Statistical Analysis For The Development Of The 31 Regions' Human Resourses Of Higher Education In Our Courtry

By synthesizing principal components analysis and factor analysis, this article gives an order and an classification for the development of the 31 regions' human resources of higher education in our country in 1999. From the results of ordering and classifying, we find that the development of human resources of higher education is greatly influenced by the region distribution. As we can see, except three municipalities directly under the Central Government, which have better education bases and whose education is higher than other regions, in general, the development of education human resources is closely related to the region distribution. This can be shown as follows: the development in the east is better than that in the west and the north better than the south; The center provinces is better than those remote border provinces; coastal areas better than interior areas; the center regions have medium level of development. These results provide some scientific basis for programming and developing the education of each region.Now, we give the procedure of analysis. First, we make a principal components analysis and a factor analysis for the collected data. Then give orders respectively according to the result of above analyses. Next we test the two methods through Concordant Coefficients' Kendall-w test. We find that the two methods are highly concordant, which means that the method by synthesizing the above two methods are reasonable.So finally we synthesize the two methods. Comparing the results by synthesizing method to those by single principal components analysis or by single factor analysis, we can see that synthetical application of two assessing methods eliminates the one-sideness of single assessing method, and the results are more comprehensive and objective.In addition, this paper choices indexes according to the correlation coefficient matrix and the actual meaning of the indexes. The method is simpler and more practicable. We compare our results with those in reference [8] and find that they are almost the same.