The distribution of the groups, delineated by their literacy and computer use profiles, varied by country (Chart 15). In Italy, Switzerland and the United States the largest group consisted of users with below average literacy skills and lowto- medium intensity computer use7. In Italy this group was particularly large, accounting for over 60% of all respondents. Conversely, in Bermuda, Canada and Norway, the largest group consisted of users with average or higher literacy skills and low-to-medium intensity computer use. For all countries except Italy, respondents with strong prose literacy skills and high intensity computer use represented the third largest group. The smallest group contained individuals with high computer use but below average literacy skills.
In order to study the effect of literacy skills and computer use on income it was important to control for gender, age and other variables. Results indicate that literacy skills and computer use were strongly associated with personal income. With the exception of Italy, respondents who were in the high group for either literacy skills or computer use (Group 2 or 3) had approximately twice the odds of being top quartile income earners compared to respondents who had below average literacy skills and low-tomedium intensity computer use (Group 1). Moreover, the odds of being a top income earner effectively doubled again for respondents with both average or higher literacy skills and high computer use (Group 4) compared with Groups 2 and 3 – again with the exception of Italy. In fact, in Canada, Bermuda and Switzerland, respondents with average or higher literacy skills and high computer use had from about five to more than six times the odds of being top income earners than respondents with below average literacy and low-to-medium intensity computer use (Table 12).
Chart 15. Distribution of literacy and computer use profiles, by country, 2003
Source:
Veenhof, Clermont and Sciadas, 2005
| Group 1 | Group 2 | Group 3 | Group 4 | |
|---|---|---|---|---|
| Bermuda | 1,00 | 2,38** | 2,06** | 5,68** |
| Canada | 1,00 | 2,63** | 2,52** | 5,18** |
| Italy | 1,00 | 1,27 | 1,69** | 1,80* |
| Norway | 1,00 | 1,79** | 1,95* | 3,85** |
| Switzerland | 1,00 | 2,25** | 3,27** | 6,30* |
| United States | 1,00 | 1,86** | 2,07** | 3,75** |
Note: Group 1 is used as the reference group. Odds are adjusted for gender, age, educational attainment, and labour force status. For definition of Groups 1-4, see Box 4.
* statistically significant at the 5% level.
* * statistically significant at the 1% level.
Source: Veenhof, Clermont and Sciadas, 2005
7 By definition, one would expect this group to be relatively large because we define low-to-medium intensity computer use as the lowest 75% of values obtained for the index score representing use of computers for task-oriented purposes.