Are the mean GCSE scores in secondary school different for boys and girls?
We mentioned above that one easy way of seeing if there is a difference in GCSE score between different groups of people is to calculate and compare the mean values of GCSE score in each group you are investigating. Let’s say we are interested in any differences in educational achievement based on sex. We can attempt to answer this question by comparing mean GCSE scores between girls and boys in Year 11. We can use the Compare Means command to determine if sex has any bearing on GCSE scores. (We could obviously ask many questions like this – which you can do by yourself after we’ve run this first mean comparison together.)
Before we run any tests on s1gender, the variable containing data on the sex of YCS survey participants, we should use the Frequencies functions to see if there are any missing values, as these could upset our analyses.
Are there any missing values in the variable s1gender? Do we need to adjust this variable at all?
Select Analyze, Compare Means, and then Means.
Move s1gcseptsnew to the Dependent List box and s1gender to the Independent List box. Click OK.
Your output should look like the table on the right.
Here, we’re given the total number of survey respondents included in s1gcseptsnew, the number of survey respondents we’ve excluded (those 233 values we coded as missing during our univariate analysis), the percentage of the total survey represented by our GCSE score variable, the mean GCSE scores for men and women, the standard deviations of those means, and the number of respondents in each category.
What do you notice about the mean GCSE scores for men and women?
Is this mean difference what you might expect?
Summary
You have calculated the mean GCSE score for both males and females and interpreted the results. This is done so that you know the potential relationships within the data. Now, try to use the mean comparison method listed above to calculate the mean difference in GCSE scores for survey participants who answered s1targs2, a categorical variable cataloguing whether or not a young person held an A/AS level qualification during Sweep 1 of the YCS.
You can go on to use a t test to test the statistical significance of the difference in means for any explanatory (or independent) variables that are categorical with two categories.
Note: as we are making changes to a dataset we’ll continue using for the rest of this section, please make sure to save your changes before you close down SPSS. This will save you having to repeat sections you’ve already completed.