Factors Affecting the Choice of Statistical Test (AQA A Level Psychology)
Revision Note
Written by: Claire Neeson
Reviewed by: Lucy Vinson
Factors affecting the choice of statistical test
The purpose of statistical testing
A statistical test determines if a difference/correlation is statistically significant
The outcome is more than a chance occurrence
The outcome determines whether the null hypothesis is accepted or rejected
What factors determine the choice of statistical test?
There are 3 distinct criteria that a researcher must consider before deciding which statistical test to use:
Have they conducted a test of difference (e.g. a lab experiment) or a test of correlation?
If they have conducted a test of difference, did they use an independent measures design, repeated measures design, or a matched pairs design?
an unrelated design refers to independent measures/groups
a related design refers to repeated measures and matched pairs
Have they collected nominal, ordinal or interval data?
The table below illustrates which test should be used and when:
Tests of Difference | Tests of association or correlation | ||
|---|---|---|---|
Unrelated design | Related design | ||
Nominal data | Chi-Squared | Sign test | Chi-Squared |
Ordinal data | Mann Whitney U | Wilcoxon T | Spearman's rho |
Interval data (Parametric tests) | Unrelated t-test | Related t-test | Pearson's r |
Chi-Squared is a test of both difference and association
Spearman's rho and Pearson's r are the only tests of correlation
Examiner Tips and Tricks
The statistical tests which feature on the AQA specification are as follows:
Parametric tests:
Unrelated t-test
Related t-test
Pearson’s r
Non-parametric tests:
Mann-Whitney U
Wilcoxon T
Chi-squared
Spearman's rho
The Sign Test
You do not need to know how to calculate each test, but ensure that you can justify when to use each one. The mnemonic below can help you learn this:
Tests of Difference | Tests of association or correlation | ||
|---|---|---|---|
Unrelated design | Related design | ||
Nominal data | Chi-Squared Carr | Sign test Should | Chi-Squared Come |
Ordinal data | Mann Whitney U Mashed | Wilcoxon T With | Spearman's rho Swede |
Interval data | Unrelated t-test Under | Related t-test Roast | Pearson's r Potatoes |
Parametric tests
Parametric tests assume the following:
A normal distribution
This occurs when data is symmetrical around the mean: scores near the mean value are more frequent than scores which are far from the mean
The normal distribution has the 'bell curve' appearance when in graphical form
Example: height is a measurement that has a normal distribution
The use of interval data
This is because interval data is the most sensitive and precise type of data
Homogeneity of variance
If the set of scores per data set/condition are similar in terms of their dispersion, then this means they have homogeneity of variance
If both conditions show a similar standard deviation, for example, then this indicates that there was not a large amount of variability in each condition, i.e. the scores clustered about the mean
Non-parametric tests
Non-parametric tests do not follow the same criteria as parametric tests
There is no assumption of a normal distribution
This is because what is being measured may not fall within strict, clearly defined parameters
Example: scores on a memory test
Non-parametric tests use nominal or ordinal data
Non-parametric tests do not depend on homogeneity of variance
Parametric tests are more powerful and precise than non-parametric tests
They have more statistical power than non-parametric tests, as they are more likely to lead to the detection of a significant difference or correlation
Worked Example
Here is an example of an AO2 question that you might be asked on this topic.
AO2: You need to apply your knowledge and understanding, usually referring to the ‘stem’ in order to do so (the stem is the example given before the question).
Bella has conducted research using elite marathon runners as her sample. She measured the body temperature of her participants who had either just run 10K or had rested for 30 minutes. Bella wants to carry out a parametric test on this data.
Q. Explain why Bella should carry out a parametric test on this data. Justify your answer.
[3 marks]
Model answer:
Link to type of data:
Bella should carry out a parametric test as her data is interval (temperature measurements have distinct and equal intervals between each measurement, plus body temperature is a relatively stable variable) [1 mark]
Link to type of distribution:
Bella can expect to see a normal distribution of data because all of the participants are elite athletes, which means that their scores are likely to converge around the mean [1 mark]
Link to variance:
She can also expect homogeneity of variance as the standard deviations per condition are likely to be similar due to the nature of the sample – all elite athletes who are used to running long distances [1 mark]
Last updated:
You've read 0 of your 5 free revision notes this week
Sign up now. It’s free!
Did this page help you?
