Experimental Evaluation of a Debiasing Method for Analysis in Engineering Design
Year: 2019
Editor: Wartzack, Sandro; Schleich, Benjamin; Gon
Author: Nelius, Thomas; Matthiesen, Sven
Series: ICED
Institution: Karlsruhe Institute of Technology (KIT)
Section: Design Education
DOI number: https://doi.org/10.1017/dsi.2019.53
ISSN: 2220-4342
Abstract
During analysis in engineering design, systematic thinking errors - so-called cognitive biases - can lead to inaccurate understanding of the design problem. With a simplified version of the Analysis of Competing Hypotheses - ACH method and a simplified decision matrix, the confirmation bias in particular can be minimized. To evaluate this method, it was taught to experienced design engineers and mechanical engineering students. During the experimental evaluation the participants analysed a real technical problem. The procedures and results were compared with a previously conducted study with the same task. The design engineers have not changed their approaches and could not further improve their analysis success. The students profited considerably from the training. They have mentioned twice as much supporting evidence and six times as much contradicting evidence through the training indicating a more extensive analysis. As a result, the students showed significantly fewer signs of confirmation bias than without training. The findings suggest that debiasing strategies should be introduced early in engineering design education.
Keywords: Decision making, Design cognition, Human behaviour in design, Confirmation Bias, Debiasing