女性教育與地位指標關聯之探索

Abstract

本研究旨在瞭解女性教育與地位指標的關聯。本研究運用女性國會地位占有率、女性管理地位占有率（即女性行政與管理人力占有率）、女性國民所得占有率、女性識字率、三級教育量為指標，進行女性地位的國家群集分類；探討各指標之間的關聯，以及分析教育女性地位的因果模式。研究資料取自UNDP（1995/96/97）。本研究提出五個虛無假設，考驗結果顯示：假設一、以五個指標進行分類，高度女性地位國家、中上女性地位國家、中度女性地位國家及低度女性地位國家各有33、34、23、11個國家，分類的準確度在95%以上；假設二、三與假設四、經過回歸分析發現，女性識字率、女性國民所得占有率、三級教育量對女性在國會占有率、女性管理地位占有、女性專業地位占有均是重要的正向影響因素；假設五、以教育潛在變項對女性地位潛在變項進行因果模式探索，卡方值並未達到顯著，同時AGFI、GFI、RMR的適配指標都符合檢定標準，且教育潛在變項對女性地位潛在變項影響值為.77，達到.01的顯著水準。也就是說，教育程度愈高，女性地位也較高。 The main purpose of this study is to analyze the correlation between the status of female and educational indicators. The study uses the seats in parliament held by women（%）, female administrators and managers（%）, and a combination o the ratio of first-, secondary-, and third-level education（%）for analysis. The raw data are collected from the UNDP（1995/96/97）. There are five null hypotheses in the study to test. The results are as follows：Using Cluster Analysis, five indicators are used to categorize 101 countries into four groups：the status of higher-level females（33 countries）, the status of middle-upper-level females（34 countries）, the status of middle-level females（23 countries）, the status of lower-level. females（11 countries）. In order to test the consistency of the clustering, Discriminant Analysis is used to reclassify the clustered countries. Above 95% of countries are correctly classified by the five indicators （hypothesis 1）. Hypotheses 2, 3 and 4 are tested by multi-regression analysis which uses the seats in parliament held by women（%）, female administrators and managers（%）, female professional and technical workers（%） as dependent variables, respectively, and use the ra6tio of first-secondary-and third-level education（%）as independent variables in the model for analysis. All the independent variables are significant in the three models. To understand the representative of the female indicators causality model, the research employed the LISREL for this study. It tested female indicators with 101 countries included in the model, and found that there are two latent variables in this model, such as educational and status of female latent variables. There resul6s showed that the X2 value is not significant, that is , the model is fitted better, and other indices, GFI, AGFI, and RMR, are also better. Also, the status of female latent variables is influenced by the educational latent variables, that is，it is significant （p＜.01）, too.