Measuring the effectiveness of undergraduate programs by a nonparametric method

• Main Authors: Mat Kasim, Maznah, Kashim, Rosmaini, Rahim, Rahela, Hassan, Siti Safura
• Format: Conference or Workshop Item
• Language: English
• Published: 2015
• Subjects: QA75 Electronic computers. Computer science
• Online Access: http://repo.uum.edu.my/15611/1/PID113.pdf; http://repo.uum.edu.my/15611/; http://www.icoci.cms.net.my/proceedings/2015/TOC.html
• Institution: Universiti Utara Malaysia

Measuring the effectiveness of programs is not an easy task since the related variables may be affected by environmental factors which are often not under the control of the policy maker. Besides that, the use of output(s) alone without considering the input(s) to measure the effectiveness of programs may depict a wrong picture about the programs. If the ratio of output(s) to input(s) is defined as the basic measure of efficiency, effectiveness relates the input or output to the final objective(s) to be achieved, that is the outcome(s).The aim of this paper is to propose a model to estimate effectiveness of undergraduate programs by using a nonparametric method, known as Data Envelopment Analysis (DEA).In the tertiary educational setting, if the graduates are defined as the output, then it is argued that the employed graduates could be considered as the outcome, whereas the main inputs are the academic staff. As the academic staff are of different rank with different contribution, this feature should be reflected in any performance measure of academic institutions.The effectiveness model for undergraduate programs is proposed as the ratio of the number of employed graduates six months after graduation to numbers of academic staff at five different rankings.This model was used to analyze the effectiveness of 26 selected undergraduate programs in a university. The results show that only 10 were effective, and 16 were not effective for year 2009.Besides the results analyzed by the proposed model, the effectiveness measure based on a ratio of one outcome to one input, that is without considering the superiority of the inputs is also included for comparative analysis.As expected the two results are different.The result based on the proposed method is suggested to be used for further analysis and improvement action. Besides that, the proposed model may be used in different context with certain adjustment especially in the selection of input and outcome.