research methods: experimental designs Flashcards
What does experimental design mean?
The arrangement of groups of participants in an experiment
What are the three types of experimental design?
- repeated measures
- independent groups
- matched pairs
What is the repeated measures design?
All participants take part in all the conditions of the experiment
What kind of data is produced from the repeated measures design?
Produces related data as each pps score in one condition can be paired with a data point in the other condition
Evaluation of repeated measures: order effects -> limitation
- Taking part in the first condition influences performance in the second condition
- in the second condition, they may do better e.g. due to practice or worse e.g. due to tiredness or boredom
- therefore, it is unclear whether the change in the second condition was due to improvement or the experimental IV
- Order acts as a confounding variable
Evaluation of repeated measures design: demand characteristics -> limitation
Participants are also more likely to work out the aim of the experiment when they experience all the conditions of the experiment
How can order effects in the repeated measures design be controlled?
- Through counterbalancing which uses the ABBA format
- half the participants complete condition A first and B second, the other half of the same start with condition B then A
What is the independent groups design?
Different participants are used in each condition of the experiment
How are pps allocated in the independent groups design?
Participants are randomly allocated to each condition to avoid researcher bias when assigning to conditions
What kind of data does the independent groups design produce?
Produces unrelated data as the individual data points in one condition cannot be paired with any of the data points in the other condition
Evaluation of the independent groups design: participant variables -> limitation
- The pps who occupy different groups are not the same in terms of participant variables
- If more participants with a particular characteristic are randomly assigned to one of the groups e.g. age, this can influence the measurement of the DV (an extraneous variable)
- reduces the validity of the results
How can participant variables be dealt with?
Using random allocation where pps are randomly allocated to different experimental conditions
- attempts to evenly distribute participant characteristics across the conditions of the experiment using random techniques
Evaluation of independent groups design: economical -> limitation
The independent groups design is less economical than the repeated measures design as each participant contributes as a single result only
- twice as many participants would be required to produce an equivalent data to that collected in a repeated measures design
Evaluation of independent groups design: order effects -> strength
Order effects are not a problem as pps take place in one condition
- while they are in repeated measures as participants are more likely to guess the aims
What are matched pairs?
Pairs of participants are matched in terms of key variables, such as age or socioeconomic status
- each participant only takes part in one condition
What is the effect of having one partipant from each pair allocated to a different condition?
An attempt to control the confounding variables and participant variables
What kind of data does the matched pairs design produce?
Related data as each pps score in one condition can be paired with a data point of their partner in the other condition
What are the advantages of the matched pairs design (evaluation)
- Reduced participant variables as participants are matched on a relevant characteristic
- no order effects as participants only take part in one condition
What are the limitations of the matched pairs design (evaluation)
- Takes longer to set up than other experimental designs (more time consuming)
- Needs twice as many participants as a repeated measures design (more expensive)
- Participants are similar but not identical so there may still be some participant variables between conditions that influence the dependent variable
