What are the three basic group designs?
Related/within groups/within subjects/repeated measures
Unrelated/between groups/independent measures
Matched pairs - two different groups but individuals are paired or related
What are advantages of using a repeated measures design?
Participant variables such as intelligence, are kept constant between conditions
Less variation between conditions so more statistical tests can be used
As people are used more than once, fewer participants are required
What are disadvantages of a repeated measures design?
Order effects such as fatigue could occur so you would need to counter balance to reduce errors
Demand characteristics may develop as the participant experiences both conditions and guesses the hypothesis therefore guessing differently
Different tests may be needed
Participants may not be available for the second part of the experiment
Cannot be used in quasi or natural experiments as participants fall into one of the conditions e.g. Male or female
What happens in a between groups design?
Each participant receives one treatment
One participant in one and then another participant in another
The dependent variable would be measured once and compared between groups
What are some advantages of between group designs?
Participants are only used once so fatigue/boredom etc do not influence the second condition
Demand characteristics are less of a problem
The same test can be used e.g. Each group can learn the same words
It is the quickest and easiest way of allocating people to groups
What are the disadvantages of using a between groups design?
Participant variables such as intelligence would differ unless in a matched pairs design
A pre test or large sample means it is more expensive and time consuming to complete
Variation between conditions means less sophisticated statistical tests can be used
More participants are required than for repeated measures design
What is a matched pairs design?
A matched pairs design uses similar but different participants in each condition and are matched if there are any characteristics that could affect the results or performance
What are the disadvantages of a matched pairs design?
Participant variables can never be perfectly matched which could affect the results
Matching participants is time consuming and can be difficult
In preparing for the study more people would need to be sampled to make a match
Participants can only be used once
It could be difficult to identify appropriate criteria for matching
What are some control techniques to reduce threats to validity?
In a related design participants are tested more than once under different conditions
Differences may be due to order effects (fatigue/practice effects)
In a between-groups design, different participants are used in each group so differences may be due to differences in people in the groups
Factors other than the independent variable need to be controlled
What is randomisation used for?
To assign participants to groups in a between groups design
What does randomisation account for?
As long as the groups are big enough, randomisation would smooth out differences between groups
An alternative would be to use matched pairs
But we don’t know what factors to match people on
What could be a problem in a related design?
Order effects
In a repeated measures design, what can be used to control the order?
A cross over design
What is called when the order is controlled?
Counterbalancing
When there are a large number of conditions, what can be used to ensure control?
Latin Squares
What is a Latin square design?
It is where each condition occurs in each order positionas equal number of times e.g. For four conditions we only need four different orders instead of 24
Although it doesn’t control for all order effects
What can effect validity?
Demand characteristics
Placebo effect
Groups should be matched except everything except IV and that includes expectation
What is a single blind study?
Where participants don’t know what condition they are in but the researchers do
What is a double blind study?
Where neither the researcher nor the participant know what condition the participant is in
-
The Importance Of Theory9
-
Asking Good Research Questions And Critical Thinking20
-
Research Designs And Ethics20
-
Qualitative Approaches To Research19
-
Analysing Qualitative Data33
-
Introduction To Quantitative Research31
-
Designing Questionnaire Surveys23
-
Interventions And Experimental Studies13
-
Designing A Good Study19
-
Samples And Populations8
-
Descriptive Statistics28
-
Z tests For Single Scores19
-
Z Test For Samples5
