Some Problems In The Physical Education Evaluations And Their Addressing Measures

In the past decade, many evaluation methods are widely applied in vary fields of Physical Education (P.E. for abbreviation), and such a trend gives a great put forward on the development of P.E. science. However, for the demerits of the evaluation methods or the limitations of the conditions that such methods should be satisfied, there arise lots of problems in the applying the above mentioned evaluation methods. So it is meaningful and necessary to find measures for solving these problems, for this purpose, this paper devote to the studying of those problems by focusing on the basic elements of P.E. evaluation. The construction of the index, the distribution of the weight and some mistakes should be escaped in the applying of P.E. evaluation are mainly discussed. The main results include the following five aspects.1. The research on the basic elements of evaluation.2. Some principles should be obeyed in the construction of index; make sure not dropping any principle components or major factors in the principle analysis and factor analysis. Treatments for the indexes and principle components in special cases.3. Some problems in the evaluating of the weights and their settlement measures. Analysis Hierarchy Process (AHP for abbreviation) and principle component methods are two widely used methods on this aspect. The application of AHP is based on the analysis and the evaluation of the judgment matrix, thus the hinge of the application of this method is its uniformity testing. Some modification must be made for the matrixes which do not satisfy the uniformity for eliminating the errors in the evaluation. The advantage effects should not be ignored in AHP, by integrating the Delphi method and principle component methods; one can avoid the advantage effects.4. The key of the fuzzy comprehensive evaluation is how to construct the fuzzy matrix and fuzzy comprehensive result vector, so the evaluation indexes should not too many and each index should be perfect independent. The fuzzy operation model should be proper, or else, the corresponding results can't be compel lent.5. The uniformity coefficient and x~2 test can be used to test the errors of evaluation. Since the errors of the evaluation reflect the quality of the evaluation, it follows that the uniformity coefficient and the x~2 test can be used to check the validity of the evaluation. If the errors are too large, one should try to find the source of the errors and measures for reducing errors by decomposing the evaluation process.