Identifying factors that affected student enrolment in Additional Mathematics for urban areas of Kuantan district

The skilful and qualified Science, Technology, Engineering and Mathematics (STEM) workforces are expected to be in high demand in the 21st digital economy. In Malaysia public education system, the principal key to STEM education is Additional Mathematics. However, the statistics depicted that the number of upper secondary students enrolled in Additional Mathematics have been severely delineated. Furthermore, the prerequisite to enrol in STEM and Technical and Vocational Education and Training (TVET) courses in tertiary education is achieving a minimum grade C for Additional Mathematics. Therefore, the principal objective of this article is to identify the significant factors that affected the upper secondary students enrolled in Additional Mathematics using Pearson’s chi-square test without Yates continuity correction and unadjusted odds ratio (OR). A validated questionnaire comprised nine potential factors that affected enrolment in additional mathematics was distributed to 389 Form Four students’ cohort 2020 from four selected urban government schools to pursue this objective. Based on the finding of this article, several initiatives might be taken by the policymakers, such as the teachers may enhancing and throughout the broad of STEM and TVET careers in the 21st digital economy era, while the students’ parents can participate in schools in building the communicating and coordinating mechanism. Consequently, the number of upper secondary enrolled in STEM education might be increased to pipeline the future STEM and TVET human capitals in sustaining and stabilising the national economy in the digital era.


Introduction
The first National Science and Technology Enrolment Policy, namely 60:40 Science: Arts policy, has been implemented in Malaysia's upper secondary education since 1970 [1,2]. In particular, this policy's principal objective is to provide the qualified and skilful Science, Technology, Engineering and Mathematics (STEM) workforces highly demanded in the future job market to ensure the sustainability and stability of the national economy in the 21st-century digital era [3,4]. For instance, the Academy of Sciences Malaysia outlined that the nation needs about one million new STEM human capitals in 2020. Despite that, the National Science and Technology Enrolment Policy remains unattainable [1][2][3][4][5][6].

Methodology
This section provides a brief overview of the methodology for data collection and data analysis. The insights collected from the selected sampling units via survey questionnaire were self-reported, and the collected data set was analysed using Microsoft Excel 2019 and SPSS Version 23. The methodology of this study has two major stages: the first stage (subsection 2.1) is the sampling and data collection, while the second stage (subsection 2.2) is the data analysis.

Sampling and data collection
Since the previous fully adapted Integrated Secondary School Curriculum (Kurikulum Bersepadu Sekolah Menengah, KBSM) for all subjects on the secondary education in 1989 is ineffective to provide the human capitals that meet the needs of workforces in the 21st century digital era [4]. Therefore, the Ministry of Education (MOE) has revised and implemented a new revised Standard Based Curriculum (Kurikulum Standard Sekolah Menengah, KSSM) in 2017 from Form One. The aim is to guide students towards the Science, Technology, Reading, Arts and Music (STREAM) education system to produce more STEM-literate students that meet the future job market needs [4][5]. In other words, the targeted population involved in this study is all the potential Form Four student's cohort 2020 from the urban areas of Kuantan District, the first cohort upper secondary students enrolled in Additional Mathematics using the KSSM syllabus. Due to time and cost constraints, this study has selected 389 Form Four students from four distinct urban national secondary schools from the areas using the cluster sampling technique. The geographical insights corresponding to the four selected schools are presented in figure  1 and table 1.   In addition, the sample size employed meets the required minimum sample size, , which was computed using Yamane's equation as given in equation (1).
where 3215 a N = represents the population size for all Form Four students' cohort 2020 from Kuantan District, while 0.05 ε = represents the sampling error. Moreover, the main reason for using the cluster sampling technique is to capture the heterogeneous effects due to the distinct subject packages offered among the secondary schools and teaching approaches among the teachers.
A new 5-point Likert scale survey questionnaire was developed for data collection by re-designing the existing survey questionnaire in the previous educational studies [18][19][20]. A pilot study was conducted to consolidate the reliability of the developed survey questionnaire. The results of analysis for the pilot study shows that the Cronbach's alpha measurement of all four intrinsic (self-efficacy: 0.929) and extrinsic (teacher influence: 0.854; parental influence: 0.923; peer influence: 0.885) motivational factors are greater than 0.7, which means the reliability of the new set of the survey questionnaire is accepted. Henceforth, a full-scale study was conducted using the validated survey questionnaire on the 389 selected Form Four students in urban areas of the Kuantan District between mid-February and mid-March before the implementation of the Malaysian Government Movement Control Order (MCO) 2020.

Data analysis
In this study, all the collected data from the 389 selected samples is complete and used. Suppose that ;    χ asymptotically approaches to the 2 χ distribution with degree of freedom 1. C − In general, the bc e is frequently assumed to be 5 ≥ in approaching the 2 χ distribution. The combination of sub-categories is true if the assumption does not satisfy [21]. Consequently, the p-value of 2 test χ is computed using equation (3).
is the upper incomplete gamma function and There is a significant association between 1 X and X j if and only if p-value < 0.05. Since the Pearson's chi-square without Yates continuity correction test can merely present the significant association between the two attributes, this study has further analysed the strengths of the association between 1 X and X j using OR, which is computed using equation (4).
where C E is the reference sub-category with the lowest odds of enrolment in Additional Mathematics, and the p -value(OR) ( ) function of the standard normal distribution. The corresponding 95% confidence interval (C.I.) for the OR is given as In general, X j does not significantly affect the odds of 1 X when the OR 1 = with the p-value (OR) 0.05 ≥ , and the corresponding 95% C.I. of the OR does overlap with the hypothesised value of OR 1. = In contrast, X j associated with significant higher odds of 1 X when the OR > 1 with the p-value (OR) < 0.05, and the corresponding 95% C.I. of the OR does not overlap with the hypothesised value of OR 1, = which the lower and upper limits for the 95% C.I. of the OR are customarily 1. > In other words, the Form Four student with the attribute ; c E c C ≠ has a significant higher tendency to enrol in Additional Mathematics compared to the student with the attribute C E in which both the attributes c E and C E belonged to the similar main factor investigated in this study.

Results of analysis and discussion
In this article, we rescaled the household incomes attribute by combining the proximity sub-categories to fit the household income classification 2019 of the Pahang state, namely B40 (< RM3,900), M40 (RM3,900 -RM7,599) and T20 (≥ RM7,600) [22]. The PT3 Mathematics achievement's sub-categories also have been rescaled into the lowest (Grades E and F), the average (Grades C and D) and the highest (Grades A and B) achievement based on the grades. The main reason for rescaling these attributes is to provide more meaningful and practical insight. In addition, the total number of items for the selfefficacy, teacher influence, parental influence and peer influence attribute on the designed survey questionnaire are 7, 5, 5 and 5 items, respectively. Hence, the average scores using the median are employed in summarising the scores corresponding to each attribute. The median is employed instead of the mean because these attributes are categorical data from the asymmetrical distribution. Likewise, we also rescale the intrinsic and extrinsic motivational factors according to the literature [23]. Furthermore, several attributes have also been combined, such as ethnicity and parental education level attributes due to the violation of the assumption for 5.  in the corresponding school, approximately 17.20% of the students with the lowest achievement in PT3 Mathematics can enrol in Additional Mathematics. Despite their lowest achievements, the students have the preference to enrol in Additional Mathematics. About the family's socioeconomic background, most of the participants are from a low-income family (B40), which numbers ≅ 171 ( ≅ 43.96%), and high parental education level (tertiary education), which numbers 210 (53.98%), respectively. Contrarily, the least of the participants are from high-income family (T20) and low parental education level (primary & others), with ≅ 88 ( ≅ 22.62%) and 14 (3.60%), respectively.
To acquire the statistical evidence about the association between the enrolment of the Form Four students in Additional Mathematics, 1 X and the potential factors, , X j Pearson's chi-square without Yates continuity correction test is employed. However, the ethnicity attribute was not considered a factor affecting the enrollment in Additional Mathematics to avoid the sensitivity issue. By using equations (2) and (3), 2 test χ and the corresponding p-values are given in table 3. In particular, the analysis results show a significant association between the upper secondary students enrolled in Additional Mathematics and the attributes such as parental educational level, PT3 Mathematics achievement, self-efficacy, teacher influence, parental influence, and peer influence. Then, further analysis was conducted to measure the significant strengths of the association between 1 X and X j using equation (4), which yielded the significant association without considering the effects of the confounder. In addition, the corresponding 95% C.I. of OR for each X j was also constructed. The results of the analysis of OR and the corresponding 95% C.I. of OR are provided in table 3. Comprehensively, the analysis results depicted that the significant affective factors for the upper secondary students in urban areas of Kuantan District enrolment in Additional Mathematics are the educational discipline, the PT3 Mathematics achievement, self-efficacy, teacher influence, and parental and peer influence. These factors yielded significant results for both Pearson's chi-square test without Yates continuity correction and OR.
In particular, the empirical results of this study show that the likelihood of a student from the STEM discipline education to enrol in Additional Mathematics is 44.317 times higher compared to those from the Humanities & Arts discipline education with the corresponding 95% C.I. of OR lies between 21.877 and 89.777. In Malaysia's public education system, Additional Mathematics is one of the elective subjects needed by the other science-related subjects for the STEM discipline education. At the same time, it was an optional subject for the Humanities & Arts discipline education. Therefore, the aforementioned empirical result is valid. The students from the STEM discipline education tend to enrol in additional mathematics than those from the Humanities & Arts discipline education.
Moreover, the likelihood of the student who has the highest achievement in the PT3 Mathematics subjects is 142.734 times higher than the student who has the lowest achievement to enrol in Additional Mathematics with the corresponding 95% C.I. of OR lies between 45.883 and 444.022. Meanwhile, the likelihood of a student with average achievement in PT3 Mathematics is 14.438 times higher than a student who has the lowest achievement to be enrolled in Additional Mathematics, with the corresponding 95% C.I. of OR lies between 5.018 and 41.542. This finding is analogous to the empirical results in previous studies [10,11].
From the psychology perspective, motivation is crucial to assist the students in preparing to face the challenges and compete [3]. Table 3 also indicated that the intrinsic motivational factor such as selfefficacy has a significant effect on additional mathematics enrollment. The likelihood of the student admitted the importance of self-efficacy to enrol in Additional Mathematics is 12.825 times higher than those denied, with the corresponding 95% C.I. of OR lies between 38.772 and 353.930. In contrast, students tolerated self-efficacy have a likelihood of 20.337 to enrol in Additional Mathematics compared to those denied, with the corresponding 95% C.I. of OR is [11.240, 36.797]. Moreover, the student's intrinsic motivation can also be affected by their parents. This is because their parents can help them make a correct decision, and the parents have better prediction skills about the child's learning skills and as the maximum support agents to their children [25]. Thus, this study found that the likelihood of the student who admitted and tolerated the parental influence are respectively 12.825 and 5.392 times higher     [4,8] have shown that external intrinsic motivational factors such as teacher and peer also significantly affected students' interest in enrolling in STEM education. Likewise, the analysis results depicted that the likelihood of a student admitting the peer influence to enrol in Additional Mathematics is 24.573 times higher than those denied. The corresponding 95% C.I. of OR lies between 12.719 and 47.473. Meanwhile, the 95% C.I. of OR for the student tolerated peer influence lies between 2.861 and 8.763. In other words, the likelihood of students accepting the peer influence to enrol in Additional Mathematics is 5.007 times higher than those who denied. In practice, the primary determinant of student engagement, success in subjects and interest in science are teaching and learning approaches and the quality of teaching [8]. Moreover, the teacher is also the primary source in enhancing the students' interest in STEM [4]. Similarly, the analysis results reveal that student who admitted that the teacher influence has contributed to a higher tendency to enrol in Additional Mathematics. The student's likelihood for those admitted the teacher influence is 15.786 times higher to enrol in Additional Mathematics than those denied, with the corresponding 95% C.I. of OR lies between 8.147 and 30.587. The likelihood of enrolling in Additional Mathematics is reduced to 5.392 times higher if a student tolerated the teacher's influence, compared to a student who denied, with the corresponding 95% C.I. of OR is [2.343, 6.800].

Conclusion
This article is educational research on investigating the significant factors that affected upper secondary students' enrollment in Additional Mathematics in the urban areas of Kuantan District. The study uses Pearson's chi-square test without Yates continuity correction unadjusted odds ratio. The analysis results reveal that educational discipline, PT3 Mathematics achievement, self-efficacy, teacher influence, parental influence and peer influence are the significant factors affecting the upper secondary students' enrolment in Additional Mathematics. In summary, this article provides new literature for educational research. The finding might be beneficial to the policymakers in instilling student's interest in STEM and pipelines the STEM human capitals demanded in the 21st digital economy era. For instance, teachers' skill can be enhanced throughout the broad career of STEM in the 21st digital market, and consciously review the role of their students in gender socialisation such as curriculum materials, expectation, educational tracking and peer relations. Meanwhile, parents also can be participated in school for building the communicating and coordinating mechanism. Henceforth, the number of upper secondary students enrolled in STEM education might be increased, and Additional Mathematics is the heart of science-related subjects.