With a within-subject design, a major problem is that once a subject has completed the first part of the study, the experience or changed circumstances could influence performance in later parts of the study. This problem is known as a SEQUENCE or ORDER effect.
1. The Problem of Controlling Sequence Effects -- COUNTERBALANCING.
a. Testing once per condition (Incomplete design) -- participants are tested in each of the conditions, but only once per condition.
- Complete counterbalancing ( All possible orders)
- Partial counterbalancing (Selected orders) – (i) random selection of combinations or (ii) Latin square
b. Testing more than once per condition (Complete design)
- Reverse counterbalancing (ABBA) -- experimenter simply presents the conditions in one order and then presents them in the reverse order
- Block randomization -- the basic rule is that every condition occurs once before any condition is repeated a second time
2. Problems with Biasing
a. Experimenter bias -- biased experimenters might treat the research participants in the various conditions differently.
b. Controlling for experimenter bias – standardized (mechanized) procedures; double-blind design.
c. Participant bias
- HAWTHORNE EFFECT
- DEMAND CHARACTERISTICS
- EVALUATION APPREHENSION
3. Controlling for participant bias – deception; manipulation check; field research