About the book


Statistical Tables




Exercises from Chapter 4

University students in Britain have been paying tuition fees for several years. The government reasoned that a university education leads to graduates benefiting from higher average pay when they start work.You want to investigate this issue to see if it is true.

Using the GHS, look at a frequency distribution of the variable EDLEV. Recode EDLEV into three groups; those with higher qualifications (including teaching and nursing); those with any qualifications up to A levels, and those with no qualifications. Call this new variable, QUALS.

Select only those in full-time employment using Data |Select Cases …, and use WKSTATE=1 as the selection criterion.

Then recode average weekly pay (PAY) into two categories: those earning above the £350 per week and those earning below that amount. This cut-off value is around the median value of weekly pay. Call this new variable SALARY.

Create a clustered bar chart where the category axis is SALARY and define the clusters by QUALS.

Make sure you select Bars Represent … % of cases. Does gaining a degree or higher qualifications improve your chances of a higher salary?

Look at the same problem but separately for men and women. This is done by selecting first men working full-time (WKSTATE=1 and SEX=1) and repeating the bar chart, and then women working fulltime (WKSTATE=1 and SEX=2) and repeating the bar chart. Are there any differences for men and women?