Literacy and Life Skills for Pasifika Adults - Further Investigation: Results from the Adult Literacy and Life Skills (ALL) Survey

3. Discussion

In this report we looked at the following two questions:

• In the ALL data, for Pasifika adults, is educational attainment or is familiarity with English most closely associated with English literacy?
• For Pasifika adults, are the different patterns of change in prose and document literacy between 1996 and 2006 attributable to different relationships between educational attainment, familiarity with English and each of these two forms of English literacy evident in the ALL data?

The answer to the first question is that educational attainment (in the form of time spent in formal education) is statistically significantly more closely related to both prose and document literacy skills than is familiarity with English. However familiarity with English is more closely related to both prose and document literacy than any of age, gender, labour force status, educational attainment, language most frequently spoken when at home, first language, place of birth, and income (and for most of these factors the difference is statistically significant).

The answer to the second question hinges on the existence of differences in the relationship between prose literacy, familiarity with English and educational attainment and that between document literacy, familiarity with English and educational attainment. Looking at the Stage 3 estimates in Appendix Tables 1 and 2, we can see that there are differences in the magnitudes of the coefficients, but that these differences are small. Although the models in this document don’t account for all factors that could potentially influence these relationships, they do account for many of the important ones. Consequently, for Pasifika adults, we can reasonably conclude that there is no strong evidence in the ALL data to suggest that any differences between performance in prose and document literacy are attributable to different relationships between educational attainment, familiarity with English and each of these two forms of English literacy.

So what can we conclude about the relationship between changes in the distributions of educational attainment and familiarity with English between 1996 and 2006 and changes in the distributions of prose literacy and document literacy over the same period?

There are two possible conclusions we could draw. The first is that the relationships between each of prose and document literacy and educational attainment and familiarity with English remained relatively stable from 1996 to 2006. The second is that these relationships changed between 1996 and 2006.

Suppose we were to hypothesise the first of these possible conclusions. In that case, there are relatively large changes in the distributions of educational attainment and familiarity with English. But there are only small differences in the relationships between each of these and prose and document literacy. Presumably there would have been a trade-off between increases in average literacy skill due to increases in average educational attainment offset by decreases in average literacy skill due to increases in the proportion of Pasifika adults who are unfamiliar with English.

Suppose now, we were to hypothesise the second, and more complex, of these possible conclusions. In that case, the changing relationships between educational attainment and familiarity with English and each of prose and document literacy would also contribute to the changes and differences in the distributions of each of these literacy skills.

In the absence of other data, these two conclusions might have quite different policy implications. For example, if the relationships between English literacy and other factors were relatively stable (case 1), then this stability might be expected to continue in the medium term. Consequently, policy might encourage shifts in the distribution of background factors among Pasifika adults toward conditions associated with higher English literacy skill. If instead the relationships between English literacy and other factors were changing (case 2), then close analysis of the direction of those changes together with information about changes in distribution would be needed to inform well-directed policy.

In either case however, well-directed policy will be informed by the possibility of unintended outcomes and will have contingencies to ameliorate these. In the present setting, the potential socio-cultural costs of encouraging familiarity with English among Pasifika adults should not be ignored.

To determine which of these two cases held in future, the analyses in the current document would need to be repeated using data similar to that of the ALL survey collected some time in the future.