Dowload and study the file:
This contains International Adult Literacy Survey (IALS) tables 6.2, 6.3 and 6.4 from Carey, S et al, (1998) Adult Literacy in Britain. London: The Stationery Office in MS Excel worksheet format.
Age and education are known to be correlated as access to education has changed over recent decades and both age and education are independently correlated with performance on the direct assessments of literacy. How countries perform relative to each other will partly be a function of how age and education and other characteristics are distributed in the population. The distribution of the population in terms of age is not the same in each country. For example, Germany has a higher percentage of its population in the oldest age-group than other countries while the US distribution tends toward the younger end of the age range. If all other things were equal you would expect Germany to have a greater proportion at the lower literacy levels than the US, just because of their age distribution. Similarly in comparing the distribution of educational attainment the Netherlands had a much higher proportion educated to second level, 1st stage (ISCED 2) than the other countries. Canada and the United States had higher proportions of their population who had completed third level education at that time. Almost one in four Americans (24%) aged 25-64 had completed a university programme compared with just 12% of the UK population in the same age range and 13% of Germans.
These differences were not simply due to the different age profiles of the countries, there were differences in educational attainment within age-groups or cohorts. Table 6.4 shows the percentage who had attained at least upper secondary education (ISCED 3) in each country for different age-groups. Among the older age-groups, 84% of those aged 45-54 in Germany and 85% of the same age group in United States had completed at least upper secondary education compared with about 70% of the population of Canadian, Swedish and British population. Apart from the Netherlands the proportion of those aged 25-34 who attained at least upper secondary education in each country is fairly similar ranging from 82% of Canadians to 90% of Germans in that age-group.
It is evident from the distribution of literacy skills as assessed in the survey that the differing age and education distributions alone do not explain differences in the profile of literacy skills between countries. Although the Swedish population was very similar to that of Britain both in terms of the age profile of the labour force and the distribution of educational attainment, even across age-groups, the distribution of literacy skills in the two countries is quite different. Sweden had larger proportions of people performing at the higher literacy levels in each age-group and in each level of education compared with Britain.
The opposite phenomenon was also evident when comparing Britain and the US. Both countries had similar age profiles and in all but the youngest age-group the US had higher proportions of the population who had third level education yet the literacy profiles of the two countries were very similar.
When education was held constant there was no consistent pattern across countries other than Sweden who consistently outperformed those in other countries.
This type of external benchmarking and setting the context in which the observed measures can be set is useful in understanding the observed differences. Without it the reader doesn't know whether the measures are simply a reflection of the characteristics of the population in those countries.
In many international surveys the analysis of the correlates of the distribution are often more illuminating than the comparison of the distribution alone. For example it may be more interesting to look at whether the relationships between education and literacy is of the same power across countries rather than focusing on which countries have higher scores than others.
