counterbalancing, there may be carryover effects: effects of 1 level of IV are still present when another level is introduced
-
Selection of subjects
- probability (random) vs. non-probability (non-random) sampling
- convenience sampling, quota sampling, purposive sampling
|
|
