Premium
There is no perfect evaluator: An investigation based on prospect theory
Author(s) -
Lee Min Chul,
Park Jaehyun
Publication year - 2018
Publication title -
human factors and ergonomics in manufacturing and service industries
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.408
H-Index - 39
eISSN - 1520-6564
pISSN - 1090-8471
DOI - 10.1002/hfm.20748
Subject(s) - usability , reliability (semiconductor) , likert scale , computer science , field (mathematics) , scale (ratio) , validity , psychology , human–computer interaction , applied psychology , psychometrics , mathematics , clinical psychology , developmental psychology , power (physics) , physics , quantum mechanics , pure mathematics
Psychophysical measurement has been widely used in various scientific and engineering fields. In the field of human–computer interaction (HCI), many prominent constructs, such as usability, affective satisfaction, user value, and user experience, have been assessed by psychophysical measurement, including Likert, Borg, and 0–100 scales. In spite of considerable empirical evidence that psychophysical measurement is reliable and valid, these measures also have fundamental limitations, such as distortion at both extremes. This study introduced prospect theory, which compensates for decision‐making behavior in uncertain situations, and applied the theory to the transformation of assessed data. In a typical usability assessment study with target selection tasks, this study revealed that psychophysically assessed data might improve reliability and validity. In addition, this study showed that people tend to regard around 75 as judgment criteria on a 0–100 scale. The results of this study can be used and applied to improve the reliability and validity of subjective measurements. This study suggested a method to improve psychological measurements based on prospect theory. The method is expected to be applied to practical engineering experiments, which has been widely conducted in the field of human factors and HCI.