台灣技職院校商學院專科部之表現評估-資料包絡法

Performance Evaluation of Junior Colleges of Business in Taiwan: A Data Envelopment Analysis

陳文琦(Wen-Chi Chen)

指導教授 ： 傅祖壇

共同指導教授 ： 張靜貞

國立臺灣大學/生物資源暨農學院/農業經濟學研究所/碩士(2007年)

https://doi.org/10.6342/NTU.2007.10042

摘要

國內高等技職教育技職院校數量、科系大幅度的擴增，而其所增加招收的學生數量更極為可觀，再加上我國的人口出生人數下降，以及受國內高中職比例調整的影響，直接影響到專科學校的招生。因此學生未來在選擇就讀學校時將有更多的機會，如何慎選一所適合自己的學校，將是一門重要的課題。
然而，坊間似乎並未曾提供任何商業類專科學校的相關選校資訊，因此本研究針對26所台灣技職院校商學院的專科部為研究對象，探討畢業生就業表現及在校時多元能力培育能力之績效評估，並以三階段 DEA (Data Envelopment Analysis，資料包絡法)，分析考量環境變數後對於學校績效之影響，並採用資料包絡法(Data Envelopment Analysis)、交叉效率(cross efficiency，CDEA)、超級效率(Super efficiency，SDEA)之評估方法進行學校表現評估，並藉由不同的學校組織特性分為權屬別、學制別、專業導向，從這三類之特性做進一步的討論。分析結果如下：
一、在三階段DEA的調整分析，公私立學校在第一次工作的薪資調整前與調整後並無太大的差異。在目前薪資中，未調整前以公立學校表現較好，經過調整後私立學校在目前薪資部份表現較佳；在學制別的部份，未調整前的第一次工作的起薪及目前薪資，技職院校皆比科技院校高，調整之後的第一次工作薪資，二者差異不大，但就目前薪資來看，技職院校高於科技大學；商科與工科導向學校在第一次工作的起薪調整前後皆沒太大的差距，不過在未調整前的目前薪資部份，商科導向較高於工科導向的學校，而調整之後則差距不大。
二、從「認知、技能、情意」能力績效中，可得知三種DEA方法求出的結果相當一致，公立學校在三種DEA方法下的績效皆比私立學校好；科技大學在這三種能力績效皆高於技職院校；在專業導向上，商科導向在認知及情意能力上，績效比工科導向的學校好，而在技能部分則低於工科導向的學校。從三種方法上分析，CDEA所評估出的績效值為最低，SDEA為最低，其次為DEA。
三、 以在校時多元能力培育中來看，公立學校表現較好且標準差較小；科技大學表現比技術學院優秀；商科導向除了CDEA表現較優異外，DEA和SDEA的績效值皆比工科導向學校差。以三個模型來看，DEA在在校時多元能力培育的部份數值較高，其次為SDEA，數值較低為CDEA。
四、在薪資調整後之就業市場表現方面，公立學校除了CDEA較低於私立學校外，其他表現皆優於私立學校。技術學院在三種不同的績效評估方法中的表現皆高於科技大學。商科導向的學校在三種績效模型的衡量表現全都高過工科導向。
五、從薪資調整前與調整後可發現，薪資調整後，每個模型的效率值皆提昇為6%左右且變異數較小，表示經環境變數對薪資的調整後，每間學校的差距不甚明顯，使每間學校效率值皆提昇，說明了調整完後的各間學校就業表現績效較為一致。
六、以全部表現來看，公立學校在全部表現中的交叉效率低於私立學校，不過整體表現還是以公立學校表現優異；在學制別部份，科技大學在DEA和SDEA模型中皆高於技術學院，技術學院則在CDEA表現較突出；商科導向的學校在DEA和CDEA優於工科導向的學校，工科導向則在SDEA表現較好。
七、從Pearson Correlation可得知，在校時多元能力培育績效的三個模型的績效評估方法差異不大。在未調整薪資的就業表現上，DEA與CDEA為高度相關，在CDEA與SDEA的部分，結果不盡相同。在調整後薪資的就業表現上，DEA與CDEA相關係數只有4成多，顯示此二種模型的效率值有明顯的不同，SDEA和DEA也只有0.527的顯著相關，CDEA和SDEA更呈現出負相關的情況，不過並無顯著。以調整前、後薪資的就業表現上做比較，調整前後的DEA、SDEA就業表現績效值並無太大的改變情況，不過在CDEA的部份相關係數只有0.694，可見經過調整前後，在交叉效率的部分二者有顯著差異。在全部績效表現部分，DEA與CDEA、DEA與SDEA只有中度相關，CDEA和SDEA相關程度更小，顯示三種模式所算出的績效值，有明顯的不同。Spearman's Correlations大致與Pearson Correlationㄧ致，但Spearman's Correlations略高於Pearson Correlation可得知經過調整之後效率值有改變但名次排序部份差異並不明顯。

關鍵字

職場表現 ； 績效 ； 技職院校

並列摘要

The increasingly large numbers of local institutions of technology and programs of study, accompanied by the decreasing local birth rate, had directly resulted in a higher school admission rate for specialized studies. As a result, students will have more choices of school in the near future. It will also be more important for the students to carefully choose the most suitable schools for themselves.
Nevertheless, there is limited public resource of information regarding business college selection. This study is conducted based on the two-year and five-year specialized studies of 26 business colleges in Taiwan. The study will explore about fresh graduates’ job performances, as well as evaluate the performance resulted from school’s ability to develop students’ diversified skills. The analysis of effects on school performance after considering the environmental variables is done by the three staged DEA (Data Envelopment Analysis). The three evaluation methods of DEA (Data Envelopment Analysis), CDEA (cross efficiency) and SDEA (Super efficiency) are adopted for school performance evaluations. Moreover, the school organization characteristics can be classified into three different kinds of authority attributions, study systems and study subjects. Further discussions will be based on the three categories of characteristics. The results of analysis are as the following:
1.It’s been found that after the wage / salary adjustment, each model’s efficiency value all increased by about 6% with less variables. In another word, there were minor differences between schools after the wage / salary adjustment due to environmental variables. This resulted in increased efficiency value for each school, meaning the job performances of the students from different schools would be much more similar after the wage / salary adjustment.
2.By viewing from the school’s ability to develop their students’ diversified skills, it’s been found that public schools performed better with less standard deviation. Also, the performances of universities of technology were better than that of institutions of technology. More business-prone schools showed superior performance on CDEA, but these schools less well performed on DEA and SDEA comparing with engineering/science-prone schools. In general, DEA expressed higher values than SDEA on the school ability to develop diversified skills of students. CDEA showed the lowest values.
3.In regards to job performances after wage / salary adjustment, public schools performed better than private schools on all models except for CDEA. The CDEA values for institutions of technology were higher than that for universities of technology. The CDEA values for business-prone schools were also higher than that for engineering-prone schools.
4.On all performances, CDEA for public schools was lower than private schools. However, the overall performances were still better for public schools. In regards to the school systems, the DEA and SDEA models of universities of technology showed better results than institutions of technology. On the other hand, institutions of technology showed superior performance on CDEA. The DEA and CDEA performances for business-prone schools were better than engineering-prone schools, which performed superior on SDEA instead.
5.From Pearson Correlation, we found little differences between the three evaluation models on the performances resulted from schools’ ability to develop students’ diversified skills. Regarding job performances before wage / salary adjustment, DEA and CDEA were highly correlated. Whereas CDEA and SDEA were not correlated and were different. Regarding job performances after wage / salary adjustment, there were major differences between DEA and CDEA. There was also only 0.527 major positive correlation between SDEA and DEA. Whereas minor negative correlation was shown between CDEA and SDEA. On comparison between job performances before and after wage / salary adjustments, the performance values of DEA and SDEA didn’t change much. However, the correlation coefficient on the part of CDEA was only 0.694, which showed there were significant differences before and after the wage / salary adjustments. There were obvious differences between all the performances values obtained from the three models. Spearman’s Correlations is basically agreeing with Pearson Correlation. Nevertheless, with Spearman’s Correlations slightly higher than Pearson Correlation, it was found that the efficiency values had changed after (wage / salary) adjustment. The ranking of these values, however, did not differ significantly before and after the adjustment.

並列關鍵字

Job Market ； Performance ； Junior Colleges

參考文獻

張育瑄，2005。「職場表現、學生滿意度與商學院各系之效率分析」。碩士論文，

臺灣大學會計學研究所。

許舒翔、周春美、沈健華，2002。「技職院校教育行銷策略之探討」，『技術及職

林容萱，2003。「台灣地區科技大學效率性之分析：資料包絡分析法的應用」，『國

教育部統計處，2006。『中華民國大專院校概況統計』。台北：教育部統計處。