- – Allocating participants to conditions
- – Independent measures, repeated measures, matched pairs
Repeated measures
Participants take part in both conditions
- + Any difference between conditions likely to be due to IV not pts variables
- + Controls individual differences (removes EV) = compares participant’s scores from one condition with their score from condition 2 = increase internal validity
- + Requires less pts
- – Consider demand characteristics = pts find it easier to work out aim of study = invalid results
- – Order effects?
How to deal with practice and order effects?
In any repeated measures design = consider controlling order effects. However, you can’t get rid of order effects, only prevent and reduce them.
Counterbalancing – specifically called ABBA
- – Half of participants do condition A then B
- – Other half do B then A
- – Assign pts randomly to conditions
Randomisation – presentational order of conditions is decided by use of random generators. Decide which half of the class gets which condition
Matched pairs
Each person is one condition is matched with someone in the other condition on a range of factors e.g. age, gender, IQ.
- + Controls individual differences (pts matched on key factors = controls individual differences)
- + Controls P&E effects (pts do only 1 condition = no P&E effects)
- – Practically difficult
- – Hard to decide which factors are important (always be other potential factors that should have been matched)
- – How do we measure the variable we match? E.g. IQ test?
- – Hard to find a match – large sample size