Are particular school subjects associated with better performance at university?

3. School achievement and school subject choice

There is an association between achievement at school and students’ choice of subjects.

The data available in this study allows us to compare, on average, student achievement in one year with the selection of subjects in the following year. Accordingly, we have used school achievement at NCEA level 2 and compared it to the choice of subject at NCEA level 3.

Figure 2 shows the frequency distribution of students across the range of NCEA level 2 achievement, calculated across all level 2 subjects, for students who did or did not take NCEA level 3 English.

Figure 2: Frequency distribution of students who did, or did not take NCEA level 3 English, against overall school achievement in NCEA level 2

It can be seen that students with achievement scores less than 50 were slightly less likely to take English, while students with achievement scores of 50 or higher were slightly more likely to take English. Although largely overlapping, the two distributions have different average level 2 achievement scores (table 1).

This situation can be contrasted with figure 3, which shows the distribution for students who did or did not take the subject mathematics with calculus⁶ as a level 3 subject. Students with above-average school achievement at NCEA level 2 are far more likely to take mathematics. In other words, the level of achievement of students taking mathematics is, on average, higher than students who did not take this subject.

The type of pattern depicted in figure 2 for English is also seen for the level 3 subjects statistics and modelling, economics, and accounting. The type of pattern seen in figure 3 for mathematics is also seen for the subjects physics, chemistry, and other languages.⁷ The frequency distribution for biology falls somewhere between that of English and mathematics, with a clear ‘shift to the right’, but is not as pronounced as that seen for mathematics. For humanities,⁸ there is a ‘shift to the left’. That is, students who chose to study humanities subjects were slightly more likely to have below-average achievement in their level 2 studies. For visual arts subjects,⁹ this ‘shift to the left’ is more pronounced. Table 1 shows the average level 2 achievement score for the various level 3 subjects.

Figure 3: Frequency distribution of students who did, or did not take NCEA level 3 mathematics with calculus, against overall school achievement in NCEA level 2

 
What these results show is that students with higher levels of school achievement are more likely to take particular subjects. These subjects are chemistry, other languages, physics, and mathematics with calculus. We would expect, given the relationship between school achievement and university performance, that students taking these particular school subjects would, on average, show higher levels of university performance across a range of bachelors-level study courses. For the average student taking these subjects, this difference in school achievement translates into a 7 or 8 percentage point increase in the likelihood of passing most first-year courses at university (extrapolated from figure 1).

It is not coincidental that higher school achievement is seen for particular groups of subjects. Table 2 shows the correlations between pairs of subjects taken by students. It can be seen that taking mathematics with calculus is positively associated with also taking physics, chemistry, economics and accounting, but negatively associated with taking statistics and modelling, English, humanities and visual arts. There is no statistical association between mathematics and other languages. It is likely to be much the same group of higher-achieving students taking a particular group of subjects.

It can also be seen that the correlations between subjects are generally small. The largest correlation, between physics and chemistry, is +0.46. Clearly, while there are ‘natural’ groupings of subject choices, in reality, there is a very large variety of subject combinations taken by students. This makes it somewhat problematic to determine the impact a single school subject has, in isolation, on university performance.

Footnotes

6. In this report, the shorthand term ‘mathematics’ will sometimes be used to refer to the subject ‘mathematics with calculus’.
7. Other Languages include French, German, Spanish and Latin.
8. Humanities, as defined in this study, include geography, history, art history and classical studies.
9. Visual arts subjects include photography, printmaking, design, sculpture and painting.