Are particular school subjects associated with better performance at university?

5. Discussion

In general, the choice of subjects at level 3 of NCEA is not strongly associated with university performance; instead, university performance is more closely associated with how well a student achieves at school, more or less independently of what subjects are studied at school. In addition, this relationship appears to be more or less independent of what is studied at university.

These results may appear to be at odds with the findings of most other research, where some school subjects—in particular mathematics—have been thought to provide some benefit in a range of degree studies. Our study suggests that this finding is likely to be due to selection effects; that is, the students taking mathematics tend to have higher ability, and it is the higher ability that is associated with the good performance at university, not the taking of mathematics.

This is not to argue that specific skills or knowledge gained in particular school subjects are not important in degree study. Our results show that accounting at school appears to be associated with higher performance in management and commerce degrees, particularly in studies in accountancy. The closer the link between the subject area of the school subject and the university study, the more likely there is an association. We have demonstrated this for mathematics and mathematical sciences, chemistry and chemical sciences, and English and law. In any discipline, if there are prerequisite skills or knowledge required of a student, especially if these are fundamental to the particular area of study, then students with those skills and knowledge will be expected to perform better.

In contrast, we have also shown that taking chemistry at school is also associated with better performance in mathematical sciences, and language and literature studies, while accounting at school is associated with better performance in mathematical sciences, but not the reverse. And the example of how accounting is positively associated with university performance in accounting, but negatively with performance in creative arts studies, whereas creative arts students at school show the opposite relationship with accountancy and creative arts studies at university, indicates that factors over and above subject-matter content are likely to be involved.

In addition, the more detailed the examination of the relationship between school subject and university performance, for example, considering two school subjects and performance in a narrow area of study (section 4.6), the more complex the findings. No consistent pattern emerges, even when one school subject is strongly associated with better university performance. For example, when controlling for school achievement, taking biology at school is associated with better performance in accountancy studies at university, but taking mathematics or economics at school is not. These counter-intuitive results suggest that more factors are involved in the relationship between school subjects and university performance than we have been able to include in our analysis.

Numerous studies have shown that prior academic achievement at school is the strongest predictor of university performance, especially in the first year of tertiary study. Even when particular circumstances modify this relationship—students taking a gap year, or studying in some particular fields of study at university—so that school achievement becomes a less strong predictor, it is still the main predictor, and this is true for the majority of students (Engler 2010). Whether a student took a particular subject or not makes only minor differences to university performance once we also control for achievement in a second subject, and it does not seem to matter which two subjects we consider. And when we look at achievement across a number of subjects, the differences in university performance become even smaller. We conclude therefore, that academic achievement at school, whether it is measured as an average over all school subjects, or for an individual subject, has the strongest association with first-year university performance. The association, if any, which arises from simply taking a subject, is relatively smaller.

It is worthwhile briefly considering the results of other work in this area, and contrasting their results with those found in this study.

Rauchas et al (2006) find results similar to ours, albeit with a proviso. They considered first-year computer science students in South Africa, and found that high school mathematics had a weak positive correlation with performance in computer science courses. At the researchers’ university, mathematics results are used as the primary criterion for admitting students into their computer courses. However, they found that the computer science students had high drop-out and failure rates, even for students who had taken mathematics. They cite Campbell and McCabe (1984) who found that a single high school subject is not useful for predicting success in computer science, but that a better indicator of success is an overall average of the high school results.²⁴ However, Rauchas et al go on to show that English, when taken as a first language, is a better predictor, but make the point that they believe that it is not about English in particular, but about language appreciation and its use in general, that is the underlying factor.

In our study, we found no association in performance in information technology studies and taking mathematics with calculus at school, after controlling for school achievement.²⁵

A second study also supports our findings. Peard (2004) studied the mathematical background of students entering the first year of a Bachelors of Education (primary) at the Queensland University of Technology. He found that, while the entering students had different levels of high school mathematics, there was no justification for denying entry to the course based on their lack, or otherwise, of year 12 mathematics. Peard had small numbers of students in his study, and he noted that his finding may only apply to primary teacher education. When we looked at teacher education students in our cohort,²⁶ also a small sample, we found that not having taken biology, chemistry or physics resulted in significantly lower university performance, but that whether a student took mathematics or not made no statistical difference.

Alcock et al (2008) considered the influence of secondary mathematics on the performance of students in introductory business courses in Australia. They indicate that there is a ‘clearly-established benefit’ of studying secondary school accounting for tertiary accounting students, but almost no advantage to performance in accounting and finance courses from having studied mathematics, despite the fact that mathematics is often required as a pre-requisite for such courses. This study did try to control for prior student achievement, using an inter-tertiary university entrance score, and their study only included students in the top 10 per cent of school achievement. They found that high school mathematics was a good predictor of success in introductory business coursework and business law. But this study did not consider student achievement in the high school subjects, and did not consider the case for students who did not take mathematics.

Our study found no association between the performance in accountancy degrees and taking mathematics at school, after controlling for school achievement. There was a weak association for performance in economics courses. Rather, it was accounting at school that was strongly associated with performance in accountancy courses.

Kok (2007) considered the influence of secondary school mathematics on the study of law in South Africa. Kok finds that it is students with mathematics and physical science at matriculation²⁷ that outperform students who do not take these school subjects. Kok finds that while A and B standard students in languages outperform the average law student, ‘even D and E candidates in mathematics (HG) and science (HG) perform better than the average’. Kok acknowledges that the mathematics and sciences students are probably an ‘elite’ group in terms of academic ability or ‘intelligence’, but does not control for this.

Our study found no association between performance in law degrees and taking mathematics at school, when controlling for school achievement.

Sadler and Tai (2007) analysed students enrolled in tertiary courses in biology, chemistry and physics in the United States, controlling for years of instruction in high school biology, chemistry, physics and mathematics, amongst other factors. They found that high school biology helped in biology courses at university, chemistry helped in chemistry, and physics helped in physics, with no cross-disciplinary effect, but that mathematics helped in each of the tertiary fields of study, including biology. Sadler and Tai do control for student achievement, using SAT/ACT exam scores,²⁸ and the last high school grade in mathematics and English. Interestingly, their results show that students’ tertiary grades were significantly associated with the students’ SAT/ACT exam score, in addition to the years of instruction in the high school subjects.

When we considered the same group of disciplines as Sadler and Tai, we found somewhat different results. In our study, we found like-for-like associations for all disciplines except school physics and physics and astronomy courses, and that taking mathematics at school was only associated with higher performance in mathematical science. We also found that chemistry at school was associated with better performance in mathematical science.

Most of our results are based on broad fields of study assigned to degrees at university, such as management and commerce, society and culture, and physical and natural sciences. Most of the previous literature on this topic has considered university performance in courses within degrees, such as Introductory finance, Fundamental algorithmic concepts, or Business law. There may be a stronger association between the subjects taken at school and the field of these courses than to the range of courses taken by a student as part of a higher degree. However, when we considered specific fields of study, such as mathematical science, chemical science, or law, we found essentially the same results as when we used broad fields of study.  It appears that only for related topics—mathematics and mathematical science, chemistry and chemical science, for example—is a particular school subject associated with an increase in university performance, but the increase in performance is more often than not marginal.

There are three main implications arising from this study.

• Firstly, there are implications for universities, and the changes some of them are making to student selection rules.
  
• Secondly, the results of this study have implications for the New Zealand Qualifications Authority, which is currently reviewing the university entrance requirements. Part of that review is to consider what form the common entrance standard should take, and what it might comprise. The results of this study and our previous analysis (Engler 2010) provide some evidence to inform this process.
  
• Thirdly, school students, and by implication their teachers and parents, may find some relevance in these results, particularly in regard to their motivation to do well at school.

Universities are facing high levels of demand for degree level study, but their enrolments are constrained by the number of places funded by government. In response, some universities are altering their general admissions criteria. While young students are still required to meet the university entrance requirement, several universities have indicated they will now also give preference to students with higher levels of school achievement. While these changes will generally identify students more likely to perform well at university, the findings of our earlier study (Engler 2010) suggest the proposed changes will disadvantage some identifiable groups of below-average students at school who, counter-intuitively, do well at university. That study found that some lower-achieving school students from low-decile schools may actually out-perform higher-achieving school students from other schools. The salient point is that school achievement is generally a good predictor of university performance, but not in all cases. The present study also suggests that any requirement for achievement in a particular school subject is also not necessary, at least for students who have met the university entrance requirement, since good achievement in one subject can offset poorer achievement in another. At one institution at least, level 3 chemistry is a prerequisite for enrolment in chemical science courses at stage one. Our results would suggest this is unnecessary. Students perform almost equally as well in these degrees whether they took school chemistry or not. The better indicator of performance is how well a student achieved at school, irrespective of what subjects they took.

The NZQA is reviewing the university entrance requirement. Our finding that the school subjects a student has taken are only weakly associated with university performance, if at all, may be important to that review. Our findings suggest a university entrance requirement based on student achievement would be more appropriate. But, basing university entrance solely on school achievement is not going to select all students who are likely to demonstrate high levels of performance at university.

Lastly, the findings are relevant to school students, their caregivers and teachers. There is evidence to suggest that when the NCEA was first introduced in 2002, students did only enough work to gain the amount of credits needed to achieve a particular qualification level (Meyer at al 2006). There was no advantage to them in working harder to gain credits since a credit earned with an achieved grade counted equally toward their credit totals as did a credit earned with a merit or excellence grade.

In 2007, the NCEA reporting system was changed to include endorsements on certificates of school achievement. Previously, a certificate only showed that a student had gained a particular NCEA level. With the change, the certificate also indicated whether the student achieved the NCEA level with merit or excellence. This had the effect of generally increasing student motivation (Meyer et al 2009); only about 10 per cent of students surveyed indicated the change did not matter to them.

The knowledge that university performance is more closely linked to the level of achievement may motivate some students. This will be reinforced by the higher entry requirements being imposed by the universities.

Engler (2010) discussed the factors that influence university performance, which included:

• motivation
  
• self-discipline
  
• confidence
  
• study habits
  
• time management skills
  
• family and peer support
  
• attending an institution of choice
  
• studying preferred courses or subjects.

Engler concluded that the NCEA level 3 achievement score, which is used to measure school achievement in that study and this one, is a proxy for some or all of the factors listed above. In general, these factors are independent of what is being studied.²⁹ Motivated, self-disciplined students with good study habits and time management skills will perform well at school, and these same attitudes and traits will stand them in good stead at university. Certainly, there needs to be an adjustment to university life and its study regime, but ultimately, successful students are those that learn new material, and then demonstrate their mastery of that material in tests or examinations of one form or another. It is therefore not surprising that the students who do well at school do well at university, nor is it surprising, that this is essentially independent of the subjects taken at school.

Footnotes

24. What Campbell and McCabe actually said was that persistence in a computer science, engineering, or other science programme is related to students’ mathematics and English scores in high school, their overall high school rank, and their background in high school mathematics and science.
25. Results not presented.
26. Results not presented.
27. The subjects needed to be taken at the Higher Grade (HG) standard, compared to the Standard Grade.
28. SAT is the Scholastic Aptitude (or Assessment) Test, a standardised test for college admission, and ACT is American College Testing, a standardised test for high school achievement and college admission.
29. The exception is the last factor, studying preferred courses or subjects. Clearly, having an interest in the subject matter helps motivation. But it would be wrong to conclude that, say, only mathematics students prefer to study mathematics.