Is Time-Sharing a General Capability?

The time-sharing ability of 18 students was measured under 8 separate dual-task conditions. Three distinct task characteristics were systematically varied across conditions in an effort to manipulate the nature of the specific time-sharing demands imposed. Each condition contained two of these characteristics in common with 3 of the remaining 7 conditions, one of the characteristics in common with 3 others, and none in common with the last condition. Time-sharing efficiency correlated across conditions that impose similar processing demands on the individual, but not across conditions imposing relatively dissimilar demands. We conclude that time-sharing performance under present conditions is determined by several poorly correlated, task-specific subcapacities rather than by a single general ability.