forecastingprinciples.com Reviews of Important Papers on Forecasting
Before 1985 Reviews
Review of:

Rodney Fralicx and Namburg S. Raju (1982), ďA comparison of five methods for combining multiple criteria into a single composite," Educational and Psychological Measurement, 42, 823-827.


Canonical correlations have been suggested in forecasting problems where a number of criteria are of interest and a number of predictors are available. The canonical weights determine the index that best predicts a criterion index. Theoretically, there is no reason to expect that a canonical index will be valid. Canonical correlations a methods that is often used when theory is lacking. This paper tests the validity of the canonical index for the formulation of a job performance index. The canonical index was compared with four alternative weighting schemes: management's subjective weights, equal weights, unit weights, and principal components factor weights. The alternatives yielded nearly identical weights for judging the overall performance of 117 bank tellers based on weight performance criteria (e.g. customer relations, attention to detail). In contrast, the canonical weights (which used the eight performance criteria as well as 13 predictor variables such as memory and arithmetic ability) had almost no correlation to the other methods. It is distressing that the canonical index had no relationship to methods with high face validity (management's subjective weights and equal weights). This study is consistent with my prior belief: it is good practice to define the criterion on the basis of its relevance to the particular decision and not merely to facilitate prediction.