Search Performance as a Function of Depth of Processing and Scrolling Rate

Abstract

Most investigations of task performance as a function of the scrolling rate have focused on display characteristics (e.g. character density and line spacing) as the major class of covariats. However, a more important class of variables pertain to the user's goal or task that caused him to envoke the scrolling function. Accordingly, the focus of the present study is the impact various scrolling rates have on performance as the complexity of the user's task is varied. The complexity of a search task was varied over two levels on a Depth of Processing dimension (Cralk & Lockhart 1972) for each of three line-by-line scrolling rate conditions as well as for a paging condition in which the contents of the screen was advanced 23 lines of the 24 line display. In the semantic search condition, (high task complexity) users searched text displayed on a CRT for a word that is a member of a pre-designated category such as an animal name. This task requires that the user process each word in the display on a semantic level to determine whether or not it is a member of the target category. Also, it is argued that the semantic search task, which is less complex than a reading task, is representative of tasks for which users typically search a display. In the feature search condition, (low task complexity) users searched a CRT display for an exact match to a pre-designated target word. For this task the user need only process each word in the display to the visual feature level to determine whether or not it is the target word. Assuming that more cognitive processing time is required for the semantic search task than the feature search task, it was expected that the optimal scrolling rate for the visual search task would be faster than that obtained for the semantic search task. The preliminary results confirmed this expectation and also indicated that the paging conditon produced better performance than any of the line-by-line scrolling conditions.