Endnotes
|
| 1. |
The IALS sample of youth for the United States was relatively
small and is probably biased because it did not adequately represent
college-bound youth. However, the National Adult Literacy Survey (NALS)
comprised a large nationally representative sample of American youth and used
the same literacy tests as the IALS. Therefore, this study based its analyses
on the NALS data for the United States, and on IALS data for Canada and five
European countries. |
| 2. |
The reader should bear in mind, though, that a policy that
resulted in an increase of one additional year of schooling for a group would
not necessarily bring about an increase in that group’s literacy scores by
0.15 of a standard deviation. Several other factors affect people’s scores
on these tests, such as parents’ education, background, reading habits and
leisure activities. |
| 3. |
For example, the analysis cannot determine whether there is an
effect associated with ethnicity for a particular state or province if there
are only a few minority youth represented in the sample. In the case of the
Canadian provinces, there were no data available on ethnicity. |