Complementary Log-Log with Random Effect Model Using Malaysian Graduate Employability Data
Md Azman Shahadan1, Rahmattullah Khan Abdul Wahab Khan2, Aimillia Mohd Ramli3
1Md Azman Shahadan, Faculty of Human Development, Sultan Idris Education University Malaysia.
2Rahmattullah Khan Abdul Wahab Khan, Faculty of Human Development, Sultan Idris Education University Malaysia.
3Aimillia Mohd Ramli, Kulliyyah of Islamic Revealed Knowledge and Human Science, International Islamic University, Malaysia.
Manuscript received on 25 March 2019 | Revised Manuscript received on 04 April 2019 | Manuscript Published on 27 April 2019 | PP: 219-225 | Volume-7 Issue-6S2 April 2019 | Retrieval Number: F10330476S219/2019©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: The objective of this research is to investigate the relationship between hazard ratio or survival function of graduate employability and 8 explanatory variables. The 8 explanatory variables are as follows: gender, CGPA, geographic region, English language proficiency, area of study, curriculum satisfaction, carrier guidance services satisfaction and monthly family income. In order to study a survival function of graduate employability, we have developed a sequence of binary numbers (employed (1) or unemployed (0)) at a particular time of being employed among first degree graduate students (N=2228). The data is based on a survival study data, which traces the survival function of graduate students within 12 months of window opportunity. We have used the complementary log-log model in various forms. For the simple complementary log-log model, the results showed that the hazard risk with baseline hazard ratio relate to graduates’ gender, CGPA, satisfaction with career guidance, the geographic region from which they came from, their English performance, their area of study and their family income. As for subject-specific (random effects), the hazard ratio also relates to the abovementioned variables.
Keywords: Modeling Discrete time to event, Employability Data, Survival Function, Subject-Specific Proportional Hazards Model, Random Effect Complementary Log-Log Regression Model.
Scope of the Article: Data Visualization