Case‐based decision support for fault diagnosis: Do salient ratings discourage situation analysis?. Issue 1 (11th July 2018)

• Record Type: Journal Article
• Authors: Müller, Romy; Ullrich, Laura; Dessel, Ann‐Kathrin; Bengsch, Amelie; Heinze, Sebastian; Müller, Tobias; Oehm, Lukas
• Is Part Of: Human factors and ergonomics in manufacturing and service industries. Volume 29:Issue 1(2019)
• Page Start: 44
• Page End: 62
• Publication Date: 2018-07-11
• Subjects: automation bias -- case‐based reasoning -- decision support -- fault diagnosis -- rating format; Computer integrated manufacturing systems -- Periodicals; Human engineering -- Periodicals; 620.8205
• DOI: 10.1002/hfm.20757
• Languages: English
• ISSNs: 1090-8471

Abstract

Abstract: In the processing industry, frequent faults call for assistance in diagnosis, and case‐based reasoning (CBR) can provide solutions applied by other operators in the past. This study investigated whether salient case ratings promote an uncritical acceptance of solutions. In 2 experiments, subjects diagnosed faults with a simulated CBR system, and ratings were presented in graphical or verbal format. In most trials, the case with the highest rating provided the correct solution, while in catch‐trials, it did not. Graphical ratings were hypothesized to speed up solutions but discourage cross‐checking and lead to errors in catch‐trials. These hypotheses were not confirmed, even though Experiment 2 maximized the incentive of relying on case ratings. While graphical ratings led subjects to start with the most highly rated case, they did not impair situation analysis and accuracy. The results suggest that during fault diagnosis people are not easily misled into overtrusting a CBR system.