Experimental/research designs:

Complete the table: designs revision

Quizizz link for teachers: https://quizizz.com/admin/quiz/5a8f0e42d84f6d002510f298

Independent groups: This means that you have different people in the two conditions or levels of the IV. This can be a problem as participant variables can affect the DV, meaning the results may lack internal validity unless the researcher has randomly allocated the participants across the two groups.

Repeated measures: This mean that the same participants take part in more than one condition. This means there are no participant variables affecting the DV as the participants act as their own control, however the internal validity may still be at stake as repeated measures designs can be affected by order effects meaning that if condition B always follows condition A, Pps may do worse in condition B not because of the nature of the IV manipulation but because they are tired or bored due to having taken part in the previous condition. On the other hand, they may be better in condition B if they have been allowed to practice in condition A. Which ever happens,if the same thing is true of all Pps then any changes in the DV may be due to order effects and not the manipulation of the IV meaning we have a limited internal validity. This can be resolved using counterbalancing, or the AB-BA design, where half the Pps do Condition A first then condition B and the other half do B first and then A thus reducing order effects and improving internal validity.

Matched pairs: This when we have different people in the groups/conditions of the IV however they have been carefully matched so that each Pp has some-one who is similar to them on certain important variables which may have affected the DV unless controlled for in this way. This is a clever design which reduces both order effects and participant variables however it can be hard to set up as recruiting suitable Pps for the two groups can be difficult and the process of screening and matching them is time consuming.

It is important that you can spot which design is being utilised in a scenario question as it affects which statistical is appropriate. For example in an experiment with independent measures you would use a chi-squared if the data was nominal or a Mann Whitney if it was ordinal, interval or ration. If the design was repeated measures or matched pairs you would use a Wilcoxons if the data was ordinal, interval or ratio.

Practice Questions

  1. Megan is a psychology student who has to carry out a practical investigation for her coursework. She has decided to study whether there is a difference in the cognitive ability between children of different ages. She will be studying two different classes at her local primary school over one day, one class of four year olds and one class of seven year olds. Which of the following experimental designs will Megan use?

A Matched pairs

B Independent groups

C Repeated measures

D Correlation

2. Mr Faraz wants to compare the levels of attendance (measured by the percentage of days they have been in school in the Autumn Term, 2016) between his psychology group and those of Mr Simon, who teaches a different psychology group.

a. Which of the experimental design would Mr Faraz use in his investigation?

b. Which stats test should he use?

c. Why might there be a problem with any conclusions drawn from this study?