International Journal of Advance Study and Research Work (2581-5997)/ Volume 6/Issue 4/Oct-Nov-Dec -2023

Mathematical Models Predicting Performance in

Licensure Examination of Engineering Graduates
Charo M. Consolacion1 and Dr. Marc D.G. Garvida2
1; 2
Graduate School, Centro Escolar University, Philippines
Email Id: [email protected] and [email protected]
ORCID Id: 0009-0009-8816-6948

DOI: 10.5281/zenodo.10439120
Abstract
The success of a tertiary educational institution’s graduates on professional licensing exams is one of the most reliable
indicators of the educational quality that the institution provides. For many years, the Philippines have produced many
engineering graduates; however, in recent years, roughly 50% successfully passed the board exams required for licensure as
engineers. The challenge that many higher educational institutions face is making all their engineering programs’ passing
rates improve and surpass the national passing rate. This study intended to develop a mathematical model capable of
forecasting the performance of BSCE and BSME graduates from a private higher educational institution in the Philippines
in the licensure examination administered by the Professional Regulation Commission (PRC). The researcher implemented a
descriptive correlational research design. The graduates’ academic performance and enrollment in a review center were
assessed through descriptive statistics, which includes means, standard deviations, frequency, and percentages. Academic
performance includes grades in courses tested by PRC in BSCE and BSME licensure examinations, which include
mathematics and engineering courses. It also involves grades in purposive communication and integration courses. Pearson
moment correlations, and ETA, were applied in determining the correlation of variables. Moreover, the strongest predictors
of the Licensure Exam Areas 1, 2, and 3 were determined using stepwise regression, and mathematical models were
developed using multiple linear regression. Mean Absolute Percentage Error (MAPE) was used to assess the models'
predictive power. The result of the research is intended to serve as a helpful guide and basis for the school in preparing
students for the licensure examination and a credible input for the institution’s curriculum enhancement.

Keywords: mathematical model, predictor variables, mathematics, licensure examination, integration courses
Introduction
Licensure examinations serve as a crucial mechanism for assessing and assuring the quality of professionals entering the
workforce, both locally and internationally, as recognized in Section 2 of Republic Act 8981. Licensure examinations are
conducted by the Professional Regulation Commission (PRC), and the importance of these examinations is inferred from the
legal framework that establishes the PRC and its responsibilities. The PRC is mandated to regulate and supervise the practice of
various professions, ensuring that individuals who enter certain professions meet the necessary qualifications and standards.
Honest and reliable licensing examinations are fundamental in establishing the competence of professionals. Consequently,
universities and colleges bear a dual responsibility: to provide students with high-quality education and to meet society's demand
for well-qualified professionals.
Habito (2019) wrote in the Philippine Daily Inquirer dated May 31, 2019, that in 2018, Philippines graduated approximately
50,000 engineering students, with civil engineering students being the largest cohort. However, only around 50% of them
successfully passed the board exams to obtain their engineering licenses. Also, based on the licensure exam results obtained
from the official PRC website (https://www.prc.gov.ph/articles), it can be concluded that over the last six years, the national pass
rate for the civil engineering board examination has averaged approximately 40.61%; and mechanical engineering, 54.33%. This
pattern of poor results is a problem that requires urgent action from academic institutions providing the programs.

This brings to light the paramount importance of licensure examination success. It is imperative for Higher Education
Institutions (HEIs) to develop accurate predictors of licensure performance. Such predictive tools help identify candidates at risk

1
Corresponding Author: Charo M. Consolacion, Graduate School, Centro Escolar University, Manila, Philippines.(Email id:
[email protected])
2
Cite as: Charo M. Consolacion and Dr. Marc D.G. Garvida (2023). Mathematical Models Predicting Performance in Licensure Examination
of Engineering Graduates. International Journal of Advance Study and Research Work, 6(4), 10–19.

10
© 2023, IJASRW, All right reserve
https://www.ijasrw.com
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License
 International Journal of Advance Study and Research Work (2581-5997)/ Volume 6/Issue 4/Oct-Nov-Dec -2023

of failing the licensure exam and enable future examinees to customize their preparation strategies. HEIs can enhance curricula
for courses identified as predictors and provide additional support to students. Educators can offer guidance, tutoring, and
support to improve student performance. The use of predictors in licensure examinations represents a proactive, data-driven
approach to improve candidate performance and the quality of education provided by institutions. It allows for early
intervention, focused preparation, and better alignment with licensure board expectations, benefiting both candidates and the
professions they aim to enter.

Several studies have already explored the prediction of graduates' performance in licensure examinations, particularly for
Nursing (Bautista et al., 2019), Agriculture (Nicolas, 2020), Elementary and Secondary Teacher Education (Amanonce &
Maramag, 2020), BS Math (Velasco, 2022), Mechanical Engineering (Dotong et al., 2019), Civil Engineering (Vallejo, 2020),
Forestry, Geology, Electrical Engineering (Callena et al., 2019), Electronics Engineering (Terano, 2018), and Agricultural
Engineering (Origines & Alo, 2019).

The key to effective prediction lies in identifying and utilizing variables that are closely associated with the outcomes.
Numerous studies have uncovered various variables that influence success in licensure examinations. Notably, academic
performance in college has been identified as the primary indicator of success in licensure examinations for graduates of Teacher
Education (Bansiong, 2019; Amanonce & Maramag; Polinar et al., 2020), Agriculture (Nicolas, 2020), Agricultural Engineering
(Origines & Alo, 2019), Electronics Engineering (Terano, 2018), and Mechanical Engineering (Dotong et al., 2019). However, a
study by Somosot et al. (2022) reported contrasting results for graduates of Teacher Education.

Another measure of students’ achievement used by earlier academics to explore predictive variables was the terminal assessment
administered before a student graduate from college in the form of a comprehensive exam for teacher education (Gabasa &
Raqueno, 2021), pre-LET (Amanonce & Maramag, 2020; Fiscal & Roman, 2022), mock board for agricultural engineering
(Origines & Alo, 2019), pre-board examination for electronics engineering (Terano, 2018), and institutional terminal
competency assessment for nursing (Oducado et al., 2019).

Among these measures, Integration Courses are three review courses designed to assist students in their preparation for the
licensure examination, specifically in the field of Engineering. The successful completion of all Integration Courses is marked
by a final exam that evaluates students' development and mastery of competencies. No study or literature was found about
Integration Courses’ ability to forecast achievement in the licensing examination.

Entrance examination in college was also found as a significant explanatory variable of the achievement of grads in the licensing
examination for Teacher Education (Somosot et al., 2022), of Agriculture (Nicolas, 2020), of Civil Engineering and Electrical
Engineering (Callena et al., 2019). Specific areas identified were intelligent quotient, IQ (Bansiong, 2019) and verbal reasoning
or reading comprehension capacity of students (Maaliw, 2021 The significance of verbal reasoning and reading comprehension
capacity as predictors has raised questions about the predictive ability of Purposive Communication.

Furthermore, some studies discovered that review center admission/attendance (Nicolas, Polinar et al., 2020), age, sex, civil
status, number of takes, period between graduation and the board examination, first or second course (Nicolas, 2020), General
Scholastic Aptitude (Bansiong, 2019, Bansiong & Balagtey, 2020), curriculum (Cristobal, 2020) and high school average grade
(Callena et al., 2019; Nicolas, 2020) are significant predictors as well, although some of these variables are weak determinants.

Given the concerning trend of poor outcomes in engineering licensure examinations, the contrasting claims about the predictive
ability of academic performance, the knowledge gap regarding the predictive capacity of Integration Courses and Purposive
Communication Course for success on the licensure examination, and the literature gap concerning the predictive ability of
enrollment in review centers as a predictor of BSCE and BSME graduates’ success in the licensure exam, the researcher chose to
conduct a study focused on developing prediction models for the performance of BSCE and BSME graduates in the licensure
examination. The research sought answers to the following specific questions.
1. Describe B.S. Civil Engineering (BSCE) and B.S. Mechanical Engineering (BSME) graduates of a private educational
institution in the Philippines in terms of the performance in the licensure examination, academic performance in
college, and enrollment in a review center.
2. Determine the strongest predictor for each of the Licensure Examination Area 1, 2 and 3.
3. Determine the relationship between licensure examination results of the BSCE and BSME graduates and the following
factors:
3.1 Academic performance:
3.1.1 The strongest predictor of the Licensure Examination Area 1.
3.1.2 The strongest predictor of the Licensure Examination Area 2.

11
© 2023, IJASRW, All right reserve
https://www.ijasrw.com
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License
 International Journal of Advance Study and Research Work (2581-5997)/ Volume 6/Issue 4/Oct-Nov-Dec -2023

3.1.3 The strongest predictor of the Licensure Examination Area 3.

3.1.4 Final grade in purposive communication.
3.1.5 Average grade in three integration courses.
3.2 Enrollment in review center.
4. Design mathematical models predicting performance in Licensure Examination of BSCE and BSME graduates.
5. Determine the prediction accuracy of the mathematical models in forecasting the performance of BSCE and BSME
graduates in the licensure examination.

Literature Review
Mathematical modeling, as described by Banerjee (2022), plays an increasingly vital role across various scientific disciplines.
According to Yohannes (2019), mathematical modeling transforms real-world problems into manageable mathematical
formulations, utilizing theoretical and arithmetic analysis to provide insights, answers, and assistance in developing applications.
Yohannes (2019) emphasizes that mathematical modeling often offers precision and a strategic approach to problem-solving,
facilitating a systematic understanding of the system under simulation. It stands out as the most effective means of
representation, applicable to analyzing all aspects of the physical or biological universe, including natural phenomena,
technological advancements, and human interventions. This effectiveness is achieved through characterizing variables using
mathematical expressions, typically through correlation (Britannica Encyclopedia, 2018).

Senapati et al. (2023) explain that a predictive mathematical model analyzes past observations to forecast future events, with
past data serving as a crucial element in anticipating likely outcomes. Shankar et al. (2021) assert that an ideal model should
incorporate the maximum number of relevant variables affecting study outcomes, aiming for the closest possible proximity to
actual results.

There have been several attempts to develop models for predicting performance in licensure examinations (Tarun et al., 2014).
Forecasting students' academic performance is crucial for anticipating potential challenges and enabling educational institutions
to implement preemptive measures to avert failure (Cabanela, 2023). Delos Angeles's study (2020) aimed to identify variables
predicting the performance in the licensing examination of teacher education graduates. The focus was on evaluating the quality
of process variables, such as instruction and curriculum, highlighting the crucial role of predictors in forecasting performance.

One of the significant impacts of mathematical modeling is its ability to identify variables influencing success in licensure
examinations and predict performance. Rustia et al.'s (2018) model, for example, has the capability to pinpoint students at risk of
failing the Licensure Examination for Teachers. It is recommended to give priority to these students during their mock board
review, thereby increasing their likelihood of passing the examination. This not only aids individual students but also plays a
role in elevating the institution's overall pass rate in the Licensure Examination for Teachers, offering potential benefits during
the accreditation process.

Moreover, the prediction models by Gabasa & Raqueño (2021) identified comprehensive exams as strong predictors of
graduates' licensing success, leading to the suggestion that the college conducts more intensive practice with comprehensive
examinations. The model established by Terano (2018) emphasized the significance of academic achievements and pre-board
examination results as indicators for the success of BSECE graduates in the electronics engineering board exam. The
interpretation of academic performance is contingent on the indicators employed to measure it, as academic achievement
encompasses a broad spectrum of educational outcomes (Steinmayr et al., 2020). This underscores the recommendation that the
institution undertake intensive evaluations of academic courses and regularly improve its curriculum.

Methods
A descriptive correlational research design was employed in this study. The descriptive correlational method was found
appropriate for the study because there is an exploration of the engineering graduates’ academic performance. In addition, the
problem calls for identifying the characteristics, frequencies, and trends necessary in developing a mathematical model for
predicting achievement of the BSCE and BSME graduates in the licensure examination.

The data were gathered from a private higher educational institution in the Philippines: 204 graduates of Bachelor of Science in
Mechanical Engineering (BSME), and 206 graduates of Bachelor of Science in Civil Engineering (BSCE). The graduates took
the professional licensure examination from September 2017 to April 2023.

The researcher divided the obtained data into 80% training set and 20% testing set. According to Gholamy et al. (2018) to avoid
overfitting, it is important to separate data into training and testing sets. He also mentioned that empirical studies show the best
result at 70/30 and 80/20 relation of training and testing sets.

12
© 2023, IJASRW, All right reserve
https://www.ijasrw.com
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License
 International Journal of Advance Study and Research Work (2581-5997)/ Volume 6/Issue 4/Oct-Nov-Dec -2023

Cleaning of data was done prior to data analysis. All six assumptions of Multiple Linear Regression and Stepwise Regression
were satisfied. The assumptions include type of data, linearity, residual independence, normality, absence of multicollinearity
and homoscedasticity.
In analyzing the graduate’s performance in the board examination and academic performance in college, frequency and
percentage was utilized as well as mean and standard deviation. To define the graduates’ average performance for each variable,
the mean was applied and to determine homogeneity/heterogeneity, the standard deviation was used. Frequency and percentage
were employed to know the graduates’ performance in the licensure examination, and frequency of graduates’ enrollment in
review center.
The Pearson moment correlation was employed for interval-to-interval correlation. Additionally, as in the research of Lumagsao
& Dellosa (2019), for nominal-to-interval, the association was determined using the eta value. Moreover, to forecast the
licensing examination achievement of the graduates, multiple linear regressions was employed. Stepwise regression, on the other
hand, was utilized to identify the strongest predictor for each licensure examination areas. And to evaluate the mathematical
model developed through multiple linear regression, this study employed Mean Absolute Percentage Error (MAPE).

Independent Variables Dependent Variables

SP1
Academic Performance in College Courses under Performance in Licensure Exam (Area 1)
Area 1

SP2
Academic Performance in College Courses under Performance in Licensure Exam (Area 2)
Area 2

SP3
Academic Performance in College Courses under Performance in Licensure Exam (Area 3)
Area 3

Academic Performance in Integration Courses

Over-all Performance
Academic Performance in Purposive
in the
Communication
Licensure Examination

Enrollment in Review Center

Developed Mathematical Model in
Predicting Performance in the Licensure
Examination of Engineering Graduates

Figure 1. Research Paradigm

SP1 – The strongest predictor of performance in the Licensure Examination Area 1,

SP2 – The strongest predictor of performance in the Licensure Examination Area 2,
SP3 – The strongest predictor of performance in the Licensure Examination Area 3

Figure 1 encapsulates the framework of the present research. It depicts the procedure involving the variables of this research.
The first frame indicates the independent variables, which were assumed to exert a significant impact on the dependent
variables. The next part of the framework pertains to the predicted variables, which consist of the graduates’ performance in the
three distinct licensure examination areas and the overall performance in the form of a general average rating. Specifically, the
first Civil Engineering Licensure Examination area is about Mathematics, Surveying, Principles of Transportation and Highway
Engineering, and Construction Management and Methods. The second area is centered around Hydraulics and Geotechnical
Engineering, while the third area deals with Principles of Structural Analysis and Design. In the mechanical engineering
licensure exam, tests are categorized also into three main areas. The first area comprises Industrial and Power Plant Engineering,
area 2 encompasses Mathematics, Engineering Economics, and Basic Engineering Services, and the third area includes Machine
Design Materials and Shop Practice courses. The last frame indicates the result of the research, which is the prediction model
each for BSCE and BSME programs that predicts how engineering graduates will do on the licensure to get licensed.

13
© 2023, IJASRW, All right reserve
https://www.ijasrw.com
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License
 International Journal of Advance Study and Research Work (2581-5997)/ Volume 6/Issue 4/Oct-Nov-Dec -2023

The dependent and independent variables were classified according to their type of data. The table below presents the research
variables, their description and how they are coded in SPSS.

Table 1. The Research Variables and their Description

Factors Description Values Descriptors Numerical

Values
Performance actual rating in the licensure 0 to 100 Excellent 94-100
In Licensure exam Very Good 88-93.99
Examination Good 82-87.99
Fair 76-81.99
Passing 70-75.99
Failed 69.99-below
Academic final grade in mathematics and 1.00 to 3.00 Excellent 1.13-1.00
Performance engineering courses tested in the Superior 1.38-1.14
licensure exam Meritorious 1.63-1.39
final grade in purposive Very Good 1.88-1.64
communication course Good 2.13-1.89
average grade in three integration Very Satisfactory 2.38-2.14
courses Satisfactory 2.63-2.39
Fair 2.88-2.64
Passing 3.00-2.89
Enrollment enrollment in any review center Yes, No No 0
In Review Center after college Yes 1

Results and Discussions

Based on the interpretation and analysis of the data gathered, the following summary of findings is enumerated:
1. In the CELE, the majority of BSCE graduates fell into the fair category, with the best performance observed in Area
2, having the highest mean, and the least performance in Area 3, marked by the lowest mean scores. On the other hand, among
the areas in the Mechanical Engineering Licensure Examination, Area 2 recorded the highest mean score, falling within the fair
category, followed by Area 1 and then Area 3. The majority of BSME graduates were classified in the passing category.

Table 2 shows that, on average, BSCE and BSME graduates who passed their respective Licensure Examination Areas scored
higher compared to those who did not pass. Furthermore, the overall average rating reveals a distinct contrast between those
who passed and those who failed, with passing BSCE and BSME graduates attaining notably higher average scores. Across both
CELE and MELE, a consistent trend emerges, indicating that graduates who passed the examination scored significantly higher
than their peers who failed in all areas and the general average rating.

Table 2. Summary of Passed and Failed Performance in the Licensure Examination

Licensure Examination Mean (Passed) Mean (Failed)

Civil Engineering Licensure Exam Area 1 78.02 69.42
(CELE) Area 2 83.77 67.74
Area 3 74.92 51.84
Gen. Ave. 78.66 62.57
Mechanical Engineering Licensure Examination Area 1 78.67 60.63
(MELE) Area 2 78.15 64.96
Area 3 74.42 57.88
Gen. Ave. 77.21 61.32

When assessing academic performance in college mathematics courses, a diverse range was observed. The majority of BSCE
graduates’ performance was distributed across the passing to very good range, with the greatest number falling into the very
satisfactory category. Moreover, the majority of BSME graduates’ performance spanned the passing to good range, with the
greatest number in the ‘satisfactory’ category. This also indicates that, in both BSCE and BSME programs, a low percentage of
graduates achieved an excellent grade in mathematics courses. There was a consistent trend of best performance in Advanced

14
© 2023, IJASRW, All right reserve
https://www.ijasrw.com
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License
 International Journal of Advance Study and Research Work (2581-5997)/ Volume 6/Issue 4/Oct-Nov-Dec -2023

Engineering Mathematics (including Numerical Methods for BSCE) and least performance in Calculus 2 compared to other
mathematics courses for both programs.
In various engineering courses, the majority of BSCE graduates obtained a good rating, while most BSME graduates achieved a
very satisfactory rating. BSCE graduates performed best in Hydrology, followed by Geology, and then Construction Materials
and Testing. On the other hand, BSME graduates excelled in Manufacturing & Industrial Processes with Plant Visits, followed
by ME Law, Ethics, Codes, and Standards, and then Machine Shop Theory. Nevertheless, a majority of both BSCE and BSME
graduates demonstrated good performance in Integration Courses.
Regarding Purposive Communication, the academic performance of BSCE graduates is generally rated as very good, whereas
BSME graduates’ performance is generally rated as good. Additionally, most of these graduates enrolled in review centers,
recognizing the importance of these centers in enhancing their preparation for the licensing examination.
2. As exhibited in Table 3, for the BSCE program, academic performance in Calculus 1, Geology, and Principles of
Reinforced/Prestressed Concrete were found to be the strongest predictors of the Civil Engineering Licensure Examination
Areas 1, 2, and 3, respectively, based on stepwise regression. Moreover, the strongest predictors of the Mechanical Engineering
Licensure Examination in Areas 1, 2, and 3 are the academic performance in Heat Transfer, Calculus-Based Physics, and
Materials Science and Engineering for ME, respectively, according to stepwise regression.

Notably, these courses displayed r-squared values ranging from .025 to .421 with significant F-changes. Although this finding
suggests that the performance of graduates in the corresponding college courses explains less than 50% of the variability in the
performance in the Licensure Examination Areas, these results underscore the significant role of these courses in explaining
variance in the licensing examination areas.

Table 3. Strongest Predictors (SP) of the Licensure Examination Areas

Licensure Examination (DV) College Courses (IV) R r2 F-change Sig. F Change

Civil Area 1 Calculus 1 .545 .297 4.636 <.05
Engineering Area 2 Geology .159 .025 4.133 <.05
(CELE)
Area 3 Principles of Reinforced/ .314 .099 4.041 <.05
Prestressed Concrete
Mechanical Area 1 Heat Transfer .445 .198 6.528 <.05
Engineering Area 2 Calculus-Based Physics .649 .421 8.498 <.05
(MELE)
Area 3 Mat Sci & Eng’g for ME .355 .126 4.502 <.05

3. All three strongest predictors of the areas of CELE, along with academic performance in, integration courses,
purposive communication, and enrollment in review centers, were found to have statistically significant positive correlations
with the graduates’ performance in the Civil Engineering Licensure Examination. Furthermore, the strongest predictors of the
MELE in Areas 1, 2, and 3, in addition to purposive communication, and integration courses, were found to have statistically
significant positive correlations with the Mechanical Engineering Licensure Examination.

Table 4. Relationship between Licensure Examination Performance and Academic Performance

Licensure Examination College Courses R VI p-value Remarks

Civil Calculus 1 (SP1) .574 Substantial <.05 S
Engineering Geology (SP2) .275 Low <.05 S
(CELE)
Princ of Reinforced/Prestressed .249 Low <.05 S
Concrete (SP3)
Purposive Communication .151 Very Low <.05 S
Integration Course .160 Very Low <.05 S
Mechanical Heat Transfer (SP1) .471 Substantial <.05 S
Engineering Calculus-Based Physics (SP2) .540 Substantial <.05 S
(MELE)
Mat Sci & Eng’g for ME (SP3) .470 Substantial <.05 S
Purposive Communication .449 Substantial <.05 S
Integration Course .573 Substantial <.05 S

15
© 2023, IJASRW, All right reserve
https://www.ijasrw.com
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License
 International Journal of Advance Study and Research Work (2581-5997)/ Volume 6/Issue 4/Oct-Nov-Dec -2023

Correlation was based on the contrasting rating scales used in college courses and licensure examinations. There is high positive
correlation when grade in college courses gets closer to 1.00 which indicates better performance, as the licensure examination
rating gets closer to 100 signifying better performance also.

Table 5 reveals an association between enrolling in a review center before taking the board exam and performance in the CELE
as specified by the obtained eta value (η = 0.290, p < .001 based on ANOVA. The finding, however, is different with the BS
Mechanical Engineering graduates, the association between enrolling in a review center and licensure examination performance
in terms of rating was found to be very weak and not significant.

Table 5. Relationship between Licensure Examination Performance and Enrollment in Review Center

Program η *p-value Remarks

Enrollment in Review Center BSCE .290 <.05 S
BSME .066 .405 NS
* based on ANOVA

4. For BSCE graduates: Calculus 1, Principle of Reinforced/Prestressed Concrete, and Enrollment in a Review Center
were found to be significant predictors of the performance of graduates in the Civil Engineering Licensure Examination (CELE).
While the predictive ability of these three factors is relatively small, as evidenced by the r-squared value of 0.420, signifying that
these factors explain approximately 42% of the variability in BSCE graduates’ performance in the Licensure Examination, it
remains statistically significant. The developed mathematical model for the BSCE program of a private higher educational
institution in the Philippines suggests that, for a grade of x1 in Calculus 1, a grade of x2 in the Principle of
Reinforced/Prestressed Concrete, and a binary response of x3 in Enrollment in a Review Center, the predicted performance in
CELE can be computed as: ŷ= 86.946 – 7.634x1 – 1.824x2 – 11.371x3.

Academic
Academic Performance in Enrollment in Predicted
Constant
+ Performance + Principles of + =
 𝛽 = 86.946 in Reinforced
Review
Center
Performance in
CELE
Calculus 1 /Prestressed (𝛽 = 11.371) 𝑟 2 = .420
(𝛽 = −7.634) Concrete
(𝛽 = −1.824)

Figure 2. Mathematical Model for BSCE Program

For BSME graduates: Integration Courses, Calculus-Based Physics, and Purposive Communications were found to be
significant predictors of BSME graduates’ performance in the Mechanical Engineering Licensure Examination (MELE). While
the predictive ability of these three factors is relatively small, as evidenced by the r-squared value of 0.477, signifying that these
factors explain approximately 47.7% of the variability in BSME graduates’ performance in the Licensure Examination, it
remains statistically significant. The developed mathematical model for the BSME program of a private higher educational
institution in the Philippines suggests that, for a grade of x1 in Integration Courses, a grade of x2 in Calculus-Based Physics, and
a grade of x3 in Purposive Communication, the predicted performance in MELE can be computed as: ŷ= 106.079 -
5.360x1 - 5.772x2 - 2.965x3

Academic
Academic
Performance Predicted
Constant
+ in
+ Performance in + Academic Performance in =
Calculus- Performance
𝛽 = 106.079 Integration Purposive Communication
Based Physics in MELE
(𝛽 = −2.965)
Courses
(𝛽 = −5.772) 𝑟 2 = .477
(𝛽 = −5.360)

Figure 3. Mathematical Model for BSME Program

5. Table 6 presents MAPE results for the BSCE and BSME mathematical models developed using multiple linear
regression applied to both training and testing sets. The low MAPE values of 6.47, 9.1, 6.37 and 13.2 of the training and testing

16
© 2023, IJASRW, All right reserve
https://www.ijasrw.com
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License
 International Journal of Advance Study and Research Work (2581-5997)/ Volume 6/Issue 4/Oct-Nov-Dec -2023

sets suggest that the mathematical model’s predictions are close to the actual values, indicating a good to very good forecasting
performance. The higher MAPE values for the testing set, matched to the training set, implies that the model’s performance
might have slightly deteriorated when applied to unseen data.

Furthermore, in both cases, the models perform better in predicting students who passed compared to those who failed. MAPE
result of 23.8 and 26.6 indicates fairly good mathematical model’s predictions.

Table 6. MAPE for the Mathematical Models

Program Set MAPE Result Outcome MAPE Result

BSCE Training Set 6.47
Testing Set 9.1 Passed 7.4
Failed 23.8
BSME Training Set 6.37
Testing Set 13.2 Passed 6.4
Failed 26.6

Conclusions
In light of study’s finding, the subsequent conclusions were formulated:
1. In both Civil Engineering Licensure Examination (CELE) and Mechanical Engineering Licensure Examination
(MELE), a consistent pattern emerges graduates who successfully passed the examination achieved significantly higher
scores than those who did not, across all specific areas and in the general average rating. This analysis indicates that
performance in specific areas is correlated with the overall success of graduates in these licensure examinations.
Graduates need to perform well in all areas to pass the exam.
2. The strongest predictors for Civil Engineering Licensure Examination Area 1, 2, and 3 based on stepwise regression are
Calculus 1, Geology, and Principles of Reinforced/Prestressed Concrete, respectively. Similarly, for Mechanical
Engineering Licensure Examination Area 1, 2, and 3, the strongest predictors based on stepwise regression are Heat
Transfer, Calculus-Based Physics, and Materials Science and Engineering.
3. The strongest predictors for their respective licensure examination areas were significantly and mostly substantially
correlated with the general average rating of the licensure examination.
4. Mathematical model predicting performance of BSCE graduates from a private higher educational institution in the
Philippines in the Civil Engineering Licensure Examination:
ŷ= 86.946 – 7.634x1 – 1.824x2 – 11.371x3
(Where: x1- grade in Calculus 1, x2– grade in Principle of Reinforced/Prestressed Concrete, and x3- a binary outcome in
Enrollment in a Review Center.)
Mathematical model predicting performance of BSME graduates from a private higher educational institution in the
Philippines in the Mechanical Engineering Licensure Examination:
ŷ= 106.079 - 5.360x1 - 5.772x2 - 2.965x3
(Where: x1– grade in Integration Courses, x2– grade in Calculus-Based Physics, and x3– grade in Purposive
Communication).
5. The developed mathematical models can forecast passing performance of BSCE and BSME graduates in the licensure
examination with high prediction accuracy. However, the mathematical models failed to provide high prediction
accuracy in forecasting failing performance in the licensure examination of the BSCE and BSME graduates.

Conflict of Interest
Authors have no conflict of interest to declare.

Author Contribution Statement:

C.C. contributed to the study conception and design, data collection, processing, analysis and interpretation of results, and
manuscript writing. M.G. contributed to the critical revision or editing of the article, gave final approval for the version to
be published, and provided supervision.

17
© 2023, IJASRW, All right reserve
https://www.ijasrw.com
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License
 International Journal of Advance Study and Research Work (2581-5997)/ Volume 6/Issue 4/Oct-Nov-Dec -2023

References
[1]. Amanonce, J.T., & Maramag, A. M. (2020). Licensure examination performance and academic achievement of teacher
education graduates. IJERE, 9(3), 510 – 516. http://doi.org/10.11591/ijere.v9i3.20614
[2]. Banerjee, S. (2022). Mathematical Modeling: Models, Analysis and Applications (2nd ed.). CRC Press.
[3]. Bansiong, A. J., &Balagtey, J. L. M. (2020). Predicting success in teacher education: Revisiting the influence of high school
GPA, admission, and standardized test scores on academic and licensure performance. Journal of Research, Policy & Practice
of Teachers, and Teacher Education, 10(2), 1-17. https://doi.org/10.37134/jrpptte.vol10.2.1.2020
[4]. Bansiong, A.J. (2019). Entry-to-exit academic variables as predictors of Board Licensure Examination for Professional Teachers
(BLEPT) Ratings. The Normal Lights, Journal on Teacher Education, 13(2). https://doi.org/10.56278/tnl.v13i2.1392
[5]. Bautista, J. R., Ducanes, G. & David, C. (2019). Quality of nursing schools in the Philippines: Trends and evidence from the
2010–2016 Nurse Licensure Examination results. Nursing Outlook, 67, 259-269.
http://dx.doi.org/10.1016/j.outlook.2018.12.012
[6]. Britannica, T. Editors of Encyclopedia (2018, July 27). mathematical model. Encyclopedia Britannica.
https://www.britannica.com/science/mathematical-model
[7]. Cabanela, L.C. (2023). Data-Driven Career Placement Examination System with Prediction Model in Forecasting Licensure
Performance Using Regression Techniques. IJMTHK, 1(2), 107-121. https://doi.org.10.56868/ijmt.v1i2.37
[8]. Callena, E., Gabales, B., Tutor, R., Villanueva, S., Gonzales, C., de Vera, A., Caberte, S., Nillas, V. B., Acerbo, J., & Pantaleon,
A. (2019). Predictors of passing probability in the licensure examination for selected programs in the University of
Southeastern Philippines. Southeastern Philippines Journal of Research and Development, 24(1), 1-16.
https://doi.org/10.53899/spjrd.v24i1.12
[9]. Cristobal, J.C. (2020). Licensure examination for teachers (let): a quality control towards k-12 curriculum efficacy. Journal of
critical reviews. https://doi.org/10.31838/jcr.07.11.26
[10]. Delos Angeles, M. G. (2019). Predictors of Performance in Licensure Examination for Teachers. Universal Journal of
Educational Research 8(3), 835-843. DOI: 10.13189/ujer.2020.080314
[11]. Dotong, C.I., Hicaro, A., &Laguador, J.M. (2019). Licensure Examination Performance of Mechanical Engineering Graduates
and Its Relationship with Academic Performance. Asia Pacific Journal of Academic Research in Social Sciences, 4, 7-14. ISSN
2545-904X. https://research.lpubatangas.edu.ph/wp-content/uploads/2019/12/APJARSS-2019.002.pdf
[12]. Fiscal, R. & Roman, A. (2022). Pre-licensure examination as predictor of licensure examination for teachers result. IJERE, 11,
136-142. http://dx.doi.org/10.11591/ijere.v11i1.22133
[13]. Gabasa, M.G., & Raqueno, A.R. (2021). Predicting performance of graduates in the licensure examination through path analysis
toward curriculum improvement. International Journal of Advance Study and Research Work, 4(1), 11-19.
https://paper.researchbib.com/view/paper/270200
[14]. Gholamy, A., Kreinovich, V., & Kosheleva, Olga. (2018). Why 70/30 or 80/20 Relation Between Training and Testing Sets: A
Pedagogical Explanation. University of Texas at El Paso, ScholarWorks@UTEP.
[15]. Habito, C. F. (2019, May 31). In search of top engineers. Inquirer.net. https://opinion.inquirer.net/121678/in-search-of-top-
engineers
[16]. Lumagsao, R., & Dellosa, R. (2019). Application of Eta Correlation in Determining the Effect of Demographic Profile on the
Assessment of Quality of Higher Educational Institution of Cordillera Administrative Region. IJRTE, 8(1).
https://www.ijrte.org/wp-content/uploads/papers/v8i1/A1179058119.pdf
[17]. Maaliw III, R. (2021). Early prediction of Electronics Engineering Licensure Examination Performance using Random Forest.
IEEE, 0041-0047. https://doi.org/10.1109/AIIoT52608.2021.9454213
[18]. Nicolas, H. (2020). Student determinants in the licensure examination for agriculturists of a state college in the Philippines.
International Journal of Educational Sciences. 28. http://dx.doi.org/10.31901/24566322.2020/28.1-3.1117
[19]. Oducado, R.M. &Cendaña, D. &Delariarte, R. G. (2019). “Institutional competency assessment and other factors influencing
the nurse licensure examination.” (“Institutional Competency Assessment and Other Factors Influencing the ...”) International
Journal of Scientific & Technology Research, 8, 268-279. http://dx.doi.org/10.6084/m9.figshare.11778324.v1
[20]. Origines Jr., D. & Alo, M. (2019). Correlation and linear regression analysis (CALRA): A predictive decision support system
for LEAE. International Journal of Advanced Research in Computer and Communication Engineering, 8. 1-14.
http://dx.doi.org/10.17148/IJARCCE.2019.8401
[21]. Polinar, E., Delima, A. & Vilchez, R. (2020). Students’ performance in Board Examination Using Naive Bayes and C4.5
Algorithm. International Journal of Advanced Trends in Computer Science and Engineering, 9, 753-758.
http://dx.doi.org/10.30534/ijatcse/2020/107912020
[22]. Republic Act No. 544. Civil Engineering Law -An Act to Regulate the Practice of Civil Engineering in the Philippines.
https://www.prc.gov.ph/sites/default/files/Civil%20Engineering%20-%20Board%20Law_0.PDF
[23]. Republic Act No. 8981. Professional Regulation Commission (PRC) Modernization Act of 2000. (Phil)
https://legacy.senate.gov.ph/lisdata/55574592!.pdf
[24]. Republic of the Philippines Professional Regulation Commission. News. https://www.prc.gov.ph/articles
[25]. Rustia, R. & Cruz, M. & Burac, M. &Palaoag, T. (2018). Predicting student’s board examination performance using
classification algorithms. ICSCA 2018: Proceedings of the 2018 7th International Conference on Software and Computer
Applications, 233-237. (“ICSCA - ACM Other conferences”) https://doi.org/10.1145/3185089.3185101

18
© 2023, IJASRW, All right reserve
https://www.ijasrw.com
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License
 International Journal of Advance Study and Research Work (2581-5997)/ Volume 6/Issue 4/Oct-Nov-Dec -2023

[26]. Senapati, A, Maji, S., & Mondul, A. (2023). Data Representation and Performance in a Prediction Model. Wiley Online Library,
Mathematics and Computer Science, 1. https://doi.org/10.1002/9781119879831.ch13
[27]. Shankar, S., Mohakuda, S., Kumar, A., Nazneen, P., Yadav, A., Chatterjee, K., & Chatterjee, K. (2021). Systematic review of
predictive mathematical models of COVID-19 epidemic. Medical Journal Armed Forces India, 77(2), S385-392.
https://doi.org/10.1016/j.mjafi.2021.05.005
[28]. Somosot, I. & Duran, J. & Rodriguez, B. (2022). Success under pressure: a probabilistic analysis of the predictors of the
licensure examination for teachers (LET) results. IJSRMS 8, 15-20.
https://www.isroset.org/journal/IJSRMS/full_paper_view.php?paper_id=2787
[29]. Steinmayr, R., Meißner, A., Weidinger, A.F., & Wirthwein, L. (2020). Academic Achievement. Oxford Bibliographies. DOI:
10.1093/OBO/9780199756810-0108
[30]. Tarun, I., Gerardo, B., & Tanguilig III, B. (2014). Generating Licensure Examination Performance Models Using PART and
JRip Classifiers: A Data Mining Application in Education. International Journal of Computer and Communication Engineering,
3(3), 203-207.
[31]. Terano, H.J.R. (2018). Regression model of the licensure examination performance of electronics engineering graduates in a
state college in the Philippines. Advances and Applications in Mathematical Sciences, 18(2), 197-204.
https://www.proquest.com/scholarly-journals/regression-model-licensure-examination/docview/2193969110/se-2
[32]. Vallejo, S.C. (2020). Analysis of the performance in the licensure examination for Bachelor of Science in civil engineering
graduates of a University in Isabela Province. https://www.semanticscholar.org/paper/Analysis-of-the-Performance-in-the-
Licensure-for-of Vallejo/fab27d7f3a51c8654e798513f26ab697a49c3c7b
[33]. Velasco, J. R.D. Jr. (2022). Mathematical models of performance in licensure examination for teachers of Nueva Vizcaya State
University. AIP Conference Proceedings 2472, 050025 (2022). https://doi.org/10.1063/5.0092741
[34]. Yohannes, H. M. G. (2019). Why is Mathematical Modelling so Important? Skyline University College.
https://www.skylineuniversity.ac.ae/knowledge-update/from-different-corners/why-is-mathematical-modelling-so-important

19
© 2023, IJASRW, All right reserve
https://www.ijasrw.com
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License