Does differential strategy use account for age-related deficits in working-memory performance?