How to Answer a Statistical Test Question (AQA A Level Psychology) : Revision Note
How do I answer an AQA A Level Psychology statistical test question?
Statistical or inferential test questions can appear in Papers 1 , 2 and 3, and they assess your ability to:
demonstrate knowledge and understanding of inferential testing and be familiar with the use of inferential tests (AO1)
apply knowledge and understanding of scientific ideas, processes, techniques and procedures (AO2)
A statistical/inferential test question is presented with a scenario or stem (usually a research study) and is designed to test your AO2 skills
E.g., which statistical test should be used to calculate whether there is a significant difference in reported smoking behaviour between the two surveys? Give three reasons for your answer
This would be presented as a 4-mark question
It is important to recognise that when a statistical/inferential test question says 'in this study' or 'with reference to this investigation' you must apply your response explicitly to the context given
Which statistical test would be most suitable to analyse the data in this investigation? With reference to this investigation, explain three reasons for your choice of test
This would be presented as a 7-mark question
Worked Example
Here is an example of a 4-mark statistical test question:
Q. A psychology teacher read a researcher’s study on sport and happiness. She considered whether setting group tasks could improve her students’ level of happiness. She decided to conduct an independent groups experiment with 30 students taking A-level Psychology using the same happiness questionnaire.
Suggest an appropriate statistical test the psychology teacher could use to analyse the data. Justify your choice of test.
[4 marks]
Identify the chosen statistical test:
The Mann-Whitney U test
Provide three justifications:
This is because:
the teacher is looking for a difference
it is an independent groups design
she collected ordinal data
Marking commentary:
This receives full marks as the statistical test is correctly identified for 1 mark along with three correct justifications for 3 marks. It is acceptable to bullet point the justifications for a statistical test question like this. There is no need to justify the choice of test in the context of the study, as the question does not explicitly demand this.
Worked Example
Here is an example of a 7-mark statistical test question:
Q. A psychologist decided to design an experiment to test the effects of recreational screen time on children’s academic performance.
The psychologist randomly selected four schools from all the primary schools in her county to take part in the experiment involving Year 5 pupils. Three of the four schools agreed to take part. In total, there were 58 pupils whose parents consented for them to participate. The 58 pupils were then randomly allocated to Group A or Group B.
For the two-week period of the experiment, pupils in Group A had no recreational screen time. Pupils in Group B were allowed unrestricted recreational screen time. At the end of the experiment, all pupils completed a 45-minute class test to achieve a test score.
The psychologist wanted to test the statistical significance of the data.
Identify the most appropriate choice of statistical test for analysing the data collected and explain three reasons for your choice in the context of this study.
[7 marks]
Identify the chosen statistical test:
The Mann-Whitney U test
Provide three justifications in the context of the study:
Because the teacher is looking for a difference
between no recreational screen time/unlimited screen time on exam performance
It is an independent groups design
as the pupils either had no recreational screen time or unlimited recreational screen time
She collected ordinal data
as the difference between each test score is not fixed
Marking commentary:
This receives full marks as the statistical test is correctly identified for 1 mark along with three correct justifications for 3 marks. Each justification is explicitly linked to the study and therefore receives a further 3 marks.
How do I know which statistical test to choose?
There are 3 factors that determine the choice of statistical test:
Is the researcher looking for a difference or correlation?
If it is a test of difference, is the experimental design related or unrelated?
Is the data collected nominal, ordinal or interval?
The best method for remembering which statistical test to use is to commit the following table to memory:
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 |
It is important to remember that Chi-Squared is a test of both difference and association (not correlation)
Spearman's rho and Pearson's r are the only tests of correlation
Examiner Tips and Tricks
One revision tip is to frequently fill out the above table (without notes) to see if you can recall which test to use and when
Some students find that the following mnemonic helps them to remember the order of the tests in the table:
Carrots should come mashed with swede under roast potatoes
Make sure you are able to identify whether the data collected in the stem is nominal, ordinal or interval data
Nominal data is categorical or frequency data, e.g. being securely or insecurely attached, being able to swim or not
Ordinal data is collected via a subjective scale, e.g., via a questionnaire a researcher has created only for the purposes of their study
Interval data is collected via an objective scale, which is a measurement used for purposes inside and outside of psychological research, e.g., heart rate, kilograms, miles per hour, temperature, kilometres
You've read 0 of your 5 free revision notes this week
Sign up now. It’s free!
Did this page help you?
