1. Introduction

What are the day-to-day experiences of 16-year-olds in New Zealand? How much of what they value and the way they approach their school learning is related to their out-of-school activities and relationships? How much of the way they approach life in and out of school is shaped by the paths they have previously taken, the successes and supports they have known? What are the main links with their current school participation, achievement, and engagement and these other past and present dimensions in their lives? What is their experience of the new and sometimes controversial secondary qualification system, the NCEA?

These questions have shaped much of the material we gathered with the participants in the Competent Children, Competent Learners study as they turned 16: the young people themselves, their parents, and teachers, and these questions shape this report. It aims to provide both a description of what young people are doing at 16, both those still in school, and those who had already left (6 percent of the young people); and an analysis of the patterns we found, how things are connected across different dimensions of their lives both in space and time.

Our study, which is funded by the Ministry of Education and NZCER, began in 1993 when the young people were in their final early childhood education centre, within the Wellington region (including the Wairarapa and Kapiti Coast). The main aim of the study then was to see what contribution early childhood education made to the development of competencies we thought would be important to being lifelong learners. These included skills and knowledge that are now being threaded through the draft revised curriculum as key competencies. Thus, this study can also provide some particularly relevant insight into not only why they are indeed important for lifelong learning (Wylie & Hipkins, 2006), but how they might be supported and developed.

Like other longitudinal studies, this study shows what it means to develop individual identity within sets of relationships and experiences that occur within social frameworks. There were some real differences evident in the day-to-day experiences of the study participants when they were preschoolers. For children in homes where parents had good education themselves, and sufficient money to provide good resources for learning, more was offered in the way of both support and challenge, particularly around language and symbol use, the prime vehicles for learning. Also like other longitudinal studies, this one also challenges some of our assumptions about the role of different social categories in children’s development. For example, we have found it is the financial poverty in which sole parents often live that lies behind apparently lower levels of performance for some, not the fact that a child has one rather than two parents living with them.

When we have the relevant data, we often find that associations between obvious social categories and differences in competencies are linked in turn to different experiences, which in turn lay down habits and ways of being that consolidate into identities. We have found the framework of “learning identity” or “learning career” (Bloomer & Hodkinson, 2000; Ecclestone & Pryor, 2003) a particularly useful one for making sense of the patterns we find. These researchers invoke the metaphor of a “career” to paint a picture of a dynamic and evolving sense of self as a learner—one that is mediated by the structures and rituals of different learning contexts. From this perspective, young people are actively making and remaking themselves through their relationships and experiences, and these experiences and relationships in turn reflect back to them images of who they are, and what matters in life, what is real.

This is not a closed circle, but more of a spiral. We can discern the imprint of the past in the patterns we found at age-16, but that imprint did not allow us to confidently predict what all of the young people in this study would be doing as we analysed the data they shared with us, as we wrote this report. As they came close to adulthood, there were some who were soaring confidently, but in contexts in which they could encounter both positive and negative risks to their identity; there were others who were closing their relationships and experiences so tightly around them that one did fear for their future well-being; and still others whose next steps could take them along a number of different paths.

The descriptive picture we provide here of what 16-year-olds were doing, thinking, and feeling, is not intended to be representative of all New Zealand 16-year-olds. The Competent Children, Competent Learners sample was originally chosen in relation to the main focus of the first phase of the study, which was the role of early childhood education experiences and quality. This meant our units for sampling were early childhood education types, other than ngā köhanga reo, rather than social characteristics. This and the fact that our sample was chosen from the Wellington region, has resulted in a sample that is not nationally representative in terms of social characteristics. Compared to the national average, our sample has higher proportions of young people from high-income families, and those whose mothers have trade or tertiary-level qualifications, and lower proportions of Māori and Pacific young people. The young people who decided not to continue in the study after age 12 also tended to be from homes with fewer resources. Almost half the young people who were attending decile 1–2 schools when they were aged 14 decided not to participate at age 16, which is likely to be the reason why we see fewer differences associated with this school characteristic in this phase than we saw for the first two years of secondary schooling. Thus, where there are differences in experiences and perceptions associated with these social and school characteristics, our findings will give probably a somewhat more positive picture than a sample that had been drawn to be representative of population and school characteristics.

The table below describes the sample at age 16 in terms of the four social characteristics we analyse in the study.

Our analysis

Data

We have a range of different kinds of data, with different properties, and requiring slightly different forms of analysis. We have data related to categories or groups; scale data, from answers to questions asking young people, their teachers, or adults to rate something; and cluster data, grouping individuals according to their responses across a set of questions.

Categorical data

Some data, for instance maternal qualifications, gender, and attendance, put the young people into groups or categories. Some of the categories, like gender, have no implication of amounts of difference between categories. Some of the categories, like maternal qualifications and attendance, do have some implication of amount of difference (these are sometimes called ordinal or ordered categories). Someone with excellent attendance attended school more often than did someone with very good attendance, say. But the difference between the attendance of two people with good and excellent attendance may not be the same as that between two people with very good and fair attendance. The categories cannot be represented on a numeric scale by numbers that represent the amount or quality of attendance.

Scale data

Other data were derived from responses to a series of questions with responses on Likert-type scales. These data and the competency data were used to form “scales” or numeric measures where the numbers on the scale do give some idea of the relative amount of difference between two measures. We have used a series of scales each with a minimum of 1 and a maximum of 10 to set up our competency measures and all of the measures related to experiences and views (e.g., views of classes; relations with friends) that we have developed. The measures for experiences and views were developed through analysis of Likert-scale items (e.g., where students or teachers were asked to express their level of agreement with a statement).

Cluster data

Many questions asked in the study were of the “tick all that apply”, or multiple response, kind, giving a third category of “raw data”. For such questions, and where we wanted to compare answers across a set of questions, we defined clusters or categories of respondents who have relatively similar response profiles (they tended to have similar patterns of the options they selected).

An overview of the measures and categories we used in our analysis is given in ‎Figure 1. A summary description is given in Appendix 1, along with a table of their means. Their derivation is described in more detail in the technical report accompanying this report, along with the details of our analyses (Hodgen, 2008).

Figure 1: Overview of measures used in this report

Analysis

We wanted to describe the interconnections between experiences, competencies, and views, which we could do in a number of ways, depending on the nature of the data.

In the first instance we have used cross-tabulations and correlations. We generally have reported only the associations based on cross-tabulations that had a chi-square test of independence statistically significant at the p ≤ 0.01 level (indicating a one in 100 odds that the association has occurred by chance). We have generally reported nontrivial correlations,¹ but sometimes include all correlations with variables of interest, to show overall patterns.

When we use models to look at the relationship between two or more measures, we generally report associations that were significant at the 0.01 level, and indicate the relative importance of the explanatory variable by the percentage of the variance or difference in student scores on one variable (e.g., NCEA Level 1 credit numbers) that can be accounted for, or predicted by, another (explanatory) variable (e.g., student attendance levels).

The various measures are all interassociated: this multiplicity of associations is shown when we report the results of the cross-tabulations, correlations, and models involving only two measures (a single explanatory measure or variable). Typically, in the models, each explanatory variable explains a relatively small proportion (often, under 20 percent) of the variability in the outcome measure. To attempt to investigate which variables are most important in explaining differences in the outcome measure, we fitted larger models, with more than one explanatory measure.

The scale explanatory measures or variables included in the multivariate models were selected on the basis of the strength of their association with the outcome measure, and the weakness of their associations with each other (two very strongly associated measures, measuring almost the same thing, could not simultaneously be used as explanatory variables in the model as the second would add little new information—the model would have problems of multicollinearity). The scale and categorical explanatory measures retained in the model were those that were still statistically significant after accounting for all the other variables in the model (they made a unique contribution to the model). Many of our multivariate models accounted for at least 40 percent of the variance in scores (outcome measures).

These multivariate models cannot provide final, definitive answers however, since, like any model, they cannot include all relevant factors, and the pictures they do provide are best regarded as the “tip of the iceberg”, signalling further layers (or interconnections) beneath. The models also cannot be interpreted as providing evidence for causal relationships between variables, not even where one of the explanatory measures predates the outcome measure. For example, enjoyment of reading up to age 14 has shown a strong association with literacy at age 16. This does not mean that enjoyment on its own causes good literacy, merely that the two tend to go together. It is likely that this association points to a “virtuous” cycle with enjoyment of reading leading to more reading (both in quantity and sophistication), and more enjoyment as more is gained from reading experiences, all of which (together with some other factors) result in good levels of literacy a little later in life.

The models fitted give a sense of which variables carry the most weight (account for the most of the variability in the outcome variable), and this may indicate some of the best levers for changing outcomes where this is desirable.

We have tried to describe our findings as simply as we can, but what we are investigating is not simple or one-dimensional. This report is not so much the story of a journey along the road to a single destination, but like discovery in an art gallery, with paintings grouped together in each room, each painting in that room offering fresh insight into a similar group of themes.

The first section of this report describes the study participants’ levels of participation in school (using the Ministry of Education term “presence”), achievement, and engagement in school, analysing the factors that seem to have most bearing on these. These factors include a mix of the previous performance levels and experiences that can be thought of as forming the ground on which an individual student might stand in relation to their current learning opportunities, the nature of those current learning opportunities, and the way they are spending their current time—the kinds of activities and friendships that either support them to grow further, turn them away from their established identities, or entrench them further in circular patterns of behaviour and understanding.

We also comment on the students’ performance on items that are related to the new key competencies introduced into the revised New Zealand Curriculum, which was launched in late 2007.

Next, we describe differences in these factors between the 16-year-olds who had already moved on from school, and those still at school, to get some further insight into the kinds of learning identities that can make the most of school, and those that find their school experiences lacking, and would rather seek fresh experiences, sometimes to have a different kind of learning, sometimes to feel validated in their preference for activities that are not encouraged by school.

The second section in this report takes a closer look at the kinds of current learning opportunities available to those who remained in school through the sets of courses they took, teaching practices in their classes, and what they and their parents were making of the senior school qualification introduced in 2002, the NCEA. We find through our analyses that approaches to the NCEA are largely influenced by previous patterns and experiences, and current classroom experiences, rather than by the new structure of the senior school qualification, which is based on a more modular structure than the previous set of qualifications.

This section also examines teacher and student descriptions of class practice to see how far they already support the new key competencies.

In the third section, we move away from school to look at what the young people (both school stayers and school leavers) were doing with their friends and families, and with their time, and how these different relationships and experiences interlink. While we can point to some steady patterns over time, we also show that the meaning of a particular activity is not drawn just from the activity, but the way it occurs, and its place in the whole picture of an individual’s interconnections.

For example, the support and growth opportunities presented by involvement in organised sport will be quite different for the players who can fit it into a context of friendships that are focused on mutual support, compared with the players who fit it into a context of friendships that take place through risky behaviour.

In the concluding section, we take a closer look at some differences in perceptions and experiences associated with the four social characteristics we focus on in the Competent Children, Competent Learners study (maternal qualification, family income, gender, and ethnicity), before providing some final food for thought by drawing together key themes from the work done for this report.