Report to Te Rūnanga o Ngāi Tahu: Māori Educational Achievement in the Christchurch Health and Development Study

David M. Fergusson, Geraldine F.H. McLeod, Te Maire Tau, Angus H. Macfarlane Copyright © 2014 David M. Fergusson, Geraldine F.H. McLeod, Te Maire Tau, Angus H. Macfarlane

The moral rights of the authors have been asserted. ISBN 978-0-473-27220-3

A catalogue record for this book is available from the National Library of (Te Puna Matauranga o ). This book is copyright. Except for the purpose of fair review, no part may be stored or transmitted in any form or by any means, electronic or mechanical, including recording or storage in any information retrieval system, without permission in writing from the publisher.

The opinions expressed and conclusions drawn in this Working Paper are solely those of the authors. They do not necessarily represent the views of the Ngāi Tahu Research Centre, University of Canterbury. For further information or additional copies of the Working Paper, please contact the publisher.

Ngāi Tahu Research Centre University of Canterbury Private Bag 4800 Christchurch 8140 New Zealand Email: [email protected] www.ntrc.canterbury.ac.nz

Cover: Background image – from Tom Green’s notebook (c. 1860s, Christchurch, University of Canterbury, Macmillan Brown Library, Ngāi Tahu Archives, M 22.) Report to Te Rūnanga o Ngāi Tahu: Māori Educational Achievement in the Christchurch Health and Development Study

Authors David M. Fergusson Christchurch Health and Development Study at the University of Otago, Christchurch Geraldine F.H. McLeod Christchurch Health and Development Study at the University of Otago, Christchurch Te Maire Tau Ngāi Tahu Research Centre, University of Canterbury, Christchurch Angus H. Macfarlane Professor of Māori Research, University of Canterbury, Christchurch

Acknowledgements This research was funded by grants from the Health Research Council of New Zealand, the National Child Health Research Foundation, the Canterbury Medical Research Foundation and the New Zealand Lottery Grants Board. Contents

Tables ...... 1 Figures ...... 2 Executive summary ...... 4 Chapter 1 Introduction ...... 8 Chapter 2 The early childhood educational experiences of the Māori cohort ...... 19 Chapter 3 School educational achievement (7-13 years) ...... 27 Chapter 4 Secondary school and tertiary educational attainment and participation (15-30 years) ... 36 Chapter 5 Summary, conclusions and recommendations ...... 45 References ...... 54 Appendix ………………………………………………………………………………………………………………………………….… 57

i Tables

Table 1.1 Number and percentage of Māori cohort members based on parental ethnicity...... 10 Table 1.2 Christchurch Health and Development Study cohort members identified as Māori at birth and interviewed at each assessment age, up to age 30...... 10 Table 1.3 Self-reported ethnic identification of the entire CHDS cohort studied at 21 and 25 years. 11 Table 1.4 Number and percentage of cohort members of Māori descent (n=169) by self-reported affiliation by region………………………………………………………………………………………………………………………… 16

1 Figures

Figure 1.1 Maternal educational attainment characteristics of the Māori cohort (n=142) and non- Māori cohort (n=1,123) defined at birth...... 13 Figure 1.2 Family socio-economic status of the Māori cohort (n=142) and non-Māori cohort (n=1,123) defined at birth...... 13 Figure 1.3 Family type of the Māori cohort (n=142) and non-Māori cohort (n=1,123) defined at birth...... 14 Figure 1.4 Maternal age for the Māori cohort (n=142) and the non-Māori cohort (n=1,123) defined at birth...... 14 Figure 2.1 Percentage of the Māori cohort (n=123) who attended Early Childhood Education (ECE) from age 2 to age 5 years, and who had ever attended ECE...... 20 Figure 2.2 Percentage of the Māori cohort (n=123) attending Early Childhood Education (ECE) by duration of attendance (in years)...... 20 Figure 2.3 Percentage of the Māori cohort (n=123) who had ever attended Early Childhood Education (ECE) (n=110) by type of facility chosen...... 21 Figure 2.4 Percentage of the Māori cohort (n=123) and the non-Māori cohort (n=975) attending Early Childhood Education (ECE) at each age, from 2 to 5 years, and who had ever attended ECE...... 22 Figure 2.5 Mean number of years of Early Childhood Education (ECE) attendance by the Māori cohort (n=123) and the non-Māori cohort (n=975)...... 22 Figure 2.6 Adjusted percentages of Early Childhood Education (ECE) attendance for each age group from 2 to 5 years, and those who had ever attended ECE by the Māori cohort (n=123) and the non- Māori cohort (n=975)...... 24 Figure 2.7 Adjusted means for the number of years of Early Childhood Education (ECE) attendance by the Māori (n=123) and non-Māori cohorts (n=975)...... 25 Figure 3.1 Mean teacher-rated Grade-Point Average (GPA) scores (ages 7, 10 and 13) for the Māori and non-Māori cohorts...... 28 Figure 3.2 Mean for the standardised test scores of the Burt Word Reading Test (ages 8, 11, 13) for the Māori and non-Māori cohorts...... 29 Figure 3.3 Mean for the standardised test scores of the Progressive Achievement Test (PAT) in Mathematics (age 11) and Reading (age 12) for the Māori and non-Māori cohorts...... 29 Figure 3.4 Adjusted mean teacher-rated Grade-Point Average (GPA) scores (ages 7, 10 and 13) for the Māori and non-Māori cohorts...... 30 Figure 3.5 Adjusted mean standardised test scores of the Burt Word Reading Test (ages 8, 11, 13) for the Māori and non-Māori cohorts...... 31 Figure 3.6 Adjusted mean standardised test scores of the Progressive Achievement Test (PAT) in Mathematics (age 11) and Reading (age 12) for the Māori and non-Māori cohorts...... 32 Figure 3.7 Mean teacher-rated Grade-Point Average (GPA) scores (ages 7, 10, and 13) by cultural identity at ages 21 and 25...... 33 Figure 3.8 Mean Burt Word Reading Test scores (ages 8, 11, and 13) by cultural identity at ages 21 and 25...... 33 Figure 3.9 Mean Progressive Achievement Test (PAT) scores (ages 11: Mathematics, and 12: Reading) by cultural identity at ages 21 and 25...... 34

2 Figure 4.1 Secondary and tertiary educational attainment for the Māori and non-Māori cohorts. ... 37 Figure 4.2 Tertiary educational participation among the Māori and non-Māori cohorts...... 39 Figure 4.3 Educational attainment of the Māori and non-Māori cohort following statistical adjustment for socio-demographic factors...... 40 Figure 4.4 Educational participation of the Māori and non-Māori cohort following statistical adjustment for socio-demographic factors...... 41 Figure 4.5 Educational attainment for the Māori cohort by cultural identity at age 21 and 25...... 42 Figure 4.6 Educational participation for the Māori cohort by cultural identity at age 21 and 25...... 43 Figure 5.1 Stem-and-leaf plot of unadjusted and adjusted standardised differences (Cohen’s d) between the Māori and non-Māori cohorts...... 47

3 Executive summary

Overview

This report has been prepared for Te Rūnanga o Ngāi Tahu by members of the Christchurch Health and Development Study (CHDS) in conjunction with Professors Te Maire Tau and Angus Hikairo Macfarlane. The general aim of this report was to provide an account of the educational history of a cohort of Māori studied from birth to the age of 30 as part of the CHDS. The CHDS is a longitudinal study of a birth cohort of 1265 children who were born in the Christchurch/Ōtautahi region during mid-1977. Of these children, 142 were classified as Māori. The aims of the report were to address three key issues relating to the educational history of the Māori cohort.

• To provide a description of the educational history of the Māori cohort and the non-Māori cohort. • To examine the extent to which any educational disparity between the Māori and non-Māori cohorts could be explained by socio-economic differences between the two cohorts. • To examine the role of cultural identity in the educational achievement of the Māori cohort.

The key findings of the report are summarised below.

Chapter 1 Introduction

This chapter sets the scene for the report by providing a description of the Māori cohort at birth. The main findings of this analysis were:

• On the basis of parentally reported descent, 11.2% of the cohort was defined as being of Māori descent. • The CHDS cohort members were studied to the age of 30, at which time 111 Māori cohort members were studied, with this group representing 78.2% of the initial Māori cohort. • Rates of sample retention in the CHDS to the age of 30 did not vary with ethnicity. • Rates of sample loss for the Māori cohort did not vary with a series of measures assessed at birth. These measures included: maternal education, family socio-economic status at birth (Elley & Irving, 1976), family type, maternal age, gender, birth weight, planning of pregnancy, and whether the child was breast fed. • In addition to the measures of ethnicity collected at birth, all sample members were questioned at age 21 and 25 on their ethnic identity using questions from the 1996 New Zealand Census of Population and Dwellings. At each age, nearly 12% of the cohort reported some Māori identity. Of these, about 5% reported a sole-Māori identity and 6% to 8% reported a Māori/other ethnic affiliation. • Measures of Māori ethnicity based on parental report of descent at birth, and reports of self-identity at 21 and 25, were generally strongly correlated (r = 0.78-0.87; p<.05). These findings suggest that it is valuable to compare results using differing definitions of ethnicity. • Examination of the socio-demographic features of the Māori cohort at birth showed that in comparison to the non-Māori cohort, this group was subject to relative disadvantage in the

4 areas of: maternal education (p<.001), family socio-economic status at birth (p<.001), family type (single parent or two parent family) (p<.001), and younger maternal age (p<.001). • While the Māori cohort was disadvantaged relative to the non-Māori cohort, by no means did all Māori children come from a disadvantaged background. • Of those reporting Māori descent at age 21 and 25 (n=169), 74.0% were able to nominate at least one iwi affiliation. The predominant affiliation nominated was Ngāi Tahu/Kāi Tahu (42.0%). It is notable that of the nominated iwi affiliations, the majority (56.9%) were from iwi, with Ngāpuhi (11.8%) and (6.5%) being the most frequently mentioned. • At ages 21 and 25, of the 169 cohort members who reported being of Māori descent: 40.8% described themselves as sole-Māori, 38.5% described themselves as Māori/other ethnic affiliation and 20.7% as non-Māori.

Chapter 2 The early childhood educational experiences of the Māori cohort

The aims of this chapter were to examine participation in early childhood education (ECE) in the Māori and non- Māori cohorts. The key findings of this chapter were:

• Overall, there was a high rate of participation of Māori in ECE; nearly 90% of the Māori cohort attended ECE by age five. • ECE participation of the Māori cohort clearly reflected the availability of ECE services in the 1980’s, with the majority of cohort members attending kindergarten, Playcentre and day- care centres. Attendance at the then emerging Kōhanga Reo services was not recorded in this study, but it is clear that less than one-quarter of the Māori cohort would have used these services. • In comparison to non-Māori, the overall Māori ECE attendance rate was significantly lower; by age five, 89.4% of Māori had attended ECE, compared to 95.3% of non-Māori (p<.01). On average, Māori received 1.91 years of preschool education, compared to 2.13 years for non- Māori (p=0.007). • Further analyses suggested that these ethnic differences in rates of ECE participation were explained by socio-demographic differences at birth (maternal educational attainment, family socio-economic status, family type and maternal age). After statistical adjustment for these factors there were no detectable differences in the rates of ECE participation by Māori and non-Māori

Chapter 3 School educational achievement (7-13 years)

This chapter examined ethnic differences in educational achievement during the middle school years (7-13). The conclusions of this chapter are summarised as follows:

• There were consistent differences between the academic performance of the Māori and non-Māori cohorts. Māori were rated by their teachers as performing less well in key subject areas and performed less well on standardised tests of word recognition, reading comprehension and mathematics. These findings suggest that irrespective of the way in which achievement was measured (teacher ratings; standardised tests) there was consistent

5 evidence of Māori/non-Māori differences in educational achievement. However, these differences were relatively small; typically the differences between Māori and non-Māori were between .2 and .4 standard deviations. • Further analysis revealed that any differences between Māori and non-Māori achievement were accounted for by socio-demographic factors. Specifically, these factors included: maternal educational attainment, family socio-economic status (based on the Elley and Irving (1976) scale), family type (single parent/two parent households) and maternal age. When due allowance was made for these factors, no statistically significant associations remained between Māori/non-Māori and educational achievement. • Finally, the results showed that educational achievement between ages 7 and 13 years was not associated with cultural identity among the Māori cohort (categorised as Māori; Māori/other ethnic affiliation; or non-Māori) assessed at ages 21 and 25.

Chapter 4 Secondary school and tertiary educational attainment and participation (15-30 years)

This chapter provides an account of the educational history of Māori and non-Māori cohorts at the point of school leaving and beyond. The conclusions of this chapter are presented below:

• Consistent with the findings in previous chapters, there were small to moderate associations between ethnicity and educational achievement, with Māori having lower levels of attainment as measured by: attainment of any secondary school qualification by age 21; attainment of any tertiary qualification by age 30; and attainment of a Bachelor’s degree or higher by age 30. The most marked difference was for degree attainment, with non-Māori having a rate of degree attainment (before statistical adjustment) that was nearly three times higher than the rate for Māori. Māori also had lower rates of tertiary educational participation as measured by: participation in any tertiary education by age 30; and enrolment at a recognised university by age 30. • Nearly all of the associations between ethnicity and educational attainment/achievement were explained by differences in the socio-demographic backgrounds of the Māori and non- Māori cohort. However, even following statistical control a small but significant (p<.05) association remained between ethnicity and enrolment in any tertiary education. Significant socio-demographic factors included: maternal educational attainment, family socio- economic status and maternal age. These findings suggest that when due allowance was made for socio-demographic factors there was no consistent evidence to suggest that Māori ethnicity was associated with increased rates of educational underachievement. • An examination of the relationship between Māori cultural identity assessed at ages 21 and 25 and educational achievement/attendance showed that there were no detectable linkages between cultural identity and educational achievement/attendance.

6 Chapter 5 Summary, conclusions and recommendations

This chapter provides a summary and comment on the findings of the previous chapters. The major points made include:

• First, there was clear and consistent evidence of Māori educational disadvantage from early childhood through to tertiary education. These differences however, fell into the small to moderate range, with the implication that substantial numbers of young Māori were successful within the New Zealand education system. For example, in this study over two- thirds of young Māori left school with an educational qualification (see Chapter 4, Figure 4.1). • Adjustment for socio-demographic factors showed that nearly all of the associations between ethnicity and educational achievement/attendance could be explained by socio- demographic differences between the Māori and non-Māori cohorts. Key socio-demographic factors included: maternal educational attainment, family socio-economic status and maternal age. • There was no significant association between cultural identity and educational achievement/attendance over the life-course. Those of sole-Māori identity had similar levels of attainment to those holding Māori/other ethnic affiliation and non-Māori identities. • Collectively, these findings lead to two general conclusions. The first conclusion is that the major route to reducing ethnic differences in educational attainment is through social policies which foster the upward social mobility of young Māori. • At the same time it is important to recognise the role of the educational system in sustaining Māori language, culture and values. For these reasons, policies which seek to increase the upward social mobility of Māori need to be accompanied by policies aimed at sustaining Māori language, culture and values. • These considerations suggest the importance of carefully balancing two viewpoints on Māori education, reflecting the socio-economic and cultural perspectives. The best route to addressing these issues is probably through the development and evaluation of new policies, strategies and approaches based around the integration of the socio-economic and cultural perspectives.

7 Chapter 1: Introduction

This report is the first in a series of reports to Te Rūnanga o Ngāi Tahu using data from the Christchurch Health and Development study (CHDS). The CHDS is a longitudinal study of a birth cohort of 1,265 children born in Christchurch/Ōtautahi in 1977, who have been studied up to the age of 30. On the basis of parental reports gathered at birth, 142 children (11.2% of the cohort) were described as having some Māori descent.

This report traces the educational history of the Māori cohort and examines the following issues:

• Description of educational outcomes for both the Māori and non-Māori cohorts. • Influence of socio-demographic background at birth on educational attendance and attainment. • Influence of ethnic identification of the Māori cohort (at ages 21 and 25) on attendance and attainment.

The report describes Māori educational attendance/attainment at three developmental stages:

• Early Childhood Education attendance. • Primary and intermediate school achievement. • Secondary school and tertiary attendance and attainment of qualifications.

In general, the aims of this report are to build up a picture of the educational participation and attainment of a well-studied Māori cohort, from early childhood to tertiary education. In the remainder of this chapter we develop the background to this analysis.

1.1 Description of the Christchurch Health and Development Study

The CHDS is a longitudinal study of a birth cohort of 1,265 children (635 males; 630 females) born in the Christchurch/Ōtautahi, New Zealand urban region over a 4-month period during 1977. These children represented 97% of all children born in public and private maternity units within the Christchurch/Ōtautahi urban region. This cohort has been studied at birth, 4 months, 1 year, annually to age 16 years, and again at ages 18, 21, 25 and 30 years. Information has been obtained from a variety of sources including: parental interviews, teacher reports, self-report, psychometric assessments, medical and other record data. The data gathered over the course of the study now comprises some 50 million characters of information describing the life history of this cohort. This information has been used to address a wide range of issues relating to the social circumstances, health, development and well-being of the cohort (Fergusson & Horwood, 2001; Fergusson, Horwood, Shannon, & Lawton, 1989).

1.2 Defining the Māori cohort at birth

Over the course of the study, different definitions and measures of the Māori population have been used in New Zealand research and official statistics. These were generally based on questions from

8 the New Zealand Census of Population and Dwellings, which have changed to reflect evolving views of ethnicity over the years (Broughton, Fergusson, Rimene, Horwood, & Sporle, 2000; Cormack, 2007; Lang, 2002; Statistics New Zealand, 2000). Specifically:

• The Māori Affairs Amendment Act 1974 defined Māori as “any person of Māori descent”. This definition prevailed at the time the CHDS cohort was recruited. • The definition of ethnicity was changed in the 1986 Census in favour of a definition based on affiliation with, and self-identification of, an ethnic group. Furthermore, in 1991, the Census definition was expanded to again record possession of Māori ancestry.

As a consequence of these variations in the definition of ethnicity, the way in which Māori has been defined over the course of this study has also varied. At birth, the cohort members’ ethnicity was based on parent-reported ethnicity, and at ages 21 and 25, cohort members were asked about their Māori ethnic affiliation and degree of Māori cultural identity (Broughton, et al., 2000). This questioning made it possible to classify the CHDS cohort into three subgroups based on self-reported ethnic identity. These subgroups were:

• Sole-Māori: Those respondents identifying as being exclusively Māori. • Māori/other ethnic affiliation: Those respondents identifying as both Māori and as another ethnic group. • Non-Māori: Those respondents who reported no identification as Māori.

The availability of different measures of ethnicity (based on descent or identity obtained at different historical times of the cohort’s development: birth; 21 years; 25 years), poses difficulties in developing classifications of ethnicity that can be applied at all ages. In this report we have adopted the following strategy:

• The Māori cohort being studied was defined as all those classified as being of Māori descent at birth. This classification is used to describe the history of the cohort, but wherever applicable this assessment is supplemented by information on self-reported identity at age 21 and 25.

While this approach is not entirely satisfactory, it reflects the problems that have arisen in assessing Māori ethnicity over a historical period, in which official methods for assessing Māori ethnicity have shifted from descent to self-identification.

1.3 The Māori cohort description and sample retention

Data gathered at birth showed that 142 (11.2%) cohort members were described as being of Māori descent on the basis of parental reports of ethnicity. Table 1.1 shows the parentage of the 142 Māori cohort members. The table shows that 83.1% of the cohort had one non-Māori parent with 16.9% having two parents who reported Māori descent.

9 Table 1.1 Number and percentage of Māori cohort members based on parental ethnicity.

Parental ethnicity n %

Mother Māori/Father non-Māori 44 31.0

Father Māori/Mother non-Māori 74 52.1

Both parents Māori 24 16.9

Table 1.2 shows the numbers of Māori cohort members studied up to the age of 30. For the purposes of display, the numbers studied at birth and ages 5, 10, 15, 21, 25, and 30 years are reported. The table shows that at age thirty, 111 Māori cohort members were studied, with this sample representing 78.2% of the Māori cohort. For comparative purposes it should be noted that 78.0% of the non-Māori cohort were studied to age 30, indicating that rates of sample retention for Māori and non-Māori were virtually identical.

Table 1.2 Christchurch Health and Development Study cohort members identified as Māori at birth and interviewed at each assessment age, up to age 30.

Assessment age Interviewed (n) % Māori cohort % Surviving Māori cohort

Birth 142 100.0 100.0

5 years 128 90.1 90.8

10 years 115 81.0 81.6

15 years 111 78.2 78.7

21 years 113 79.6 80.7

25 years 113 79.6 80.7

30 years 111 78.2 79.3

Further analysis was conducted to examine the extent to which the 31 members of the Māori cohort lost to follow up differed to the 111 studied at age 30. These comparisons were made on eight measures collected at birth (maternal education, family socio-economic status at birth (Elley & Irving, 1976), family type, maternal age, gender, birth weight, planning of pregnancy, and whether the child was breast fed). These findings are reported in Table A1 in the Appendix to this report. Two of these comparisons (percentage of children in single parent families at birth and the pregnancy

10 being unplanned) showed a statistically significant difference (p<.05)1. Specifically, children born into single parent families and those where the pregnancy was described as unplanned were more likely to be lost to follow-up. However, given the number of statistical comparisons made, it is possible that these results were due to chance as a consequence of multiple statistical significance testing. Correction for this multiple testing showed that the associations between unplanned pregnancy, family situation at birth and losses to follow-up were not statistically significant. These findings suggest that losses to follow up were not systematic and that the cohort studied to 30 was broadly representative of the initial Māori cohort (n=142).

1.4 Ethnic identity at age 21 and 25

In addition to the measures of ethnicity collected at birth, all sample members were questioned at age 21 and 25 on their ethnic identity using the questions in the 1996 New Zealand Census of Population and Dwellings (Statistics New Zealand, 1997). These questions assessed individual cultural identity (as opposed to descent) and permit the classification of those reporting Māori identity into two groups: a) those reporting a sole Māori identity and b) those reporting a Māori/other ethnic affiliation. Table 1.3 shows the self-reported ethnic identification of the entire CHDS cohort studied at 21 and 25 years. The table shows that at each age, nearly 12% of the cohort reported some Māori identity. Of these, about 5% reported a sole-Māori identity and 6% to 8% reported a Māori/other ethnic affiliation.

It should be noted that combining over ages 21 and 25, a total of 169 cohort members reported being of Māori descent with 69 (40.8%) describing themselves as sole-Māori, 65 (38.5%) describing themselves as Māori/other ethnic affiliation and 35 (20.7%) describing themselves as non-Māori.

Table 1.3 Self-reported ethnic identification of the entire CHDS cohort studied at 21 and 25 years.

Māori/other Assessment age Sole-Māori ethnic affiliation Non-Māori Total

21 years n (%) 52 (5.1) 62 (6.1) 897 (88.7) 1011 (100.0)

25 years n (%) 45 (4.5) 77 (7.7) 881 (87.8) 1003 (100.0)

1.4.1 Associations between measures of Māori descent and identity

The availability of different definitions of ethnicity raises important issues about the inter- relationships of these measures. To address this, comparisons were made between measures of ethnicity based on descent at birth, self-identification at age 21 and self-identification at age 25. This analysis revealed that all measures were strongly correlated, suggesting a substantial agreement between measures of ethnicity based on descent and self-identity obtained at different ages. The

1 The p-values in this report may be interpreted as follows: p-values less than 0.05 (p<.05, or smaller) represent a statistically significant difference (i.e., between means) or a statistically significant association (i.e. regression model parameters). Conversely, p-values greater than 0.05 (p>.05, or greater) represent a statistically non-significant difference or association.

11 correlation between the measure of ethnicity based on descent and self-identified ethnicity at 21 was r=+0.79 (p<.001), the correlation between ethnicity based on descent and self-identity at 25 was r=+0.78 (p<.001) and the correlation between self-reported ethnicity at ages 21 and 25 was r=+0.87 (p<.001). All of these findings suggest substantial agreement and overlap between different measures of ethnicity obtained at different times and in different ways. At the same time this agreement was not perfect. As noted earlier, the strategy used in this report was to compare the conclusions drawn on the basis of different measures of ethnicity, wherever this was feasible. Appendix A2 to A4, show the associations between: a) ethnicity assessed at birth and self-reported Māori ethnic identity at age 21, b) ethnicity assessed at birth and self-reported Māori ethnic identity at age 25 and c) self-reported ethnicity at age 21 and self-reported ethnicity at age 25.

1.5 The socio-demographic background of the Māori cohort at birth

Figures 1.1 to 1.4 describe the socio-demographic background at birth of the 142 cohort members classified as Māori on the basis of descent and the non-Māori cohort. These comparisons included: maternal educational attainment, family socio-economic status at birth (based on the Elley and Irving (1976) scale), family type at birth (single parent/two parent), and maternal age. Inspection of the figures reveals two general trends:

• First, there was quite substantial variability within the Māori cohort: many of the measures examined indicated that the cohort was not a socially and economically homogenous population. These findings are consistent with Durie’s (1994) account of the diverse worlds of Māori and have important implications for later analysis since they suggest it is misleading to view Māori as being a uniformly disadvantaged population. • At the same time, inspection of the figures suggests that the Māori cohort was subject to relatively high levels of disadvantage. For example, over two-thirds of mothers lacked formal educational qualifications and 43% of families were classified as having low SES (Elley and Irving (1976) level 5/6). Rates of single parenthood were relatively high (20.4%) and parents were generally young; nearly 20% of mothers of Māori cohort members were aged 19 years or younger. All of these findings reflect the fact that in comparison to the non-Māori cohort, the Māori cohort was a relatively disadvantaged population. As shown in Figures 1.1 to 1.4, relative to non-Māori, the Māori cohort had significantly lower levels of maternal educational attainment (p<.001), lower mean family socio-economic status (p<.001), higher rates of single parenthood (p<.001) and younger maternal age (p<.001).

12 80 Māori 70 Non-Māori 60 Note: the 50 differences % 40 between Māori/non- 30 Māori 20 maternal 10 education are 0 No school Secondary school Terary statistically qualificaons qualificaon qualificaons significant (p<.001) Maternal educaon

Figure 1.1 Maternal educational attainment characteristics of the Māori cohort (n=142) and non- Māori cohort (n=1,123) defined at birth.

60 Māori

50 Non-Māori

40 Note: the differences 30 % between 20 Māori/non- Māori 10 Family socio- 0 economic Professional/ Clerical/ technical/ Semi-skilled/ managerial skilled trade unskilled/ status are unemployed statistically significant Family socio-economic status (p<.001)

Figure 1.2 Family socio-economic status of the Māori cohort (n=142) and non-Māori cohort (n=1,123) defined at birth.

13 100 Māori 90 Non-Māori 80 70 Note: the 60 differences between % 50 Māori/non- 40 Māori 30 family type 20 are 10 statistically 0 significant Single parent household Two parent household (p<.001) Family type

Figure 1.3 Family type of the Māori cohort (n=142) and non-Māori cohort (n=1,123) defined at birth.

70 Māori

60 non-Māori

50 Note: the differences 40 between Māori/non- 30 % Māori 20 Maternal age are 10 statistically significant 0 (p<.001) 19 years or younger 20 to 24 years 25 years or older Maternal age

Figure 1.4 Maternal age for the Māori cohort (n=142) and the non-Māori cohort (n=1,123) defined at birth.

14 The findings in Figures 1.1 to 1.4 were replicated using measures of cultural identity gathered at ages 21 and 25 (see Appendix A5 and A6). The same patterns were also found for these measures. In general, the results of this analysis clearly demonstrate that compared to the non-Māori cohort, the Māori cohort was disadvantaged on a number of measures, including: maternal educational attainment, family socio-economic status, family type and maternal age. These associations have clear implications for understanding the educational history of the Māori cohort.

1.6 Iwi identification

At ages 21 and 25, those reporting Māori descent were asked if they knew the name of their iwi and to specify this. At ages 21 and 25, a total of 169 respondents reported being of Māori descent. This number is larger than the number classified as Māori at birth, suggesting that some cohort members had discovered Māori whakapapa later in life. Of the 169 respondents, 125 (74.0%) were able to nominate iwi affiliation(s). The results of this questioning are summarised in Table 1.4, which shows the iwi affiliations of the respondents reporting Māori descent. The iwi described in the table are organised into a number of iwi regions using the classification from Statistics New Zealand (Statistics New Zealand, 2000). This classifies New Zealand into 11 iwi regions representing broad groupings of affiliated iwi. The table also shows the numbers and percentages of the sample (n=169) reporting each iwi. Note that the percentages in this table sum to more than 100% since some respondents reported more than one iwi affiliation. By age twenty-five, 97 (57.3%) reported one iwi affiliation, 24 (14.2%) reported two iwi affiliations and 4 (2.3%) reported three iwi affiliations.

15 Table 1.4 Number and percentage of cohort members of Māori descent (n=169) by self-reported iwi affiliation by region.1

Iwi region/affiliation n % Iwi region/affiliation n % Te Waipounamu Ngāi Tahu/Kāi Tahu 71 42.0 Ngāti Ruanui 2 1.1 Ngāti Māmoe/Kāti Māmoe 2 1.1 Ngā Rauru 1 0.6 Total 73 43.2 Ngāti Kahu 1 0.6 Te Tau Ihu Ngāti Mutunga 2 1.1 1 0.6 Te Āti Awa 4 2.4 Ngāti Toa 2 1.1 Total 10 5.9 Total 3 1.8 Takitimu Ngāti Paoa 1 0.6 7 4.1 Ngāti Maniapoto 4 2.4 Total 7 4.1 6 3.6 Rangitikei Ngāti Hauā 3 1.8 Ngāti Hauiti 1 0.6 Ngāti Raukawa 2 1.1 Total 1 0.6 Total 15 8.9 Mataatua Te Ati Hau Nui-A-Paparangi 1 0.6 Ngāti Awa 2 1.1 Total 1 0.6 Ngāi Tūhoe 1 0.6 Arawa Te Whakatōhea 1 0.6 Te Arawa 11 6.5 Total 4 2.3 Ngāti Tūwharetoa 7 4.1 Tarawhiti Ngāti Whakaue 1 0.6 Ngāti Porou 8 4.7 2 1.1 Ngāi Tāmanuhiri 1 0.6 Tūhourangi 1 0.6 Total 9 5.3 Ngāti Rangiwewehi 1 0.6 Tai Tokerau Total 22 13.0 Te Aupouri 1 0.6 Ngāpuhi 20 11.8 Ngāti Whātua 2 1.1 1 0.6 Total 24 14.2 No affiliation stated 44 26.0 1 Only regions and iwi affiliations that were reported by the cohort have been listed in this table. Percentages sum to more than 100% as some respondents nominated more than one iwi affiliation.

16 The table leads to the following general conclusions:

• As might be expected, the largest iwi grouping was from Te Waipounamu, with 42% of the sample reporting an affiliation with Ngāi Tahu/Kāi Tahu. • Sample members also reported affiliations with a broad range of North Island Iwi; the most frequent of these being Ngāpuhi (11.8%) and Te Arawa (6.5%).

These findings highlight the complex mix of iwi within the cohort and also draw attention to the fact that 26.0% of respondents were unable to nominate an iwi affiliation despite being questioned on this matter on two occasions.

Summary and conclusions

In this chapter we have provided a summary and overview of the ethnic and related background of the Māori cohort(s) participating in the CHDS. The following conclusions may be drawn from this analysis:

• Ethnicity at birth: on the basis of parentally reported descent, 11.2% of the cohort was defined as being of Māori descent. • The CHDS cohort members were studied to the age of 30, at which time 111 Māori cohort members were studied, with this group representing 78.2% of the initial Māori cohort. • Rates of sample retention in the CHDS to the age of 30 did not vary with ethnicity. • Rates of sample loss for the Māori cohort did not vary with a series of measures assessed at birth. These measures included: maternal education, family socio-economic status at birth (Elley & Irving, 1976), family type, maternal age, gender, birth weight, planning of pregnancy, and whether the child was breast fed. • In addition to the measures of ethnicity collected at birth, all sample members were questioned at age 21 and 25 on their ethnic identity using questions from the 1996 New Zealand Census of Population and Dwellings. At each age, nearly 12% of the cohort reported some Māori identity. Of these, about 5% reported a sole-Māori identity and 6% to 8% reported a Māori/other ethnic affiliation. • Measures of Māori ethnicity based on parental report of descent at birth, and reports of self-identity at 21 and 25, were generally strongly correlated (r = 0.78-0.87; p<.05). These findings suggest that it is valuable to compare results using differing definitions of ethnicity. • Examination of the socio-demographic features of the Māori cohort at birth showed that in comparison to the non-Māori cohort, this group was subject to relative disadvantage in the areas of: maternal education (p<.001), family socio-economic status at birth (p<.001), family type (single parent or two parent family) (p<.001), and younger maternal age (p<.001). At the same time, there was evidence of quite considerable variability in the Māori cohort, suggesting that it would be misleading to view this cohort as a homogenously disadvantaged group. These conclusions are consistent with the observations of Durie (1994) on the diverse worlds of Māori and highlight the importance of examining variability within Māori populations.

17 • Of those reporting Māori descent at age 21 and 25 (n=169), 74.0% were able to nominate at least one iwi affiliation. The predominant affiliation nominated was Ngāi Tahu/Kāi Tahu (42.0%). It is notable that of the nominated iwi affiliations, the majority (56.9%) were from North Island iwi, with Ngāpuhi (11.8%) and Te Arawa (6.5%) being the most frequently mentioned. • At ages 21 and 25, of the 169 cohort members who reported being of Māori descent: 40.8% described themselves as sole-Māori, 38.5% described themselves as Māori/other ethnic affiliation and 20.7% as non-Māori.

Against the background of this introduction, the remainder of this report focusses on the educational history of the Māori cohort(s), their participation in various forms of educational activity and the factors associated with academic success and achievement.

18 Chapter 2: The early childhood educational experiences of the Māori cohort

In this chapter, we examine the early childhood educational experiences of the Māori cohort. The key questions addressed are:

• What percentage of the Māori cohort attended Early Childhood Education (ECE)? • What types of ECE facilities were chosen? • Were there differences in ECE attendance by the Māori and non-Māori cohort members? • To what extent were any ethnic differences in ECE attendance explained by differences in the socio-demographic characteristics (maternal educational attainment, family socio- economic status, family type, and maternal age) of the Māori and non-Māori cohorts?

2.1 What percentage of the Māori cohort attended Early Childhood Education (ECE)?

Of the initial birth cohort of 142 children, a total of 123 (86.6%) were studied from age 3 to 5 years, during which time they were available to enter various forms of ECE. At the time of the study, the major sources of ECE were: Playcentre, kindergarten, and daycare centres.

Figures 2.1 and 2.2 summarise the ECE experiences of the Māori cohort. Figure 2.1 shows the percentages of the cohort attending ECE at ages 2-3, 3-4, and 4-5, and those who had ever attended ECE. The figure shows that by age five, 89.4% had attended ECE, with 32.5% starting their ECE experience at 2-3 years. The results in Figure 2.1 imply that the duration of ECE attendance by the Māori cohort varied quite substantially. This variation is summarised in Figure 2.2, which shows the percentage of the Māori cohort attending ECE by duration of attendance. On average, the Māori cohort attended ECE for 1.9 years.

19 100 90 80 70 60 % 50 40 30 20 10 0 2-3 years 3-4 years 4-5 years Ever Age

Figure 2.1 Percentage of the Māori cohort (n=123) who attended Early Childhood Education (ECE) from age 2 to age 5 years, and who had ever attended ECE.

60

50

40

% 30

20

10

0 0 <1 year 1-2 years 2-3 years Number of years aended ECE

Figure 2.2 Percentage of the Māori cohort (n=123) attending Early Childhood Education (ECE) by duration of attendance (in years).

20 2.2 Sources of Early Childhood Education (ECE)

Figure 2.3 shows the facilities attended by the 110 Māori cohort members who had ever attended ECE. Kindergarten was the most commonly reported educational facility, attended by nearly three- quarters of those attending ECE (72.7%). Slightly more than one-quarter reported attending Playcentre (27.2%), with the remainder reporting attendance at daycare centres (10.9%) or other educational facilities (22.7%).

Details of the number of children attending Kōhanga Reo were not obtained because at the time the CHDS cohort was eligible for ECE, the Kōhanga Reo movement was relatively new and unknown. It is likely that some of those describing attendance at other ECE facilities would have been attending Kōhanga Reo, but the actual fraction cannot be ascertained.

80

70

60

50 % 40 30

20

10

0 Playcentre Kindergarten Daycare centre Other Type of facility

Figure 2.3 Percentage of the Māori cohort (n=123) who had ever attended Early Childhood Education (ECE) (n=110) by type of facility chosen.

2.3 Early Childhood Education (ECE) attendance among the Māori and non-Māori cohorts

Figure 2.4 compares rates of ECE participation amongst the Māori cohort (n=123) and the non-Māori cohort (n=975) studied to the age of 5. The figure shows that at all ages, rates of participation in ECE were higher amongst the non-Māori cohort than the Māori cohort, with these differences being statistically significant at ages 3-4 (p=0.005) and 4-5 (p<.001); overall, 89.4% of the Māori cohort had attended ECE sometime before age 5, compared to 95.3% of the non-Māori cohort (p=0.007).

21

Figure 2.5 compares the mean duration of attendance at ECE for the Māori and non-Māori cohorts. The figure shows that Māori attended ECE for a mean of 1.91 years (SD=0.90) compared with a mean of 2.13 years (SD=0.80) for non-Māori. While the differences between Māori and non-Māori ECE participation were comparatively small (mean difference=0.22 years), they were nonetheless statistically significant (p=0.009).

100 * * Māori 90 * Non-Māori 80 70 Note: * denotes a 60 statistically % 50 significant 40 Māori/non- 30 Māori 20 difference 10 (p<.01) 0 2-3 years 3-4 years 4-5 years Ever Age at aendance

Figure 2.4 Percentage of the Māori cohort (n=123) and the non-Māori cohort (n=975) attending Early Childhood Education (ECE) at each age, from 2 to 5 years, and who had ever attended ECE.

2.5 Māori Non-Māori 2

Note: Mean years 1.5 Māori/non- aended Māori ECE 1 differences are statistically 0.5 significant (p<.05) 0 Māori Non-Māori Ethnicity

Figure 2.5 Mean number of years of Early Childhood Education (ECE) attendance by the Māori cohort (n=123) and the non-Māori cohort (n=975).

22 2.4 Differences in Early Childhood Education (ECE) attendance between the Māori and non-Māori cohorts after adjustment for socio-demographic factors at birth

One explanation for the ethnic differences in Figures 2.4 and 2.5 is that these differences can be explained by variations in the socio-demographic backgrounds of the Māori and the non-Māori cohorts. As shown in Chapter 1 (see Section 1.5), the Māori and non-Māori cohorts differed on a number socio-demographic factors at birth, including: maternal educational attainment, family socio-economic status (based on the Elley and Irving (1976) scale), family type (single parent/two parent) and maternal age. These comparisons suggested that the Māori cohort was relatively disadvantaged when compared to the non-Māori cohort. This conclusion implies that the differences in ECE attendance by the Māori and non-Māori cohorts may be explained by socio-demographic factors correlated with ethnicity. To examine this hypothesis, the associations between ethnicity and ECE attendance (shown in Figures 2.4 and 2.5) were adjusted for the socio-demographic factors listed above using regression methods. The regression analyses are summarised in Appendix B Tables B1-B5. The results of these analyses are discussed below.

Figure 2.6 shows the percentage of Māori and non-Māori attending ECE at various ages, following statistical adjustment for socio-demographic factors (maternal educational attainment, family socio- economic status, family type, and maternal age). The adjusted percentages may be interpreted as the hypothetical rates of ECE attendance for Māori and non-Māori, had both populations experienced the same socio-demographic circumstances. The figure shows that overall there was no statistically significant difference between the adjusted ECE attendance of Māori and non-Māori (adjusted rate Māori=92.5%; adjusted rate non-Māori=95.7%; p>.05). The only exception to this trend was that at ages 4-5, Māori/non-Māori differences remained statistically significant (adjusted Māori rate=89.5%; adjusted non-Māori rate=94.4%; p<.05). These findings clearly demonstrate that most of the differences in Māori/non-Māori ECE attendance rates were explained by the socio- demographic factors correlated with ethnicity.

23 100 Māori 90 * Non-Māori 80 70 60 % 50 Note: * 40 denotes statistically 30 significant 20 Māori/non- 10 Māori 0 difference 2-3 years 3-4 years 4-5 years Ever (p<.05) Age at aendance

Figure 2.6 Adjusted percentages of Early Childhood Education (ECE) attendance for each age group from 2 to 5 years, and those who had ever attended ECE by the Māori cohort (n=123) and the non- Māori cohort (n=975).

Figure 2.7 shows the mean number of years of ECE attendance by the Māori and non-Māori cohorts, after statistical adjustment for socio-demographic factors. The figure shows that there was no significant or substantive difference in the mean number of years of ECE attendance between Māori and non-Māori after adjustment. The adjusted mean for Māori was 1.99 years compared to the adjusted mean of 2.14 years for non-Māori (p>.05).

24 2.5 Māori Non-Māori 2 Note: the Mean years 1.5 differences aended between Māori and 1 ECE non-Māori are 0.5 statistically non- 0 significant Māori Non-Māori (p>.05) Ethnicity

Figure 2.7 Adjusted means for the number of years of Early Childhood Education (ECE) attendance by the Māori (n=123) and non-Māori cohorts (n=975).

The findings summarised in Figures 2.6 and 2.7 show that the differences between the rates of Māori and non-Māori ECE attendance were explained by socio-demographic factors associated with Māori ethnicity. The regression model parameters reported in Appendix B Tables B1 to B5, show that key covariate factors were: maternal educational attainment, family socio-economic status and family type as assessed at the birth of the cohort member. In general, these findings imply that the lower rates of ECE attendance were a reflection of the disadvantaged status of the Māori cohort, rather than being a consequence of ethnic factors.

2.5 Summary and conclusions

In this chapter we have examined the first stage of the educational journey of the Māori cohort as it moved from birth to adulthood. The central findings of this analysis may be summarised as follows:

• Overall, there was a high rate of participation of Māori in ECE; nearly 90% of the Māori cohort attended ECE by age five. • ECE participation of the Māori cohort clearly reflected the availability of ECE services in the 1980’s, with the majority of cohort members attending kindergarten, Playcentre and daycare centres. Attendance at the then emerging Kōhanga Reo services was not recorded in this study, but it is clear that less than one-quarter of the Māori cohort would have used these services. • In comparison to non-Māori, the overall Māori ECE attendance rate was significantly lower; by age five, 89.4% of Māori had attended ECE, compared to 95.3% of non-Māori (p<.01). On average, Māori received 1.91 years of preschool education, compared to 2.13 years for non- Māori (p=0.007).

25

• Further analyses suggested that these ethnic differences in rates of ECE participation were explained by socio-demographic differences at birth (maternal educational attainment, family socio-economic status, family type and maternal age). After statistical adjustment for these factors there were no detectable differences in the rates of ECE participation by Māori and non-Māori.

In summary, while there was some evidence of ethnic differences in ECE attendance, these differences were small and the great majority of Māori children attended some form of ECE before the age of 5. The most commonly used forms of education were kindergarten, Playcentre and daycare centres. Differences in ECE participation between the Māori and non-Māori cohorts were explained by differences in the socio-demographic backgrounds of the two groups. Once social background had been taken into account, rates of participation in ECE by Māori were virtually identical to rates of participation by non-Māori. These findings clearly suggest that any differences in rates of ECE participation for Māori and non-Māori were due to socio-demographic factors rather than to cultural or ethnic differences.

26 Chapter 3: School educational achievement (7-13 years)

In this chapter, we examine the educational history of the Māori cohort at primary and intermediate schools (when aged 7 to 13 years). The key questions addressed are:

• To what extent were there differences between the educational achievement of Māori and non-Māori when assessed by: a) teacher ratings of school performance; and b) standardised tests of reading and mathematics? • Can ethnic differences in educational achievement be explained by socio-demographic factors (maternal educational attainment, family socio-economic status, family type, and maternal age) associated with ethnicity? • To what extent did educational attainment vary with the cultural identity of the cohort member?

3.1 Educational achievement of the Māori and non-Māori cohorts at school

At ages 7 to 13 years, educational achievement of the CHDS cohort was assessed by collecting data on teacher ratings of school performance and child achievement on standardised tests of reading and mathematics. These measures and the results are described below:

3.1.1 Teacher ratings of school performance

As part of the CHDS, data were gathered on teacher ratings of school performance in a number of key areas, including: reading, mathematics, spelling and handwriting. For each subject area, teachers rated performance on a 5-point scale ranging from 1 (very poor) to 5 (very good). To represent teacher ratings, an overall Grade-Point Average (GPA) score was constructed by summing the ratings and dividing by four (the number of academic areas assessed).

Figure 3.1 compares the GPA scores of the Māori and non-Māori cohorts at ages 7, 10 and 13. The figure shows the presence of small but statistically significant (p<.05) differences between the two groups at all ages. Māori had mean GPA scores that were between .20 and .40 standard deviations lower than non-Māori.

27

4 3.5 Māori 3 2.5 Non-Māori Mean 2 Note: all Māori, GPA score 1.5 non-Māori 1 differences are 0.5 statistically 0 significant 7 10 13 p<.05 Age

Figure 3.1 Mean teacher-rated Grade-Point Average (GPA) scores (ages 7, 10 and 13) for the Māori and non-Māori cohorts.

3.1.2 Standardised tests

Parallel to the GPA information provided by the teachers, cohort members resident in Canterbury/Waitaha were also assessed on a number of standardised tests of reading and mathematics. The cohort members were assessed on the Burt Word Reading Test (Gilmore, Croft, & Reid, 1981) at the ages of 8, 11, and 13; the Progressive Achievement Test (PAT) in mathematics (Reid & Hughes, 1974) at age 11 and the Progressive Achievement Test (PAT) in reading comprehension (Elley & Reid, 1969) at age 12. All scores were scaled to a mean of 100 and a standard deviation of 10.

Figure 3.2 compares the Māori and non-Māori cohorts on the Burt Word Reading Test assessed at ages 8, 11 and 13. The comparisons in this figure are based on those cohort members resident in Canterbury/Waitaha (8 years: Māori=89, non-Māori=792; 11 years: Māori=78; non-Māori =753; 13 years: Māori= 72, non-Māori=711). This figure shows significant differences between the Māori and non-Māori cohorts on measures of word recognition at ages 8 and 11 (p<.05), with these differences declining to statistical non-significance at age 13 (p>.05).

28 102 * * 100 Māori

98 non-Māori Mean Burt 96 Note: * denotes score a statistically 94 significant 92 Māori, non-

90 Māori 8 11 13 difference p<.05 Age

Figure 3.2 Mean for the standardised test scores of the Burt Word Reading Test (ages 8, 11, 13) for the Māori and non-Māori cohorts.

Figure 3.3 compares the Māori and non-Māori cohorts on the Progressive Achievement Test (PAT) in mathematics assessed at age 11; and the Progressive Achievement Test (PAT) in reading comprehension assessed at age 12. The comparisons in Figure 3.3 are based on those cohort members resident in Canterbury/Waitaha at ages 11 and 12 years (11 years: Māori=78; non- Māori=753; 12 years: Māori=73; non-Māori=731). This figure shows significant differences (p<.05) between Māori and non-Māori achievement at ages 11 (mathematics) and 12 (reading comprehension).

102

100 Māori Non-Māori 98 Mean PAT Note: all Māori, 96 score non-Māori 94 differences are statistically 92 significant

90 p<.05 11 (Mathemacs) 12 (Reading) Age

Figure 3.3 Mean for the standardised test scores of the Progressive Achievement Test (PAT) in Mathematics (age 11) and Reading (age 12) for the Māori and non-Māori cohorts.

29

Although Figures 3.2 and 3.3 show differences between the Māori and non-Māori cohorts on standardised measures of reading and mathematics, these differences are relatively small, with Māori having standardised test scores that were between .30 and .40 standard deviations lower than non-Māori.

3.2 Differences in school achievement between the Māori and non-Māori cohorts after adjustment for socio-demographic factors at birth

An explanation for the ethnic differences in Figures 3.1 to 3.3 is that these differences can be explained by variations in the socio-demographic backgrounds of the Māori and non-Māori cohorts. As shown in Chapter 1 (see Section 1.5), the Māori and non-Māori cohorts differed on a number socio-demographic factors at birth including: maternal educational attainment, family socio- economic status (based on the Elley and Irving (1976) scale), family type (single parent/two parent households) and maternal age. Using the same statistical techniques as Chapter 2, Section 2.5, the associations between ethnicity and school achievement (shown in Figures 3.1-3.3) were adjusted for the socio-demographic factors listed above using regression methods (see Appendix C, Tables C.1- C.8). The results of these analyses are discussed below.

3.2.1 Adjusted mean Grade-Point Average (GPA) scores of the Māori and non-Māori cohorts

Figure 3.4 shows the association between ethnicity and mean teacher-rated GPA scores (ages 7, 10, 13) after adjustment for socio-demographic factors. This figure shows that following adjustment there were no statistically significant differences between Māori and non-Māori on GPA. Statistically significant covariate factors were: maternal educational attainment and family socio-economic status (see Appendix C, Tables C.1 to C.3).

3.5 3 Māori 2.5 Non-Māori Mean GPA 2 score 1.5 Note: all Māori, 1 non-Māori differences are 0.5 statistically 0 non-significant 7 10 13 p>.05 Age

Figure 3.4 Adjusted mean teacher-rated Grade-Point Average (GPA) scores (ages 7, 10 and 13) for the Māori and non-Māori cohorts.

30 3.2.2 Adjusted mean Burt Word Reading Test scores of the Māori and non-Māori cohorts

Figure 3.5 shows associations between ethnicity and mean Burt Word Reading Test scores (ages 8, 11, 13) after adjustment for socio-demographic factors. The figure shows that following adjustment, no statistically significant differences remained between Māori and non-Māori on Burt Word Reading Test scores. Statistically significant covariate factors included: maternal educational attainment and family socio-economic status (based on the Elley and Irving (1976) scale) (see Appendix C, Tables C.4 to C.6).

102 Māori 100 non-Māori 98 Mean BURT Note: all Māori, 96 non-Māori score differences are 94 statistically 92 non-significant p>.05 90 8 11 13 Age

Figure 3.5 Adjusted mean standardised test scores of the Burt Word Reading Test (ages 8, 11, 13) for the Māori and non-Māori cohorts.

3.2.3 Adjusted mean Progressive Achievement Test (PAT) scores of the Māori and non- Māori cohorts

Figure 3.6 shows the associations between ethnicity and mean Progressive Achievement Test (PAT) scores (age 11, mathematics; age 12, reading) following adjustment for socio-demographic factors. The figure shows that after adjustment no statistically significant differences remained between Māori and non-Māori on Progressive Achievement Test (PAT) scores. Statistically significant covariate factors were: maternal age, maternal educational attainment and family socio-economic status (based on the Elley and Irving (1976) scale) (see Appendix C, Tables C.7 and C.8).

31

102 Māori 100

98 Non-Māori Mean PAT 96 Note: all Māori, score 94 non-Māori differences are 92 statistically 90 non-significant 11 (Mathemacs) 12 (Reading) p>.05 Age

Figure 3.6 Adjusted mean standardised test scores of the Progressive Achievement Test (PAT) in Mathematics (age 11) and Reading (age 12) for the Māori and non-Māori cohorts.

The major conclusion from these analyses is that the apparent associations between ethnicity and measures of school achievement from age 7 to 13 were explained by socio-demographic factors correlated with ethnicity. Key covariate factors included maternal age, maternal education and family socio-economic status (based on the Elley and Irving (1976) scale).

3.3 Māori cultural identity and school achievement

The analysis above was extended to examine the extent to which variations in Māori cultural identity were associated with differences in school achievement. As explained previously (see Chapter 1, Section 1.2), a measure of cultural identity (classified as Sole Māori; Māori/other ethnic affiliation; non-Māori) at ages 21 and 25 was used.

3.3.1 Māori cultural identity and teacher-rated Grade-Point Average (GPA) scores

Figure 3.7 shows the associations between cultural identity amongst the Māori cohort (Sole Māori; Māori/other ethnic affiliation; non-Māori) and mean teacher-rated GPA. In this analysis, no statistically significant associations were found between mean teacher-rated GPA at ages 7, 10 and 13 and cultural identity at ages 21 and 25. The figure shows that variations in cultural identity were not related to teacher-rated school performance.

32 3.5 3 Sole Māori 2.5 Māori/other Mean GPA 2 Non-Māori score 1.5 1 0.5 0 7 10 13 Age

Figure 3.7 Mean teacher-rated Grade-Point Average (GPA) scores (ages 7, 10, and 13) by cultural identity at ages 21 and 25.

3.3.2 Māori cultural identity and Burt Word Reading Test scores

Figure 3.8 shows the associations between cultural identity amongst the Māori cohort and scores on the Burt Word Reading Test at ages 8, 11 and 13. In this analysis, no statistically significant associations were found between cultural identification and word recognition scores.

100 Sole Māori 95 Māori/other Note: All of the Mean Burt Non-Māori cultural score 90 differences are statistically non- 85 significant p>.05

80 8 11 13 Age

Figure 3.8 Mean Burt Word Reading Test scores (ages 8, 11, and 13) by cultural identity at ages 21 and 25.

33 3.3.3 Māori cultural identity and Progressive Achievement Test (PAT) scores

Figure 3.9 shows the associations between cultural identity and Progressive Achievement Test (PAT) scores of mathematics (11 years) and reading comprehension (12 years). In line with the findings in the previous figures, there were no statistically significant associations between cultural identification and Progressive Achievement Test (PAT) scores.

100 Sole Māori

95 Māori/other Mean PAT Non-Māori score 90

85 Note: All of the 80 cultural 11 (Mathemacs) 12 (Reading) differences are statistically non- Age significant p>.05

Figure 3.9 Mean Progressive Achievement Test (PAT) scores (ages 11: Mathematics, and 12: Reading) by cultural identity at ages 21 and 25.

3.4 Summary and conclusions

In this chapter, we have examined the educational achievement of the Māori cohort at primary and intermediate school (ages 7 to 13). The main findings of these analyses may be summarised as follows:

• There were consistent differences between the academic performance of the Māori and non-Māori cohorts. Māori had lower teacher-rated GPA scores and performed less well on standardised tests of word recognition, reading comprehension and mathematics. These findings suggest that irrespective of the way in which achievement was measured (teacher ratings; standardised tests) there was consistent evidence of Māori/non-Māori differences in educational achievement. However, these differences were relatively small; typically the differences between Māori and non-Māori were between .2 and .4 standard deviations. • Further analysis revealed that any differences between Māori and non-Māori achievement were accounted for by socio-demographic factors. Specifically, these factors included: maternal educational attainment, family socio-economic status (based on the Elley and Irving (1976) scale), family type (single parent/two parent households) and maternal age. When due allowance was made for these factors, no statistically significant associations remained between Māori/non-Māori and educational achievement.

34 • Finally, the results showed that educational achievement between ages 7 and 13 years was not associated with cultural identity among the Māori cohort (categorised as Māori; Māori/other ethnic affiliation; or non-Māori) assessed at ages 21 and 25.

The findings in this chapter suggest that for the CHDS cohort, the origins of educational underachievement in Māori were socio-demographic rather than cultural. These findings have potentially important implications for contemporary debates about the origins of educational underachievement among Māori. In particular, many contemporary accounts of Māori educational disadvantage have focussed on the role of cultural factors including: language, school culture and related factors as determinants of Māori educational achievement (Bishop & Glynn, 1999; Fitzsimons & Smith, 2000). However, the findings of this study suggest that the ethnic differences in educational achievement were due to socio-demographic rather than cultural factors. Two lines of evidence support this conclusion: first, statistical control for socio-demographic factors adequately explained the association between ethnicity and educational achievement; and second, there was no detectable relationship between cultural identity and educational achievement. The latter conclusion is limited by the fact that cultural identity was measured at ages 21 and 25, rather than during the primary and intermediate school years. It is possible that had cultural identity been measured at this time, associations between cultural identity and educational achievement may have been found.

In the concluding chapter of this report we will consider the implication of these findings for the development of policies targeted at eliminating ethnic differences in educational achievement.

35 Chapter 4: Secondary school and tertiary educational attainment and participation (15-30 years)

In this chapter, we examine the secondary and tertiary educational history of the Māori cohort when aged from 15 to 30. The following questions are addressed:

• To what extent are there differences between the Māori and non-Māori cohorts regarding: a) attainment of either secondary school or tertiary qualifications; and b) participation at tertiary educational institutions? • Can ethnic differences in educational attainment and participation be explained by socio- demographic factors (maternal education, family socio-economic status, family type, and maternal age) associated with ethnicity? • To what extent did educational attainment and participation vary with the cultural identity of the cohort member?

4.1 Ethnicity and educational attainment and participation of the Māori and non- Māori cohorts

At ages 18, 21, 25 and 30, CHDS cohort members were questioned about their attainment of educational qualifications and their history of enrolment at educational institutions (since the previous interview). From this information, the following measures were developed; these measures and results are described below:

4.1.1 Attainment of secondary school and tertiary educational qualifications

Attainment of secondary school and tertiary educational qualifications was assessed using three measures in which the respondent reported whether or not they had:

Attained any secondary school qualification (prior to age 21). At the time the cohort members received their secondary education, the following educational qualifications were available:

• School Certificate • Sixth Form Certificate • Higher School Certificate • University Entrance Bursary/Scholarship

A more detailed account of these qualifications is available from the Ministry of Education website (Ministry of Education, 2013). At ages 18 and 21, participants were questioned about any secondary school qualifications they had attained. For the purposes of the current analyses, study participants who attained any of the above qualifications by age 21 were defined as leaving school with qualifications. This information was used to construct a dichotomous measure of those who had either: attained a recognised secondary school qualification; or those who had not.

36 Attained any tertiary qualification (by age 30). At ages 18, 21, 25 and 30, cohort members were interviewed about their attainment of any tertiary-level educational qualifications. Tertiary qualifications included: a recognised university degree, certificate or diploma; a polytechnic or other tertiary training establishment’s certificate or diploma; a trade certificate; or another qualification that required at least 3 months full-time study. This information was used to construct a dichotomous measure that classified the cohort into either: those who had attained a qualification from a tertiary educational training institution, polytechnic, or university; or those who had not.

Attained a Bachelor’s degree or higher (by age 30). At age 30, cohort members were questioned as to whether they had ever attained a Bachelor’s degree or higher from a university or another accredited tertiary educational institution (Ministry of Education, 2008; New Zealand Qualifications Authority, 2010). This information was used to construct a dichotomous measure that classified the cohort into either: those who had attained a Bachelor’s degree or higher; or those who had not.

Figure 4.1 compares the Māori and non-Māori cohorts on measures of secondary school and tertiary educational attainment between ages 15 and 30. The comparison between ethnicity and attainment of any secondary school qualifications shown in Figure 4.1 is based on 119 Māori and 934 non-Māori Christchurch Health and Development Study (CHDS) cohort members. The comparisons between ethnicity and the remaining tertiary outcomes are based on 111 Māori and 876 non-Māori CHDS cohort members. The comparisons in Figure 4.1 show that Māori had:

• Lower rates of secondary school qualifications: Māori attainment was 15% lower than non- Māori attainment (Māori=68.9%; non-Māori=83.1%, p<.001). • Lower rates of attainment of any tertiary educational qualifications: Māori attainment was 15% lower than that of non-Māori (Māori=54.9%; non-Māori=70.7%, p<.001). • Lower rates of attainment of a Bachelor’s degree or higher: Māori degree attainment was nearly 20% lower than that of non-Māori (Māori=12.6%; non-Māori=32.0%, p<.001).

90 80 Māori 70 Non-Māori 60 50 % 40 Note: all Māori. 30 Non-Māori differences are 20 statistically 10 significant p<.05 0 Any secondary Any terary Bachelor's degree or qualificaon qualificaon higher Level of educaonal aainment

Figure 4.1 Secondary and tertiary educational attainment for the Māori and non-Māori cohorts.

37 However, a limitation of this analysis is that those cohort members who had attended a tertiary educational institution, but had not gained any recognised qualifications, were excluded. Therefore, to supplement this information we have conducted an additional analysis which examines participation in tertiary education.

4.1.2 Participation in tertiary education

Participation in tertiary education was assessed using two measures in which the respondent reported whether or not they had:

Attended any type of tertiary educational institution (by age 30). At ages 18, 21, 25 and 30, cohort members were questioned about their participation in tertiary education since the previous interview. Tertiary education was defined as full-time or part-time tertiary educational participation, regardless of whether a qualification was achieved, at any of the following institutions: a university, polytechnic or other tertiary educational training institution (other tertiary training institutions included: private training establishments, industry training organisations and wānanga) (Ministry of Education, 2008; New Zealand Qualifications Authority, 2010; Tertiary Education Commission, 2009). This information was used to construct a dichotomous measure that classified the cohort into either: those who had attended any tertiary educational training institution; or those who had not.

Attended a recognised university (by age 30). At ages 18, 21, 25 and 30 cohort members were asked if they had ever attended or were attending a recognised university since the previous interview, regardless of whether a qualification was achieved. A recognised university is a state-owned entity that is accredited by the Committee on University Academic Programmes (CUAP) of Universities New Zealand (Ministry of Education, 2008; Tertiary Education Commission, 2009; Universities New Zealand, 2013). This information was used to construct a dichotomous measure that classified the cohort into either: those who had attended a recognised university; or those who had not.

Figure 4.2 compares the Māori and non-Māori cohorts on measures of tertiary educational participation up to age 30. The comparisons between ethnicity and participation shown in Figure 4.2 were based on 111 Māori and 876 non-Māori CHDS cohort members. The comparisons in Figure 4.2 show that Māori had:

• Lower rates of attending any type of tertiary educational institution: Māori participation was 10% lower than non-Māori (Māori=81.9%; non-Māori=92.2%, p<.001). • Lower rates of attending a recognised university: Māori participation was 16% lower than non-Māori (Māori=41.4%; non-Māori=57.9%, p<.001).

38

100 90 Māori 80 Non-Māori 70 60 % 50 Note: all Māori, 40 non-Māori 30 differences are statistically 20 significant p<.05 10 0 Any terary educaon Any university educaon Educaonal parcipaon

Figure 4.2 Tertiary educational participation among the Māori and non-Māori cohorts.

4.2 Differences in tertiary educational attainment and attendance between the Māori and non-Māori cohorts after adjustment for socio-demographic factors at birth

To take account of the known socio-demographic differences between the Māori and non-Māori cohorts, the associations in Figures 4.1 (educational attainment) and 4.2 (educational attendance) were statistically adjusted for socio-demographic factors (maternal educational attainment, family socio-economic status, family type and maternal age) using regression methods. The regression analyses are reported in Appendix D. The results of these analyses are discussed below.

39

4.2.1 Educational attainment for the Māori and non-Māori cohorts (15-30) years

Figure 4.3 compares the rates of educational attainment for the Māori and non-Māori cohorts after statistical adjustment for socio-demographic factors assessed at birth. The figure shows that in all cases the difference between Māori and non-Māori were not statistically significant (p<.05) after adjustment for socio-demographic factors. Significant covariates include: maternal educational attainment, family socio-economic status and maternal age (see Appendix Tables D.1-D.3).

90 80 Māori 70 Non-Māori 60 Note: all Māori, 50 % non-Māori 40 differences that 30 was statistically 20 non-significant p>.05 10 0 Any secondary Any terary Bachelor's degree or qualificaon qualificaon higher Type of educaonal aainment

Figure 4.3 Educational attainment of the Māori and non-Māori cohort following statistical adjustment for socio-demographic factors.

4.2.2 Educational attendance for the Māori and non-Māori cohorts (15-30 years)

Figure 4.4 compares the adjusted rates of educational attendance among Māori and non-Māori. The figure shows that there was no significant association between ethnicity and university attendance after adjustment for socio-demographic factors (p>.20). However, there was a small but significant tendency for Māori to attend any tertiary educational institution following statistical adjustment. The adjusted rate for Māori was 85.5%, compared with the adjusted rate for non-Māori of 91.9% (p<.05) (see Appendix D, Tables D.4 and D.5).

40

100 * 90 Māori 80 Non-Māori 70

60 Note: * denotes % 50 statistically 40 significant Māori, 30 non-Māori difference p<.05 20 10 0 Any terary educaon Any university educaon Type of educaonal parcipaon

Figure 4.4 Educational participation of the Māori and non-Māori cohort following statistical adjustment for socio-demographic factors.

4.2.3 Conclusions

In general, the findings in Figures 4.3 and 4.4 suggest that when due allowance was made for socio- demographic factors, most of the associations between achievement/attendance and ethnicity became statistically non-significant (p>.05). The exception to this was for attendance at any tertiary educational institution, where there was a small tendency for Māori to have a significantly (p<.05) lower rate of attendance.

41

4.3 Māori cultural identity and tertiary educational attainment and attendance

As explained previously, the cultural identity of the Māori cohort was classified into three groups on the basis of self-reported ethnic status at ages 21 and 25. These three groups were:

• Sole-Māori: Those members of the Māori cohort who stated their only identity was Māori (n=59). • Māori/other ethnic identity: Those members of the Māori cohort who identified as being Māori and also reported another ethnic identity (n=38). • Non-Māori: Those members of the Māori cohort who stated that they did not identify as Māori (n=13).

Figures 4.4 and 4.5 show the associations between self-reported cultural identity and educational achievement and educational attainment. While both figures suggest a small tendency for the Sole- Māori group to have lower educational achievement/attendance than the other groups, in no case were the differences statistically significant. This suggests that rates of educational achievement/attendance within the Māori cohort did not vary with variations in cultural identity.

90 80 Sole Māori 70 Māori/other 60 50 Non-Māori % 40 30 20 10 0 Any secondary Any terary Bachelor's degree or qualificaon qualificaon higher Level of educaonal aainment

Figure 4.5 Educational attainment for the Māori cohort by cultural identity at age 21 and 25.

42

100 90 Sole Māori 80 Māori/other 70 60 Non-Māori % 50 40 30 20 10 0 Any terary educaon Any university educaon Level of educaonal parcipaon

Figure 4.6 Educational participation for the Māori cohort by cultural identity at age 21 and 25.

4.4 Summary and conclusions

In this chapter, we have examined the secondary and tertiary educational history of the Māori cohort during adolescence and young adulthood up to age 30.

• Consistent with the findings in previous chapters, there were small to moderate associations between ethnicity and educational achievement, with Māori having lower levels of attainment as measured by: attainment of any secondary school qualification by age 21; attainment of any tertiary qualification by age 30; and attainment of a Bachelor’s degree or higher by age 30. The most marked difference was for degree attainment, with non-Māori having a rate of degree attainment (before statistical adjustment) that was nearly three times higher than the rate for Māori. Māori also had lower rates of tertiary educational participation as measured by: participation in any tertiary education by age 30; and enrolment at a recognised university by age 30. • Nearly all of the associations between ethnicity and educational attainment/achievement were explained by differences in the socio-demographic backgrounds of the Māori and non- Māori cohort. However, even following statistical control a small but significant (p<.05) association remained between ethnicity and enrolment in any tertiary education. Significant socio-demographic factors included: maternal educational attainment, family socio- economic status and maternal age. These findings suggest that when due allowance was made for socio-demographic factors there was no consistent evidence to suggest that Māori ethnicity was associated with increased rates of educational underachievement.

43 • An examination of the relationship between Māori cultural identity and educational achievement/attendance showed that there were no detectable linkages between cultural identity and educational achievement/attendance.

In summary, the findings above suggest that Māori have lower educational attainment and participation in tertiary education. These findings are consistent with previous suggestions of Māori educational disadvantage (Earle, 2008; Wang, Harkess, & Parkin, 2006). What the findings of this study strongly suggest is that the origins of these differences lie with socio-demographic (maternal educational attainment; family socio-economic status; family type; maternal age) rather than cultural differences.

44 Chapter 5: Summary, conclusions and recommendations

In this report we have used data from the Christchurch Health and Development Study to trace the educational history of a cohort of 142 Māori children born in Christchurch/Ōtautahi in 1977. The study has used a developmental life-course approach which has examined Early Childhood Education (ECE), primary and intermediate school achievement, and secondary and tertiary educational attendance and achievement.

Each stage of the analysis addressed a series of common research questions. These questions were:

• To what extent was the Māori cohort at an educational disadvantage when compared with the non-Māori cohort? • To what extent could any educational differences between the Māori and non-Māori cohorts be explained by socio-demographic factors (maternal educational attainment, family socio- economic status, family type and maternal age) assessed at birth? • To what extent did rates of educational participation and achievement vary with cultural identity?

The findings of the study revealed remarkable consistency in the answers to these questions over the life-course of the cohort from birth to 30 years. The major trends in the findings and their theoretical implications are reviewed below.

5.1 Māori/non-Māori differences in educational achievement

In confirmation of previous research and official statistics (Biddulph, Biddulph, & Biddulph, 2003; Earle, 2008; Ministry of Education, 2010, 2011; Peterson & Williams, 2000; Wang, et al., 2006), there was consistent evidence to show that the Māori cohort had lower levels of educational participation and lower levels of educational attainment. These differences were evident over the life-course of the cohort and were manifested in different ways at different developmental stages:

• During the preschool years, Māori children had lower rates of participation (89.4%) in ECE than non-Māori (95.5%). While this difference was relatively small, it was nonetheless statistically significant (p=0.007). • During middle childhood there were also small but consistent tendencies for Māori to perform less well on both teacher ratings of classroom performance and standardised tests of reading and mathematics. • The trends observed during the preschool and primary years were sustained at secondary school and tertiary education, with Māori performing less well in national school examinations (School Certificate, Sixth Form Certificate, Higher School Certificate or University Bursary/Scholarship,) having less participation in tertiary education, and lower rates of degree attainment.

45 To summarise the strength of the associations between ethnicity and educational achievement over the life-course, Figure 5.1 shows a plot of values of the Cohen’s d statistic for each outcome examined during the preschool, school and post-school periods. Cohen’s d is a measure of size of effect that is widely used to compare study findings (Cohen, 1988). In describing the d statistic, Cohen suggests that a value of d of .20 can be considered to be small, whereas a value of .50 can be considered to be moderate. Figure 5.1 shows the unadjusted associations between ethnicity and educational achievement. These are given by the blue bars in the graph and show that the associations between ethnicity and achievement fell within the range of small to moderate; unadjusted values of d ranged from .16 to .42 (median=.31). Nonetheless, all of the unadjusted associations between ethnicity and educational achievement were statistically significant.

The figure also shows that there was no general trend for effect sizes to vary over the life-course. The only possible exception to this was for degree attainment, which had a larger d value than the other outcomes. The main message to be taken from Figure 5.1 is that ethnicity was associated with small to moderate effects on educational outcomes throughout the life history of the cohort.

There are two perspectives that can be taken on these consistent differences between Māori and non-Māori. The first is the traditional ‘deficit’ perspective that emphasises the educational under- achievement of Māori and fails to recognise Māori achievement (Biddulph, et al., 2003; Bishop, Berryman, Cavanagh, & Teddy, 2009; Bishop, Berryman, Tiakiwai, & Richardson, 2003; Ministry of Education, 2009a). The alternative is a strength-based perspective which recognises that in nearly all areas assessed, Māori often have quite substantial achievement and participation in education, with levels of these outcomes only being slightly lower than those of non-Māori (Biddulph, et al., 2003; Chapple, 2000; Education Review Office, 2010; Ministry of Education, 2009a).

It is the view of this report, that the strength-based approach provides a more accurate perspective on the achievement of the Māori cohort and avoids transforming what are relatively small, albeit consistent disparities, between Māori and non-Māori into major deficits. The good news that emerges from much of this research is that for the CHDS cohort, the educational gap between Māori and non-Māori, while evident, is not very large. The important implication of this conclusion is that there is every reason to believe that with appropriate investment and collective effort, the gap between Māori and non-Māori educational achievement and participation can be closed.

46

Measure Cohen’s d

EARLY CHILDHOOD EDUCATION (ECE)

% ever aended ECE

Mean number of years aended ECE

SCHOOL ACHIEVEMENT

Mean Grade Point Average Age 7

Mean Grade Point Average Age 10

Mean Grade Point Average Age 13

Mean BURT Word Recognion Age 8

Mean BURT Word Recognion Age 11

Mean BURT Word Recognion Age 13 Adjusted Cohen's d Mean PAT Mathemacs Age 11 Unadjusted Mean PAT Reading Comprehension Age 12 Cohen's d

SECONDARY SCHOOL/TERTIARY EDUCATION

% with secondary school qualificaons

% with terary qualificaons

% with Bachelor's degree or higher

% aending terary educaon

% aending university

0 0.1 0.2 0.3 0.4 0.5 0.6 Small Moderate

Figure 5.1 Stem-and-leaf plot of unadjusted and adjusted standardised differences (Cohen’s d) between the Māori and non-Māori cohorts.

5.2 The role of socio-economic factors

An on-going debate about the origins of Māori/non-Māori disparities concerns the extent to which these disparities are socio-economic in origin (Biddulph, et al., 2003; Chapple, 2000; Durie, 2005; O'Neill & Nash, 2005; Strathdee, 2013; Wylie, 2001). Specifically, it has been suggested on many occasions that the reasons for lower levels of educational achievement (and other adverse outcomes) amongst Māori are due to the following:

47

• As a population, Māori are generally more socio-economically disadvantaged (Chapple, 2000; Marie, Fergusson, & Boden, 2008; Perry, 2012; Statistics New Zealand, 2002); • It is well established that children from socio-economically disadvantaged families fare less well in the education system than children from economically advantaged families (Biddulph, et al., 2003; Ministry of Education, 2004; Ministry of Finance, 2012).

Proponents of this position point to a series of processes, including institutional racism, urban migration and intermarriage, which have combined to place a disproportionate fraction of Māori into socially and economically disadvantaged groups (Chapple, 2000; Kukutai, 2010). An important implication of the socio-economic explanation is that the principal route for reducing ethnic educational disparities is to develop policies that foster the upward social mobility of Māori (Chapple, 2000; Snook & O’Neill, 2010).

To test the socio-economic hypothesis, the associations between ethnicity and educational achievement over the life-course were adjusted for a series of socio-demographic factors assessed at birth. These factors included: maternal educational attainment, maternal age, family socio- economic status and family type. Figure 5.1 shows the adjusted associations (in red) and compares these with the unadjusted results (in blue). The figure shows that the adjusted values of Cohen’s d were small and fell into the range of .01 to .24 (median value=.12). With two exceptions (ECE attendance at age 4-5, p=0.020; attendance at any tertiary educational facility, p=0.017), the associations between ethnicity and educational achievement became statistically non-significant, following adjustment for socio-demographic factors. These findings provide strong support for the socio-economic hypothesis and suggest that most, if not all, of the ethnic disparities in educational achievement were accounted for by socio-economic factors.

These findings are consistent with results from the Competent Children, Competent Learners project which also examined the associations between ethnicity and educational achievement after adjustment for socio-economic factors (Wylie, 2001). The principle conclusions of this study were:

While we found some differences in mathematics and literacy scores for children who came from different ethnic groups, most of these differences were reduced or were no longer significant once we took family income and maternal qualification into account. In other words, it is the resources available to children which matter to their progress, not their culture or ethnicity (Wylie, 2001, p.31).

The fact that two major New Zealand longitudinal studies (CHDS; Competent Children, Competent Learners project) of ethnic differences in educational achievement show that these differences can be explained by socio-economic factors, clearly suggests the importance of these factors in the educational disadvantage of Māori.

Perhaps the best explanation for these associations is provided by what has been described as the Family Resource Framework (O'Neill & Nash, 2005). This framework assumes that the educational achievements of children are strongly shaped by family resources, environment and social capital. Under this model, the higher achievement of children from socially advantaged backgrounds arises because these children are raised in families which are characterised by a constellation of factors which act to encourage, sustain and reward educational achievement. In contrast, children from socio-economically disadvantaged families lack these benefits. These conclusions imply that the

48 educational success of children from socially advantaged families does not arise simply from the higher incomes of these families, but from a combination of family related factors, all of which combine to increase levels of educational achievement for children from these backgrounds.

These findings clearly suggest that the most important route to addressing educational disadvantage amongst Māori is to reduce more general disparities in the socio-economic distributions of the two populations. This conclusion in turn raises important issues about the type of educational, social, financial and related policy needed to produce a more equal distribution of socio-economic resources amongst Māori and non-Māori (Chapple, 2000; Codd & Openshaw, 2005; Fitzsimons & Smith, 2000; Snook & O’Neill, 2010; Strathdee, 2013). A number of possible policies are addressed below. All of these policies require various forms of affirmative action to reduce Māori/non-Māori disparities in educational achievement, occupational participation and related outcomes.

While it is possible to list policies and approaches that may mitigate Māori/non-Māori socio- economic disparities, it is clear that there is no `quick fix’ for addressing these issues; change is likely to evolve over a prolonged period during which consistent investments in addressing Māori/non- Māori socio-economic disparities are required (Codd & Openshaw, 2005; Ministry of Finance, 2012; Strathdee, 2013).

5.3 The role of cultural factors and cultural identity

An alternative explanation of ethnic disparity in educational achievement is that this disparity is the result of cultural rather than socio-economic factors (Bishop, et al., 2009; Durie, 2005; Macfarlane, Glynn, Cavanagh, & Bateman, 2007; Macfarlane & Macfarlane, 2012). The cultural explanation assumes that because of a series of family, social, language and related problems caused by the process of colonisation, the Māori child enters the New Zealand education system at an inherent disadvantage when compared with a child from a Pākehā family (Bishop, 2003; Bishop, et al., 2009; Fitzsimons & Smith, 2000). Proponents of this view, point to the role of epistemological racism which places Māori students at a disadvantage within a mono-cultural school system (Bishop, 2003; Bishop, et al., 2009; Fitzsimons & Smith, 2000; Macfarlane, et al., 2007). The implication of cultural explanations is that the appropriate route for resolving educational disparities is through the development of educational systems that place greater emphasis on Māori culture, language and values (Bishop, 2003; Bishop, et al., 2009; Durie, 2005; Fitzsimons & Smith, 2000; Macfarlane, 2008; Macfarlane, et al., 2007; Macfarlane & Macfarlane, 2012; Ministry of Education, 2009b, 2011).

While the findings of the CHDS suggest that Māori/Non-Māori disparities in education were explained by socio-demographic factors, these findings require careful interpretation and discussion. First, it is important to recognise that the findings in this report relate to a birth cohort which started its educational history over 30 years ago. It is clear the educational services, systems and practices of that time were very different from the contemporary situation (Codd & Openshaw, 2005; Duncan, 2004; May, 2002). For example, while the role of Māori language, culture and practice is widely recognised in the 21st century, there would have been far less recognition of this over the period when this cohort received its education (Pihama, 1993; Smith, 2003). It may be that because of the limited attention given to the role of cultural factors in the 1980’s and 1990’s in New Zealand, the role of these factors was not evident in the educational history of the Māori cohort.

49 For these reasons, it may be misleading to conclude that cultural factors play no role in the educational success of the Māori cohort. As shown in a number of studies, educational institutions and programmes that are culturally responsive are associated with increased achievement by Māori (Biddulph, et al., 2003; Bishop, et al., 2009; Education Review Office, 2010; Macfarlane & Macfarlane, 2012). This evidence clearly suggests the importance of recognising culture within the New Zealand education system. For these reasons, it may be misleading to use the present evidence as an argument against the increased recognition of Māori culture and values within the New Zealand educational system.

In addition, it may be strongly argued that the purposes of education are not solely to teach young people skills in a specified series of academic subjects, but rather to provide a wider preparation for later life (Codd & Openshaw, 2005; Fitzsimons & Smith, 2000; Ministry of Education, 2009a; Smith, 2003). From this perspective, educational institutions which support and recognise the individual’s cultural background play an important role in the educational and socialisation process, even if these features do not facilitate achievement in a narrowly defined set of academic skills. Looked at from this perspective, the justification for culturally appropriate and responsive education for Māori is that this education will provide young Māori with the life skills and abilities to live successfully within a Māori cultural context.

5.4 Kaupapa Māori

A requirement of this report is that it be underpinned by kaupapa Māori. From the outset, it has been made clear to the Ngāi Tahu Research Centre (NTRC) by Te Here of Te Runanga o Ngāi Tahu that the data and information collected and collated by the CHDS is significant to the tribe. At the same time, Ngāi Tahu are aware that the CHDS was initiated at a time when few researchers were aware of any theoretical basis for understanding an epistemology that is Māori, let alone a research method specific to Māori. Nonetheless, the fact that the CHDS has managed to compile a longitudinal study over the last three decades on tribal members and other iwi means that the insights gained by the CHDS are important.

Because of the importance Ngāi Tahu places on the CHDS there are comments that the NTRC wish to provide.

The CHDS report has argued that there is consistent evidence to show that its Māori cohort had low levels of educational participation and achievement. In explaining the gap between Māori and non- Māori the CHDS report suggests that the differences are not that marked when aligned with the non-Māori cohort in the same socio-economic grouping. Overall, children from socio-economically disadvantaged families do less well in the education system. This finding echoes the Marie, Fergusson, Boden article of 2008 that was published in the Australian Journal of Education (Marie, et al., 2008). The implications of this finding are important for Te Runanga o Ngāi Tahu to understand. The CHDS report makes it clear that it would be a mistake to explain these differences on the basis of ethnicity. That is, poverty and a low socio-economic status will have a greater impact on low educational achievement and participation than ethnicity – or being Māori.

However, locating the reasons for low achievement as being a matter of socio-economic status does not mean that culture has no role in providing a solution within the education sector. The CHDS

50 plainly states that it would be wrong to conclude that culture does not have a role in the educational success of Māori. The CHDS references studies that show culturally responsive institutions increase Māori achievement levels. Therefore a kaupapa Māori approach to education provides a solution to resolving the problems of Māori achievement and participation. It does not address the underlying issue of poverty.

What is important to note, is that the preceding analysis provides no account of the processes that led the Māori cohort to being disadvantaged relative to the non-Māori cohort. These processes are likely to reflect a complex mix of historical factors that have combined to place Māori at a relative disadvantage within New Zealand. Further, comparing Māori and non-Māori within the same socio- economic index assumes identical explanations for each group’s relative poverty because of the equally tacit assumption that Crown legislation and economic policy are ethnically neutral. This is not the case and this is where a nuanced kaupapa Māori assessment is required.

Historians generally agree that colonization, the alienation of land and the capital base of Māori was a clear reason for the growing poverty of Māori. This can be seen in the numerous Royal Commissions initiated by the Crown in the later part of the nineteenth century and the early part of the twentieth that were concerned with the apparent decline of Māori and their expected disappearance as a people (Waitangi Tribunal, 1991). It is from this period that Dr. Isaac Featherston spoke of the European responsibility to ‘smooth the pillow of a dying race’. The situation was especially bad in the , although in the end all Māori communities suffered appalling levels of poverty.

In 1890, Judge Alexander Mackay was appointed to a Royal Commission to investigate which South Island Māori lacked sufficient land for their self-support. Mackay proceeded to give a harrowing outline of poverty in the South Island that affected all sectors of the tribe from children to working and able-bodied men to elders in distress. One tribal leader told the commission, ‘ . . . it would be better to be dead and out of the way . . . as there does not appear to be any place for us in the future’. The Mackay report is explicit in how systemic the poverty was amongst not only tribal members but also ‘half-castes’ – a group often ignored by the Crown. The lack of training to learn a skilled trade meant Māori could only work as farm labourers; the poor diet many suffered and the destitution of the elderly were all traumatic accounts recorded by Mackay (Evison, 1997). Mackay concluded that the root of all this suffering was due to the fact that Ngāi Tahu had not been allocated enough land and a greater endowment was required (Waitangi Tribunal, 1991, Ch. 20). Nothing was to happen. The great irony was that at the same time the Liberal Party took Office under the leadership of John Balance who initiated the land reform programme, breaking up the large landed estates for small farmers (Evison, 1997). Crown policy was directed towards reallocating capital and wealth to different sectors of the settler community. It did not extend to Māori. Instead the Crown directed its legislation towards impoverishing Māori. In assessing the economic decline of Ngāi Tahu from the mid to later part of the nineteenth century, the Waitangi Tribunal found:

. . . the extent to which settlement was interfering with the Ngai Tahu economy was clear, as was the inability of the tribe to make any economic gain out of the reserves granted to them in 1868 . . . . As Ngai Tahu's economic life became more and more restricted to the reserves by settlement, their condition worsened. Economic depression in the 1880s further

51

accentuated their disadvantage, making it even more difficult for Maori to gain employment (Waitangi Tribunal, 1991, Ch. 18.4, para. 12-13).

The situation for Ngāi Tahu was the same for all Māori. Crown legislation and Native Land Court policy had removed the capital base from Māori, thereby creating poverty. Furthermore, this active process of directing legislation towards the transfer of Māori land to the Crown continued through to the 1960s.

The 1958 Town and Country Planning Act, 1967 Ratings Act and the 1967 Māori Affairs Amendment Act were all designed to see the transfer of any remaining Māori land from its owners to the rural farmer economy. In its Central North Island Report, ‘Maunga ’, the Waitangi Tribunal referred to the Crown’s ‘systematic dispossession of Māori’, when examining the land legislation post 1950 (Waitangi Tribunal, 2008). That Māori feature in the lower socio-economic index is a consequence of well over a century and a half of combined legislation and policy by the Crown that was designed to remove any residues of wealth or capital from Māori. The fact that legislation was targeted towards Māori to create poverty simply indicates institutional racism as the base reason for the socio- economic status of the Māori cohort. A further factor that may have encouraged the relative disadvantage of Māori is high rates of intermarriage with disadvantaged non-Māori partners. These trends are clearly evident in the CHDS cohort where over 80% of children classified as Māori had one non-Māori parent (see Table 1.1).

5.5. Concluding comments

The findings reviewed above lead to the following major conclusions about the linkages between ethnicity and educational achievement within the CHDS cohort.

• First, there was clear and consistent evidence of Māori educational disadvantage from early childhood through to tertiary education. These differences however, fell into the small to moderate range with the implication that substantial numbers of young Māori were successful within the New Zealand education system. For example, in this study over two- thirds of young Māori left school with an educational qualification (see Chapter 4, Figure 4.1). • Adjustment for socio-demographic factors showed that nearly all of the associations between ethnicity and educational achievement/attendance could be explained by socio- demographic differences between the Māori and non-Māori cohorts. Key socio-demographic factors included: maternal educational attainment, family socio-economic status and maternal age. • There was no significant association between cultural identity and educational achievement/attendance over the life-course. Those of sole-Māori identity had similar levels of attainment to those holding Māori/other ethnic affiliation and non-Māori identities. • Collectively these findings lead to two general conclusions. The first conclusion is that the major route to reducing ethnic differences in educational attainment is through social policies which foster the upward social mobility of young Māori. • At the same time it is important to recognise the role of the educational system in sustaining Māori language, culture and values. For these reasons, policies which seek to increase the

52 upward social mobility of Māori need to be accompanied by policies aimed at sustaining Māori language, culture and values. • These considerations suggest the importance of carefully balancing two viewpoints on Māori education, reflecting the socio-economic and cultural perspectives. The best route to addressing these issues is probably through the development and evaluation of new policies, strategies and approaches based around the integration of the socio-economic and cultural perspectives.

53 References

Biddulph, F., Biddulph, J., & Biddulph, C. (2003). The complexity of community and family influences on children's achievement in New Zealand: Best evidence synthesis. Wellington: Ministry of Education. Bishop, R. (2003). Changing power relations in education: Kaupapa Māori messages for "mainstream" education in Aotearoa/New Zealand. Comparative Education, 39(2), 221-238. Bishop, R., Berryman, M., Cavanagh, T., & Teddy, L. (2009). Te Kotahitanga: Addressing educational disparities facing Māori students in New Zealand. Teaching and Teacher Education, 25(5), 734-742. Bishop, R., Berryman, M., Tiakiwai, S., & Richardson, C. (2003). Te Kōtahitanga: The experiences of Year 9 and 10 Māori students in mainstream classrooms. Wellington: Māori Education Research Institute. Bishop, R., & Glynn, T. (1999). Culture counts: Changing power relations in education. Palmerston North: Dunmore Press. Broughton, J., Fergusson, D. M., Rimene, C., Horwood, L. J., & Sporle, A. (2000). Nga Tini Aho O Te Ao Hou. The many strands of contemporary Maori society: Maori ethnicity and identity in the Christchurch Health and Development Study. Dunedin: Ngai Tahu Maori Health Research Unit. Chapple, S. (2000). Maori socio-economic disparity. Political Science, 52. Codd, J., & Openshaw, R. (2005). The education system in Aotearoa New Zealand. In P. Adams, R. Openshaw & J. Hamer (Eds.), Education and society in Aotearoa New Zealand (pp. 155-186). Australia: Cengage Learning. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. Cormack, D. (2007). The Māori population. In B. Robson & R. Harris (Eds.), Hauora: Māori standards of health IV. A study of the years 2000-2005. Wellington: Te Rōpū Rangahau Hauora a Eru Pōmare. Retrieved from http://www.hauora.maori.nz/downloads/hauora_complete_web.pdf. Duncan, J. (2004). Misplacing the teacher? New Zealand early childhood teachers and early childhood education policy reforms, 1984-96. Contemporary Issues in Early Childhood, 5(2), 160-177. Durie, M. (1994). Te Ara Ahu Whakamua: Proceedings of the Māori Health Decade Hui, . Wellington: Te Puni Kōkiri. Durie, M. (2005). Race and ethnicity in public policy: Does it work? Social Policy Journal of New Zealand, 24, 1-11. Earle, D. (2008). Hei titiro anō i te whāinga – Māori achievement in bachelors degrees revisited. Wellington: Ministry of Education. Education Review Office. (2010). Promoting success for Māori students: Schools’ progress. Wellington: Education Review Office. Elley, W. B., & Irving, J. C. (1976). Revised socio-economic index for New Zealand. New Zealand Journal of Educational Studies, 11(1), 25-36.

54 Elley, W. B., & Reid, N. A. (1969). Progressive Achievement Tests: Teacher manual: Reading comprehension, reading vocabulary. Wellington: New Zealand Council for Educational Research. Evison, H. C. (1997). The Long Dispute: Maori Land Rights and European Colonisation in Southern New Zealand. Christchurch: Canterbury University Press. Fergusson, D. M., & Horwood, L. J. (2001). The Christchurch Health and Development Study: Review of findings on child and adolescent mental health. Australian and New Zealand Journal of Psychiatry, 35(3), 287-296. Fergusson, D. M., Horwood, L. J., Shannon, F. T., & Lawton, J. M. (1989). The Christchurch Child Development Study: A review of epidemiological findings. Paediatric and Perinatal Epidemiology, 3(3), 278-301. Fitzsimons, P., & Smith, G. (2000). Philosophy and indigenous cultural transformation. Educational Philosophy and Theory, 32(1), 25-41. Gilmore, A., Croft, C., & Reid, N. (1981). Burt Word Reading Test. New Zealand Revision. Teacher's Manual. Wellington, New Zealand: New Zealand Council for Educational Research. Kukutai, T. H. (2010). The thin brown line: Re-indigenizing inequality in Aotearoa New Zealand. Ph.D., Stanford University, Stanford. Lang, K. (2002, 29.1.2013). Measuring ethnicity in the New Zealand population census. Retrieved from http://www.stats.govt.nz/Census/2001-census-data/change-in-ethnicity-question.aspx Macfarlane, A. H. (2008). Kia hiwa rā! Listen to culture: A counter narrative to standard assessment practices in psychology. The Bulletin, 111, 33-36. Macfarlane, A. H., Glynn, T., Cavanagh, T., & Bateman, S. (2007). Creating culturally-safe schools for Māori students. Australian Journal of Indigenous Education, 36, 65-76. Macfarlane, A. H., & Macfarlane, S. L. (2012). Weaving the dimensions of culture and learning: Implications for educators. In B. Kaur (Ed.), Understanding teaching and learning: Classroom research revisited (pp. 213-224): Sense Publishers. Retrieved from http://link.springer.com/content/pdf/10.1007%2F978-94-6091-864-3_15.pdf. doi: 10.1007/978-94-6091-864-3_15 Marie, D., Fergusson, D. M., & Boden, J. M. (2008). Educational achievement in Maori: The roles of cultural identity and social disadvantage. Australian Journal of Education, 52(2), 183-196. May, H. (2002). Aotearoa-New Zealand: An overview of history, policy and curriculum. Journal of Education, 37(1), 19-36. Ministry of Education. (2004). Focus on low SES students’ achievement in reading literacy results from PISA 2000. Wellington: Ministry of Education. Ministry of Education. (2008). New Zealand education system overview. Wellington: Ministry of Education. Ministry of Education. (2009a). Ka Hikitia - Managing for success. Wellington: Ministry of Education. Ministry of Education. (2009b). Ngā Haeata Mātauranga – The annual report on Māori education, 2007/08. Wellington: Ministry of Education. Ministry of Education. (2010). Participation and attainment of Māori students in National Certificate of Educational Achievement. Wellington Ministry of Education. Ministry of Education. (2011). Briefing to the incoming minister. Wellington: Ministry of Education. Ministry of Education. (2013). Secondary school qualifications prior to 2001. Retrieved from http://www.nzqa.govt.nz/qualifications-standards/results-2/secondary-school- qualifications-prior-to-2002/

55 Ministry of Finance. (2012). Treasury’s advice on lifting student achievement in New Zealand: Evidence brief. Wellington: Ministry of Finance. New Zealand Qualifications Authority. (2010). Approval and accreditation of courses leading to degrees and related qualifications for Institutes of Technology and Polytechnics (ITPs). Wellington: New Zealand Qualifications Authority. O'Neill, A., & Nash, R. (2005). Family explanations of educational attainment. In Paul Adams, Roger Openshaw & Judy Hamer (Eds.), Education and Society in Aotearoa (Vol. 2, pp. 327-346). Australia: Cengage Learning. Perry, B. (2012). Household incomes in New Zealand: Trends in Indicators of inequality and hardship 1982 to 2011. Wellington: Ministry of Social Development. Peterson, D. R., & Williams, D. (2000). Gaps between ethnic groups: Some key statistics. Wellington: Parliament Library. Pihama, L. (1993). Development of early childhood services in Aotearoa. Tungia te Ururua, kia tupu whakaritorito te tupu o te harakeke: A critical awareness of parents as first teachers. Unpublished MA Thesis. Auckland: University of Auckland. Reid, N. A., & Hughes, D. C. (1974). Progressive Achievement Tests teacher's manual: Mathematics. Wellington: New Zealand Council for Educational Research. Smith, G. H. (2003). Kaupapa Maori theory: Theorizing indigenous transformation of education and schooling. Paper presented at the ‘Kaupapa Maori Symposium’ NZARE/AARE Joint Conference, Auckland, New Zealand. Snook, I., & O’Neill, J. (2010). Social class and educational achievement: Beyond ideology. New Zealand Journal of Educational Studies, 45(2), 3-18. Statistics New Zealand. (1997). Information guide: 1996 census of population and dwellings Retrieved from http://www.stats.govt.nz/Census/1996-census-data/1996-census- information-guide.aspx Statistics New Zealand. (2000). Statistical standard for Iwi. Retrieved from http://www2.stats.govt.nz/domino/external/web/carsweb.nsf/Standards/Iwi Statistics New Zealand. (2002). Census snapshot: Maori. Welington: Statistics New Zealand. Strathdee, R. (2013). Neo-tribal capitalism, socio-economic disadvantage and educational policy in New Zealand. Journal of Education Policy. Tertiary Education Commission. (2009). Types of tertiary education. Retrieved from http://www.tec.govt.nz/Tertiary-Sector/Types-of-Tertiary-Education/ Universities New Zealand. (2013). Committee on University Academic Programmes (CUAP). Retrieved from http://www.universitiesnz.ac.nz/aboutus/sc/cuap Waitangi Tribunal. (1991). The Ngai Tahu Report 1991. Wellington: Department of Justice. Waitangi Tribunal. (2008). He Maunga Rongo: Report on Central North Island Claims, Stage 1 (Vol. 1). Wellington: Waitangi Tribunal. Wang, H., Harkess, C., & Parkin, M. (2006). Maori student achievement on the National Qualifications Framework 2002 - 2005 fact sheet. Wellington: Ministry of Education. Wylie, C. (2001). Ten years old and competent. The fourth stage of the competent children project: A summary of the main findings. Wellington: New Zealand Council for Educational Research.

56 Appendix

Table A.1 Comparison of socio-demographic and birth characteristics between Māori cohort members followed to age 30 (n=111) and those lost to follow-up (n=31)...... A-1 Table A.2 Māori ethnicity assessed at birth by self-reported Māori ethnic identity at age 21...... A-2 Table A.3 Māori ethnicity assessed at birth by self-reported Māori ethnic identity at age 25...... A-2 Table A.4 Self-reported ethnic identity at age 21 by self-reported ethnic identity at age 25...... A-2 Table A.5 Socio-demographic characteristics of the self-defined Māori cohort at age 21 (n=114). .. A-3 Table A.6 Socio-demographic characteristics of the self-defined Māori cohort at age 25 (n=122). .. A-4 Table B.1 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and ECE attendance at age 2-3 for socio-demographic factors...... B-1 Table B.2 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and ECE attendance at age 3-4 for socio-demographic factors...... B-1 Table B.3 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and ECE attendance at age 4-5 for socio-demographic factors...... B-2 Table B.4 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and ECE attendance prior to age 5 for socio-demographic factors...... B-2 Table B.5 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and duration of ECE attendance for socio-demographic factors...... B-3 Table C.1 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and teacher-rated GPA at age 7 for socio-demographic factors...... C-1 Table C.2 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and teacher-rated GPA at age 10 for socio-demographic factors...... C-1 Table C.3 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and teacher-rated GPA at age 13 for socio-demographic factors...... C-2 Table C.4 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and the Burt Word Reading Test at age 8 for socio-demographic factors...... C-2 Table C.5 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and the Burt Word Reading Test at age 11 for socio-demographic factors...... C-3 Table C.6 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and the Burt Word Reading Test at age 13 for socio-demographic factors...... C-3

57 Table C.7 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and the PAT (Mathematics) at age 11 for socio-demographic factors...... C-4 Table C.8 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and the PAT (Reading) at age 12 for socio-demographic factors...... C-4 Table D.1 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and attainment of any secondary school qualification for socio-demographic factors...... D-1 Table D.2 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and attained any tertiary qualification...... D-1 Table D.3 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and attained a recognised degree. D-2 Table D.4 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and attendance at any tertiary educational facility for socio-demographic factors...... D-2 Table D.5 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and attended a recognised university...... D-3

58 A. Appendix for Chapter 1

Table A.1 Comparison of socio-demographic and birth characteristics between Māori cohort members followed to age 30 (n=111) and those lost to follow-up (n=31).

Followed to age 30 Lost to follow-up

Measure n %/Mean n %/Mean p

Socio-demographic characteristic

Maternal education 0.519

No school qualifications 72 64.9 23 74.2

Secondary school qualifications 33 29.7 6 19.4

Tertiary qualifications 6 5.4 2 6.5

Family socio-economic status 0.873

Professional/managerial 10 9.0 2 6.5

Clerical/technical/skilled trade 53 47.8 16 51.6

Semi-skilled/unskilled 48 43.2 13 41.9

Family type 0.019

Single-parent household 18 16.2 11 35.5

Two-parent household 93 83.8 20 64.5

Mean (SD) maternal age 23.7 (4.6) 23.2 (4.7) 0.615

Birth characteristic

% Male 48.7 64.5 0.118

Mean (SD) birth weight (g) 3378 (574) 3272 (513) 0.357

% Planning of pregnancy 0.045

Planned pregnancy 43 38.7 6 19.4

Unplanned pregnancy 68 61.3 25 80.7

% Child breast fed 64 57.7 15 48.4 0.358

A-1 Table A.2 Māori ethnicity assessed at birth by self-reported Māori ethnic identity at age 21.

Self-reported Māori ethnic identity at age 21

Māori ethnicity assessed at birth Sole-Māori Māori/other ethnic affiliation Non-Māori

Māori 49 44 20

Non-Māori 3 18 877

Note. The correlation coefficient (r) = +0.79; p<.001

Table A.3 Māori ethnicity assessed at birth by self-reported Māori ethnic identity at age 25.

Self-reported Māori ethnic identity at age 25

Māori ethnicity assessed at birth Sole-Māori Māori/other ethnic affiliation Non-Māori

Māori 41 55 17

Non-Māori 4 22 864

Note. The correlation coefficient (r) = +0.78; p<.001

Table A.4 Self-reported ethnic identity at age 21 by self-reported ethnic identity at age 25.

Self-reported Māori ethnic identity at age 25 Self-reported Māori ethnic identity at age 21 Sole-Māori Māori/other ethnic affiliation Non-Māori

Sole-Māori 28 23 0

Māori/other ethnic affiliation 13 38 8

Non-Māori 1 12 855

Note. The correlation coefficient (r) = +0.87; p<.001

A-2 Table A.5 Socio-demographic characteristics of the self-defined Māori cohort at age 21 (n=114).

Measure n %

Maternal education

No school qualifications 76 66.7

Secondary school qualifications 31 27.2

Tertiary qualifications 7 6.1

Family socio-economic status

Professional/managerial 11 9.7

Clerical/technical/skilled trade 52 45.6

Semi-skilled/unskilled 51 44.7

Family type

Single parent household 17 14.9

Two-parent household 97 85.1

Mean SD

Mean (SD) maternal age (years) 23.4 4.3

A-3 Table A.6 Socio-demographic characteristics of the self-defined Māori cohort at age 25 (n=122).

Measure n %

Maternal education

No school qualifications 80 65.6

Secondary school qualifications 36 29.5

Tertiary qualifications 6 4.9

Family socio-economic status

Professional/managerial 10 8.2

Clerical/technical/skilled trade 61 50.0

Semi-skilled/unskilled 51 41.8

Family type

Single parent household 19 15.6

Two-parent household 103 84.4

Mean SD

Mean (SD) maternal age (years) 23.4 4.8

A-4 B. Appendix for Chapter 2

Table B.1 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and ECE attendance at age 2-3 for socio-demographic factors.

ECE Attendance 2-3 years

Measure B(SE) p

Māori ethnicity -0.022 (0.212) 0.914

Maternal age (years) -0.012 (0.014) 0.379

Maternal education 0.291 (0.087) <.001

Family type -0.175 (0.277) 0.526

Family socio-economic status -0.176 (0.103) 0.088

Table B.2 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and ECE attendance at age 3-4 for socio-demographic factors.

ECE Attendance 3-4 years

Measure B(SE) p

Māori ethnicity -0.375 (0.234) 0.109

Maternal age (years) -0.018 (0.017) 0.283

Maternal education 0.265 (0.119) 0.025

Family type 0.757 (0.277) 0.006

Family socio-economic status -0.267 (0.133) 0.044

B-1 Table B.3 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and ECE attendance at age 4-5 for socio-demographic factors.

ECE Attendance 4-5 years

Measure B(SE) p

Māori ethnicity -0.712 (0.306) 0.020

Maternal age (years) -0.021 (0.025) 0.402

Maternal education 0.326 (0.191) 0.088

Family type 0.498 (0.377) 0.186

Family socio-economic status -0.541 (0.204) 0.008

Table B.4 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and ECE attendance prior to age 5 for socio-demographic factors.

Ever attended ECE

Measure B (SE) p

Māori ethnicity -0.579 (0.351) 0.099

Maternal age (years) -0.019 (0.029) 0.520

Maternal education 0.260 (0.213) 0.223

Family type 0.651 (0.412) 0.114

Family socio-economic status 0.501 (0.231) 0.030

B-2 Table B.5 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and duration of ECE attendance for socio-demographic factors.

Number of years in ECE

Measure B (SE) p

Māori ethnicity -0.134 (0.076) 0.079

Maternal age (years) -0.007 (0.005) 0.169

Maternal education 0.119 (0.032) <.001

Family type 0.240 (0.097) 0.013

Family socio-economic status -0.109 (0.038) 0.004

B-3 C. Appendix for Chapter 3

Table C.1 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and teacher-rated GPA at age 7 for socio-demographic factors.

Teacher-rated GPA at age 7

Measure B(SE) p

Māori ethnicity -0.102 (0.080) 0.201

Maternal age (years) 0.004 (0.005) 0.463

Maternal education 0.172 (0.034) <.001

Family type 0.020 (0.104) 0.841

Family socio-economic status -0.174 (0.040) <.001

Table C.2 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and teacher-rated GPA at age 10 for socio-demographic factors.

Teacher-rated GPA at age 10

Measure B(SE) p

Māori ethnicity -0.051 (0.089) 0.561

Maternal age (years) -0.002 (0.006) 0.979

Maternal education 0.237 (0.037) <.001

Family type 0.021 (0.113) 0.849

Family socio-economic status -0.149 (0.044) <.001

C-1 Table C.3 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and teacher-rated GPA at age 13 for socio-demographic factors.

Teacher-rated GPA at age 13

Measure B(SE) p

Māori ethnicity -0.050 (0.095) 0.596

Maternal age (years) 0.009 (0.006) 0.155

Maternal education 0.267 (0.041) <.001

Family type 0.033 (0.124) 0.789

Family socio-economic status -0.178 (0.048) <.001

Table C.4 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and the Burt Word Reading Test at age 8 for socio-demographic factors.

Burt Word Reading Test at age 8

Measure B(SE) p

Māori ethnicity -1.078 (1.106) 0.329

Maternal age (years) 0.069 (0.070) 0.324

Maternal education 2.324 (0.451) <.001

Family type 0.337 (1.358) 0.803

Family socio-economic status -2.292 (0.527) <.001

C-2 Table C.5 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and the Burt Word Reading Test at age 11 for socio-demographic factors.

Burt Word Reading Test at age 11

Measure B(SE) p

Māori ethnicity -1.388 (1.173) 0.236

Maternal age (years) 0.056 (0.072) 0.435

Maternal education 2.622 (0.466) <.001

Family type 0.501 (1.40) 0.721

Family socio-economic status -1.873 (0.550) <.001

Table C.6 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and the Burt Word Reading Test at age 13 for socio-demographic factors.

Burt Word Reading Test at age 13

Measure B(SE) p

Māori ethnicity 0.112 (1.216) 0.926

Maternal age (years) 0.005 (0.075) 0.939

Maternal education 2.408 (0.474) <.001

Family type 2.036 (1.475) 0.167

Family socio-economic status -2.321 (0.568) <.001

C-3 Table C.7 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and the PAT (Mathematics) at age 11 for socio-demographic factors.

PAT (Mathematics) at age 11

Measure B(SE) p

Māori ethnicity -1.149 (0.850) 0.176

Maternal age (years) 0.118 (0.058) 0.041

Maternal education 2.531 (0.370) <.001

Family type 0.607 (1.038) 0.558

Family socio-economic status -3.068 (0.427) <.001

Table C.8 Model parameters and tests of significance (B, SE, p) from an ordinary least squares regression model adjusting the associations between Māori/non-Māori ethnicity and the PAT (Reading) at age 12 for socio-demographic factors.

PAT (Reading) at age 12

Measure B(SE) p

Māori ethnicity -0.075 (0.851) 0.929

Maternal age (years) 0.050 (0.058) 0.381

Maternal education 3.287 (0.371) <.001

Family type 0.493 (1.039) 0.634

Family socio-economic status -2.254 (0.428) <.001

C-4 D. Appendix for Chapter 4

Table D.1 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and attainment of any secondary school qualification for socio-demographic factors.

Attainment of any secondary school qualification

Measure B(SE) p

Māori ethnicity 0.281 (0.236) 0.233

Maternal age (years) -0.062 (0.020) 0.001

Maternal education -0.799 (0.148) <.001

Family type -0.202 (0.295) 0.494

Family socio-economic status 0.689 (0.146) <.001

Table D.2 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and attained any tertiary qualification.

Attained any tertiary qualification

Measure B(SE) p

Māori ethnicity -0.341 (0.218) 0.117

Maternal age (years) -0.004 (0.016) 0.794

Maternal education 0.615 (0.109) <.001

Family type 0.185 (0.295) 0.530

Family socio-economic status -0.575 (0.123) <.001

D-1 Table D.3 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and attained a recognised degree.

Attained a recognised degree

Measure B(SE) p

Māori ethnicity -0.622 (0.320) 0.052

Maternal age (years) 0.037 (0.017) 0.036

Maternal education 0.725 (0.102) <.001

Family type 0.998 (0.619) 0.110

Family socio-economic status -0.971 (0.134) <.001

Table D.4 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and attendance at any tertiary educational facility for socio-demographic factors.

Attendance at any tertiary educational facility

Measure B(SE) p

Māori ethnicity -0.692 (0.292) 0.017

Maternal age (years) -0.019 (0.024) 0.434

Maternal education 0.830 (0.206) <.001

Family type 0.060 (0.416) 0.885

Family socio-economic status -0.401 (0.197) 0.042

D-2 Table D.5 Model parameters and tests of significance (B, SE, p) from a logistic regression model adjusting the associations between Māori/non-Māori ethnicity and attended a recognised university.

Attended a recognised university

Measure B(SE) p

Māori ethnicity -0.237 (0.221) 0.289

Maternal age (years) 0.021 (0.015) 0.169

Maternal education 0.630 (0.099) <.001

Family type -0.143 (0.303) 0.637

Family socio-economic status -0.761 (0.119) <.001

D-3 University of Canterbury Private Bag 4800 Christchurch 8140 New Zealand Email: [email protected] www.ntrc.canterbury.ac.nz