Hei titiro anō i te whāinga: Māori achievement in bachelors degrees revisited

From school to degree study

How well did Māori degree students achieve at school?

The first question to examine is how well Māori who enter degree studies have achieved at school relative to their non-Māori peers. The NCEA data shows that, on average, Māori are starting degree studies with a lower level of school achievement.

Māori were more likely to enter degree-level studies without having attained the NCEA university entrance requirements. Twenty-four percent of Māori first-year degree students who had finished school in 2004 or 2005 did not have university entrance, compared with 16 percent of non-Māori. Similarly, Māori students were less likely to have achieved a level 3 NCEA qualification, with 75 percent of Māori having their highest qualification at this level, compared with 85 percent of non-Māori.

Māori students going into degree studies had also achieved a lower expected percentile in their highest level of NCEA than their non-Māori peers. For Māori first-year degree students who had finished school in 2004 or 2005, the mean expected percentile was 48 percent, compared with 54 percent for non-Māori.¹ The distribution for Māori students was also more heavily weighted to the lower end, as shown in Figure 4.1

Figure 4.1 Distributions of expected percentile at highest level of NCEA for first-year degree students

Note: Results are for students who finished school in 2004 or 2005 and entered degree study in 2005 or 2006.

How does school achievement influence tertiary achievement?

Scott (2008) demonstrated that the mean pass rate of first-year degree students with the same expected percentile in NCEA Level 3 achievement standards was strongly related to their expected percentile. This relationship is best described using a logarithmic curve, which means the effect diminishes as expected percentile increases.

The data used for this revised study shows a similar relationship using the expected percentile for the highest level of NCEA achievement standards studied at school. It also shows that while the nature of the relationship is generally similar for Māori and non-Māori, there are two major differences. The first is that Māori with the same expected percentile will have an average pass rate that is 7.6 percent lower than that of non-Māori.² The second is that there may be a drop in first-year performance for Māori first-year students who performed very well at school, that is with expected percentiles of greater than 80 percent. The numbers at this level are too small to provide any conclusive findings at this stage.

Figure 1 2: Expected percentile at highest level of NCEA by average first-year degree course pass rate

Note: Students enrolled in only one or two courses are excluded. Analysis is limited to expected percentiles in the range from 20 to 90 percent. Outside this range there are too few observations for meaningful interpretation. Results are for students who finished school in 2004 or 2005 and entered degree study in 2005 or 2006.

Scott (2008) also looked at the correlation between expected percentiles at each level of NCEA study and individual course pass rates for 2004 school leavers. He found a moderate correlation at each level, with the level 3 expected percentile having the strongest correlation. The same analysis was run on the dataset for this study which contains results for 2005 and 2004 school leavers, with a comparison of Māori and non-Māori.

Note: The correlations were determined using Pearson’s correlation. A coefficient of 1 means the values are completely related and a value of 0 means there is no relationship.

Table 4.1  shows that the correlations for non-Māori are similar to those reported by Scott for all students, although with slightly less differentiation between levels.³ However, for Māori students the correlations are lower at each level, with level 2 having the strongest correlation to first-year pass rates. This suggests an area of further investigation into the performance of Māori students within NCEA and its relationship to tertiary performance.

Moving from school to degree study

In 2006, there were 1,560 first-year Māori students aged under 20 in bachelors degrees. The number of Māori students aged under 20 has been steadily increasing since 2002 by an average of 4 percent per year. The participation rate⁴ in bachelors degrees for Māori aged 18-19 has increased from 8.7 percent in 2002 to 9.6 percent in 2006. Over the same period the participation rate for all students in this age group increased from 21.6 to 23.1 percent.

Just under 40 percent of Māori first-year degree students aged under 20 are male. Most of these students (81 percent in 2006) study at universities, with the next largest group studying at institutes of technology and polytechnics (12 percent). Three-quarters of the students went into degree study directly from school. Around 17 percent studied in a lower-level tertiary qualification beforehand and 6 percent were in the workforce between school and degree study.

Figure 4.3: Māori first-time, first-year students aged under 20 by gender

Ussher (2007) noted that Māori students who left school in 2004 were less likely to move into degree studies than students in other ethnic groups, including Pasifika, who had the same level of qualification. He found that of those students leaving with a university entrance qualification, 70 percent of Māori went on to degree study within two years, compared with 77 percent of Pasifika, 81 percent of Asian and 82 percent of European students.

First-year pass rates

As noted in the original study, Māori students moving from school to tertiary are less likely to pass all of their courses than non-Māori. Scott (2008) showed that this disparity remains, even once all other measurable factors were controlled for.

The new model using matched NCEA results shows that for Māori students, their level of performance at school, as measured by the expected percentile, has the largest association with their success in their first-year studies. The model shows that for Māori a 5 percent increase in expected percentile will result in about a 6 percent increase in the probability of passing 75 percent or more of first-year courses.

The subject of degree study was the second most important factor. Once school performance of students is controlled for using the expected percentile, there were few differences across most subjects. The exceptions were information technology, where Māori students were less likely than Māori students in other subjects to pass 75 percent or more of their first-year courses, and education and graphic design, where Māori students were more likely than Māori students in other subjects to pass 75 percent or more.

The differences shown in the original study between specialist and generic qualifications were no longer apparent once school performance was more fully controlled for. This supports the supposition made in the earlier study that higher rates of success in specialist qualifications were likely to be due to the higher ability of students entering these subjects.

Highest school qualification still had an effect in the model, even once expected percentile was controlled for. Students with NCEA Level 3 or above were much more likely to succeed than those with lower-level qualifications, including students who had achieved the university entrance requirements but not a level 3 qualification.

The new model also showed that Māori students who studied at a non-university provider generally had better pass rates than students studying at universities, once other factors were controlled for. A similar finding was reported for all students in Scott (2008).

A more detailed picture can be provided by looking at the proportion of students at the provider at bachelors level and above. This provides a measure of the intensity of degree-level and above provision. At non-university providers, there was a slight increase in success at providers with a greater proportion of students at bachelors level or above. The opposite was the case at universities. That is, the greater the proportion of students at bachelors level and above, the less likely Māori first-year students were to pass 75 percent or more of their courses.

Figure 4.4: Predicted probabilities of Māori students passing 75 percent or more of first-year degree courses by percent of students at bachelors level and above at the provider

Comparing mean pass rates for Māori and non-Māori with the percentage of students at bachelors level and above at universities and other providers shows a similar pattern. At non-university providers, the mean pass rate increased for both Māori and non-Māori with the percentage of students at bachelors level and above, with Māori pass rates being slightly lower.⁵ At universities, there was a fairly flat relationship for non-Māori students, but a decreasing relationship for Māori students.⁶ Alternatively, the universities could be read as having two clusters, with the left-hand cluster having a smaller difference between Māori and non-Māori students than the right-hand one.

Figure 4.5: Mean first-year pass rate by percentage of students at bachelors level and above at the provider

Non-University

University

Note: Results are for students who finished school in 2004 or 2005 and entered degree study in 2005 or 2006. The percentage of students at bachelors level and above is grouped into bands of 2.5 percent to calculate the mean pass rates. This means that some institutions are combined within one observation.

These findings suggest that degree students at providers other than universities may benefit from the presence of a greater proportion of degree and above students. However, this may not hold at universities, particularly for Māori students. Several possible effects may be occurring. Universities with below-degree-level students may be providing better support to Māori students. Also, universities with mostly degree and above students may have a greater focus on supporting postgraduate students than on supporting first-year degree students. In addition, these universities may set higher standards for their first-year degree courses and have higher expectations of students to take responsibility for their own performance.

The new model also showed that the small group of Māori students who go overseas between school and degree studies do considerably better in their first year, even once their school performance has been controlled for. No significant differences were found between students going directly from school and those taking a break within New Zealand, including undertaking lower-level study.

Students who attended Māori boarding schools were shown to be less likely to pass 75 percent or more of their first-year courses than students from other schools, once qualifications and NCEA results were controlled for. Students from kura kaupapa Māori were somewhat less likely than other Māori students to pass all of their courses. Scott (2008) also noted that, for all students, students from single-sex schools and private schools did somewhat less well at first-year degree level than students from other schools, once other factors were controlled for.

Return to study

Māori students aged 18 to 19 were less likely to return to degree study after one year than non-Māori students. The first-year retention rate in 2005 for Māori degree students in this age group was 81 percent, compared with 87 percent for all students. Scott (2008) found that Māori students were less likely to remain in degree study after one year than other students, even once other factors were controlled for. Retention rates have been fairly similar for Māori men and women in this age group, with men being slightly more likely to remain in study.

The matched NCEA data was used to create a single-year model of return to study for Māori students who left school in 2004 and studied at degree level in 2005. This model found three statistically significant factors associated with return to study, which were first-year pass rate, degree subject and the percentage of students at degree level and above at the provider. No significant effects were found related to NCEA results. The effect of school performance was fully absorbed by the first-year degree performance.

The original model was also rerun to create a four-year model of Māori degree students under 20 who started study between 2002 and 2005. This model came up with very similar results. The larger population size enabled a few other effects to show as significant. The results of both models are discussed together below.

First-year pass rate is by far the most significant factor associated with return to degree study after one year. Those who passed 75 percent or more of their first-year courses were significantly more likely to continue in study.

The major subject of degree study does have some effect on the chances of returning to study. Both models showed the differences are moderate. In both cases, education students were less likely to return than students in other subjects, even though they had a better chance of passing all of their first-year degree courses.⁷

Both models showed that Māori students who attended providers with a higher proportion of students at degree level and above were more likely to return to study, having controlled for their first-year pass rates.⁸ This is in contrast to the result for first-year pass rates at universities, which decreased as the proportion of students at degree level and above increased. This suggests that while students at universities with a high proportion of degree level and above provision are less likely to pass all of their first-year courses, those who do pass are more likely to continue in their study.

The other effects that came through in the four-year model were that:

• full-time, full-year students were somewhat more likely to continue in study than other students
• having a school qualification at NCEA Level 3 or above slightly improved the chances of continuing in study
• students from larger secondary schools were slightly more likely to continue in study.

Completion

Māori students who started study at the age of 18 or 19 were less likely to complete a degree than non-Māori students starting at the same age. Of students in this age group who started in 2002, 52 percent of Māori students had completed by 2006, compared with 65 percent of all students. Māori women in this age group were more likely to complete, at 57 percent, than Māori men, at 43 percent.

The revised model of completion included data for two entry cohorts, whereas the original only included one entry cohort. The model confirmed that total course pass rate was the most significant determinant of completion, with those passing at least 75 percent of their courses being significantly more likely to complete.

The model also showed that students taking more than one degree qualification in the period were less likely to complete any one of the qualifications. Students with more than one qualification include those enrolled concurrently in different degree programmes⁹ and students who switch from one programme to another. The model also showed that the greater the number of years a student was enrolled in degree study, the less likely the student was to complete a qualification. In other words, those who fail several courses, switch degree programmes and/or study part-time are less likely to complete successfully, even after completing at least three years of equivalent full-time study.

Students at providers with a greater proportion of provision at postgraduate level were slightly more likely to complete their degree. This reinforces the finding with regard to return to study that students in providers with a higher proportion of provision at degree level and above are likely to be more persistent in their study.

It was also found that students at wānanga were less likely to complete than students in other sub-sectors. There was no significant difference across the other sub-sectors.

Footnotes 

1. A one-way analysis of variance showed this difference in means to be statistically significant with a p-value of <.0001. Welch’s ANOVA was used, as the assumption of equal variance was rejected by the Levene’s test for homogeneity of variance.
2. A two-way analysis of variance using log of expected pass rate and Māori vs non-Māori showed that the mean pass rate for Māori is 7.57 percent lower than for non-Māori. This difference had a p-value of 0.0162.
3. This difference may be due to including ‘gap year’ students in the current study, who are not included in Scott’s study.
4. This is the number of students enrolled in bachelors degrees as a proportion of the total population.
5. A two-way analysis of variance using percentage of students at bachelors level and above and Māori vs non-Māori at non-university providers found a significant overall positive relationship between percentage of students at bachelors level and above and mean pass rates, and that pass rates for Māori were likely to be 6.1 percent lower than those of non-Māori. However, the difference between Māori and non-Māori was not statistically significant.
6. A two-way analysis of variance using percentage of students at bachelors level and above and Māori vs non-Māori at universities found a significant overall negative relationship between percentage of students at bachelors level and above and mean pass rates, and that pass rates for Māori were likely to be 8.9 percent lower than those of non-Māori.
7. This finding may be confounded by the merger of four colleges of education with universities during this period. It is possible that these changes had a negative effect on retention rates during the period of changeover.
8. Interacting this variable with sub-sector in the model rendered both the main effects and the interaction non-significant. Sub-sector was significant in the model if the percentage of students at bachelors level and above was excluded. Both variables are probably picking up a similar effect.
9. However, it excluded registered conjoint degree programmes.