The object-line effect: is it attributable to intercontext differences or the structural properties of contexts and task demands?

Briefly presented target lines are reported more accurately when part of an object-like context than alone (Williams & Weisstein, 1978). Efforts to explain this effect have tended to focus on structural properties that contexts must possess in order to be effective in facilitating target-line discrimination (e.g., Weisstein, Williams, & Harris, 1982). Enns and Prinzmetal (1984) take issue with this approach. They propose that contexts differ more from one another than do target lines presented alone, and it is this fact, and not the general structural properties of contexts, that underlies their effectiveness. Two experiments examined intercontext differences and structural factors as determinants of context effectiveness. Subjects were found to use intercontext differences to narrow the range of possible alternatives, but when this factor was controlled, structural properties determined context effectiveness. The structural interpretation of context effectiveness was found not be as straightforward as generally assumed, however. A third experiment showed context effectiveness to be affected by the character of the discrimination task employed.