Premium
INFLUENCE OF MISSING ATTRIBUTES ON SELECTING A BEST MULTIATTRIBUTED ALTERNATIVE
Author(s) -
Barron F. Hutton
Publication year - 1987
Publication title -
decision sciences
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.238
H-Index - 108
eISSN - 1540-5915
pISSN - 0011-7315
DOI - 10.1111/j.1540-5915.1987.tb01516.x
Subject(s) - selection (genetic algorithm) , computer science , set (abstract data type) , sensitivity (control systems) , econometrics , statistics , risk analysis (engineering) , operations research , mathematics , machine learning , engineering , business , electronic engineering , programming language
Alternatives involving many factors arc difficult to evaluate because of multiple underlying competing objectives. If evaluation is based on an incomplete set of factors, and if the purpose of the evaluation is to select a single overall best alternative, inferior alternatives may be selected with surprising frequency and/or severe negative impact. At the same time, sensitivity analysis of evaluation scores based on statistical criteria can easily mask the impact and the frequency of selection of inferior alternatives. In this paper, criteria appropriate to reflect the decision impact are developed and both the frequency and impact of the selection of inferior alternatives are demonstrated empirically. Previous studies based on statistical criteria [1] [9] indicated minimal impact on overall evaluation and selection. This paper demonstrates that high statistical criterion values coexist with frequent and/or serious errors of selection.