Classification of Malaysian students' tendency in choosing TVET after secondary school using analytic hierarchy process and decision tree model: a case study in northern part of Malaysia

Technical and Vocational Education and Training (TVET) programme is a channel to produce skilled workers because it provides theoretical knowledge and practical skills for students. However, the current enrolment in TVET programme among secondary school leavers is still low. Therefore, this study ai...

Full description

Saved in:
Bibliographic Details
Main Author: Hong, Chia Ming
Format: Thesis
Language:eng
eng
Published: 2021
Subjects:
Online Access:https://etd.uum.edu.my/10173/1/s826073_01.pdf
https://etd.uum.edu.my/10173/2/s826073_02.pdf
Tags: Add Tag
Description
Summary:Technical and Vocational Education and Training (TVET) programme is a channel to produce skilled workers because it provides theoretical knowledge and practical skills for students. However, the current enrolment in TVET programme among secondary school leavers is still low. Therefore, this study aims to uncover the factors that affect students’ tendency in enrolling TVET programme. The perception of students, public, instructors, employers, parents, besides facility, cost and policy have been discovered as the factors in this study. In the first phase, Analytic Hierarchy Process (AHP) is used to determine the level of importance of each factor. Based on the outcome of the first phase, various Decision Tree models are developed to classify the students’ tendency in enrolling TVET programme. The respondents in this study are students from the TVET programme and secondary schools. The result of AHP shows that the factor of parents is the most important factor, followed by instructors, employers, students, cost, facility, policy and public. Then, four types of Decision Tree namely ID3, CART, C4.5 and CHAID are generated to classify the students’ tendency based on the four most important factors (parents, instructors, employers and students). The ID3 tree in 5-fold Cross Validation is selected as the best model due to its low misclassification rate (0.1355), high accuracy rate (0.8645), high precision rate (0.9143), high recall rate (0.8828), and high F-score (0.8983) with tree depth of 7 and maximum branches of 3. Hence, in the future, this model can be used to classify the students’ tendency enrolling in TVET programme. It can also assist the government to implement strategic plans such as organizing campaigns or providing on-the-job training to students in the TVET programme. Therefore, skilled workers that can adapt to new technology and innovation could be produced.