Search

The article presents the author's comments on the spectral and cross spectral analysis of post-accord interest rates conducted by economists V. Kerry Smith and Richard G. Marcis in their article "A Time Series Analysis of Post-Accord Interest Rates," originally published in the June 1972 issue of "The Journal of Finance." The author states that the major weakness in Smith and Marcis' analysis is their failure to compute confidence intervals. Contrary to the claim made by Smith and Marcis, the analysis does not reveal any systematic lead-lag relationship between long-term and short-term interest rates.

The article discusses a study which ascertains the validity of Kuder preference record scores in predicting forester advancement and turnover in the U.S. The study examines 87 men, all graduates of professional forestry schools and employed with the U.S. Forest Service, who were engaged in allied specialties in December 1962. Results show that the standard deviation and means for the Kuder scales are found to have a .01 statistical significance. It adds that those employees represent an unemployment period between 13 and 15 years.

The article discusses a study which evaluates factors to consider on the effectiveness of performance appraisal program. The study included 31 specialists, and 31 managers in several manufacturing components of General Electric Co. Results show that the employees' attitude toward the merit-pay salary plan is significantly correlated with the constructive action taken toward the performance appraisal program. The study adds that education shows negative correlations with the criterion of the program.

The article discusses the important point in the logic of testing statistical hypotheses. The data used in the study is from an experiment involving a statistical test of significance where there are two group randomly chosen. According to the author, a random sample which is taken with respect to some particular quality, is almost certainly not random with respect to the many qualities represented in the individuals of the sample. Therefore, it is believed that the logic of the test of significance has lost its force, because some configuration of the extra-experimental qualities must always occur.