RMA Week 4: experimental methods II

what’s a within-subjects design?

aka repeated-measures, all participants receive all levels of the IV

what’s a between-subjects design?

aka independent groups, different groups of participants receive different levels of the IV

advantages of a between-subjects design

1. no order effects

2. some participants can only be between-subjects

3. naïve participants - no learning

disadvantages of a between-subjects design

1. lots of participants

2. characteristics may differ between groups

how to counteract disadvantages of between-subjects designs

random assignment to groups or participant characteristics matched in each group (but can’t account for every nuisance variable and variables can be related yet unmatchhable)

advantages of a within-subjects design

1. fewer participants required

2. reduced individual differences - confounding variables from each participant, and they’re used as ‘their own controls’

disadvantage of a within-subjects design

carryover effects - effect of one condition carries over to the next session

solution to carryover effects

order of conditions randomised/counterbalancing

counterbalancing can be too complex with more conditions - e.g. 24 orders with 4 conditions… what’s the compromise?

latin square design - each condition occurs equally often in each phase of the experiment

what’s a quasi-experiment

when 1+ IV’s are selected - not manipulated - as impractical/impossible/unethical

advantages of quasi-experiments

1. you can examine variables that would be impossible to manipulate otherwise

2. can inform potential experiments (if ethical)

disadvantages of quasi-experiments

1. no strong conclusions about cause and effect

2. no manipulation leads to more confounding variables that can’t be removed

how can improvements be made to quasi-experiments?

1. matching participants

2. if treatment study: tests before and after treatment with pp as their own control

what’s a random sample

everybody (in the population) has an equal chance of being selected - usually practically different!

what’s a stratified sample and one advantage

random selection of each subgroup of the population. advantage: key groups in sample

what’s a quota sample

representative sample that meets quotas/targets e.g. 50% males and females: ‘semi-randomised’

state the aim of psycho-physiological measurements

testing the effect of psychological variables on physiological processes

name three examples of psycho-physiological measurements

1. muscle activity

2. eye movements/eye blink rate

3. brain imaging/neurophysiological measures