Academic journal article Human Factors

Evidence from a Partial Report Task for Forgetting in Dynamic Spatial Memory

Academic journal article Human Factors

Evidence from a Partial Report Task for Forgetting in Dynamic Spatial Memory

Article excerpt


Little is understood about how people represent dynamic spatial knowledge (i.e., the knowledge of rapidly changing spatial situations needed in real-time tasks such as driving). For example, drivers must keep track of the location of nearby traffic vehicles in order to maneuver successfully and avoid hazards. Drivers' knowledge of the location of nearby vehicles is an example of what has been called situation awareness. Situation awareness can be defined as the activated knowledge a person has about a dynamic scene - knowledge that is used in performing a real-time task. While driving, operators must keep track of where they are on their route and the status of their vehicle (e.g., fuel status) in addition to nearby traffic. Research suggests that errors in maintaining situation awareness are the most frequent cause of errors in real-time tasks such as driving and flying (Hartel, Smith, & Prince, 1991; Shinar, 1993). Dynamic spatial knowledge is an aspect of situation awareness that is particularly important for tasks such as driving, flying, and air traffic control.

One reason so little is understood about dynamic spatial knowledge and the broader construct of situation awareness is that these constructs are difficult to measure. Researchers have commonly used recall measures to assess peoples' dynamic spatial knowledge and situation awareness. A common measurement technique is to have participants view a changing scene, then stop the scene at unpredictable times and have the participants recall object locations or other information about the scene. For example, Gugerty (1997) had participants recall the locations of traffic cars after three-dimensional moving driving scenes were stopped and the screen was blanked. Endsley (1995) developed the Situation Awareness Global Assessment Technique (SAGAT), in which pilots participate in a mission in a high-fidelity flight simulator. Whenever the scene is blanked, the pilots are asked a series of questions concerning information such as the location, altitude, and fuel status of their own and other aircraft.

There is a potential problem with this recall-based approach to measuring dynamic spatial knowledge and other aspects of situation awareness. In both of the previously mentioned experiments, participants recalled multiple pieces of information about a dynamic scene after the scene was removed from the simulator screen. As participants recall initial information, their knowledge about the scene may decline because of either decay or interference in memory. A number of experiments with simple, static stimuli have shown this decline, indicating that information that is recalled later in the recall sequence is recalled less accurately (Ayres, 1966; Endsley & Smith, 1996; Lawrence & LaBerge, 1956; Scheerer, 1972; Sperling, 1960). Thus this recall technique may underestimate participants' knowledge, at least for information from simple, static scenes. One purpose of the experiment presented here was to assess whether this underestimation occurs in a complex, dynamic spatial task.

In the rest of this article, I first review research showing that accuracy in recalling briefly presented scenes was affected by order of report, as well as one study using SAGAT that found that recall accuracy was not affected by order of report. Then I describe the partial report recall technique used in the current study and compare this technique with other recall-based measures of situation awareness. Finally, the results of the experiment are presented and discussed.

Order of Report

Sperling (1960) was one of the first to point out that asking participants to recall multiple pieces of information about a briefly presented stimulus may underestimate their knowledge of the stimulus information. When participants were asked to recall as many letters as they could from an entire visually presented array (whole report), they could recall only about 4 letters of a 12-letter array (53%). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.