[[alternative]]Does NTU Add More Value to Her Students Than Other Colleges?--Estimating the Performance of College Graduates in Taiwan

[[abstract]]本文運用「七十二年與七十三年專上畢業生調查」與「七十三年與七十四年專 上畢業生調查」資料，分析臺灣各大學院校初出校門的畢業生，在薪資決定、就業與進修 抉擇上是否有顯著差異。我們考慮個別樣本在就業、進修與其他狀況間的選擇，建立 multinomial logit模型，並據以修正薪資方程式中可能存在的樣本選擇偏誤。估計結果 顯示，大學院校對薪資、就業與進修三者的影響不一。利用Wald test檢定大學院校別虛 擬變數的聯合解釋能力，我們發現校際間的差異主要表現在就業與進修比例上，而薪資並 無顯著差異。此外，以臺大作為觀察參考組的實證結果顯示，在就業與薪資方面，臺大畢 業生的表現未必優於他校。 We use 1983-84 Taiwanese college graduate survey data to analyze the possible differences in college graduates' performance. Applying self- selection adjustment to our multinomia1 logit model, we estimate and test the impact of college dummy variables on graduates' wages, employment rates, and the percentage pursuing advanced study. A standard Wald test shows that there are significant across-college differences in the employment rate and the percentage pursuing advanced study, but there is no difference in graduates' salary. We find that the performance of NTU graduates in salary and employment is not significandy superior to those from other Taiwan colleges.[[notice]]補正完畢[[journaltype]]國