Future Performance Trend Indicators: a Current Value Approach to Human Resources Accounting. Report IV. An Examination and Evaluation of the Statistical Model.
Abstract
This report uses predictive relationships between the Survey of Organizations and measures of organizational effectiveness established in earlier reports (Pecorella and Bowers, 1976a, 1976b, 1977) (AD's A037 733, A033 608, and A036 907), to develop the formal equations, including parameter values, to be used in the value attribution phase. Preliminary to this a series of issues arising from the theoretical and statistical bases of the current value approach are examined. These include: characteristics of variables to be included in the model, assumptions required for prediction of future performance, extension of the univariate methodology to the multivariate case, and the consequences of applying the methodology to this particular data file, for example, standardization of variables and elimination of outliers. A summary of previous and current analyses is also provided. A particular issue is how well the assumptions underlying the multivariate model are met in this particular data file. Analyses show there is no reason to believe the assumptions of multivariate linearity and normality are not met. Other analyses show that one cannot significantly reduce the number of predictor variables in the model. It is concluded that the presented model is appropriate, and the data set is ready to begin the value attribution steps.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1977
- Accession Number
- ADA045068
Entities
People
- Alan S. Davenport
- David G. Bowers
- Jean B. Lapointe
Organizations
- University of Michigan