| Abstract: | The continuous performance task is currently used to detect attention disorders in school-age children. However, its use with younger children is problematical, because indices measuring attention indirectly (omissions, commissions, reaction time) may be distorted by difficulties in following instructions: the ability to maintain and manipulate instructions in working memory, and to inhibit inappropriate actions increases with age. In the present research, we adapted such a task for children between 2 years 6months and 5 years 6 months, and recorded visual activity in order to measure directly visual attention and to relate it to performance. Even very young children spend most of the time looking at the screen, but they display poor performances and are unable to re-engage in the task after a gap (2 consecutive omissions). At a medium level (age 3 years 6 months to 4 years) endogenous control of attention increases, children come back to the task, off-task looks shorten and reach the short values observed in older children. After 4 years 6 months no more gaps are observed, the consistency of reaction times shows that a controlled strategy, including shifts from and to the screen, is now adopted. The discussion suggests that only at this upper level do indirect performance indices evaluate attention. |
