Experiment designs

Experiments

Experiments are methods used widely in psychology, involving researchers manipulating a variable to test its effects, in order to gather information or test a hypothesis

Repeated measures design

Repeated measures design involves the same participants take part in each condition of the independent variable

Strength of repeated measure designs

High internal validity as extraneous variables are reduced and less individual differences - such as IQ levels of participants. Therefore cause and effect can be applied.

Weakness of repeated measure designs

The order of the conditions may affect performances - sequence of tasks may affect performance levels. Participants may perform better in later conditions because they've had practice with the previous task. This can be overcome by counterbalancing, each condition is tested first or second in equal amounts as participants are assigned different sequences.

Independent groups design

Independent groups design involves different participants in each independent condition of the study.

Strength of independent group design

High internal validity, as there are more controlled variables, no demand characteristics or order of effects. Therefore cause and effect can be easily established.

Weakness of independent group design

Low internal validity as extraneous variables are harder to reduce due to more individual differences, such as the extraneous effect of IQ levels. Therefore cause and effect cannot easily be established.

Matched pair design

Match pairs design involves using different participants for each condition of the experiment, but allocating each participant to a condition based on a  relevant participant variable.

Strength of matched pair design

High internal validity, as there are no demand characteristics or order of effects as the differences between two groups are minimised. Cause and effect can easily be established

Weakness of matched pair design

Low applications as the process is time consuming, as participants need to be screened and tested to be categorised into accurate task conditions. Therefore this is unsuitable for widespread experiments, using a large sample. Similarly, an exact match is rarely possible as it is not possible to control all participant variables.