Estimating Predictors of the Philippine Licensure

Examination for Engineering

Adrian M. Tamayo, PhD

University of Mindanao
Philippines
dagntamayo@yahoo.com

Over the last three years, 2009-2011, the electrical engineering exam
was found to show declining trend, with faster rate of decline observed for the
year-end than the mid-year examination. Though declining, the electrical
engineering program could still make 4 passers out of 10 takers.
The academic performance, proxied by GPA and correlation, showed a
mixed result. The GPA could predict board outcome with a grade of 2.5 as
threshold. The correlation, observed with float-up grades, did not necessarily
predict board outcome, though, a simulated grade of 3.0 may land a positive
board results.
The examination for registered electrical engineers is composed of three
subject clusters, namely: mathematics, engineering sciences, and professional
subjects. It was determined that the engineering science had the strongest
influence on the board, followed by professional subjects, and the mathematics
cluster. It was also determined that the threshold rating for each subject is 65 in
order to pass the licensure examination

Keywords: Discriminant analysis, logistic regression, academic performance

INTRODUCTION

The global demand for electrical engineers leads to the initiatives of accreditation of
engineering education (Prados et al., 2005); accreditation and assessment ensures
quality of the engineering workforce (Patil & Codner, 2007). Given the borderless
nature of the demand for electrical engineers, countries are developing curriculum
that goes beyond the country-specific competencies (Lucena, 2008; Kellam, 2013)
that are oriented on engineering education research (Borrego & Bernhard, 2011)
that addresses engineering education issues (Case & Light, 2011) like teaching
strategy.

Perenet et al (2000) describe problem-based learning as an innovation in

engineering education, while others embedded teaching with service and
cooperative learning (Smith et al., 2005). The changes of the curriculum of the
engineering education gear are much oriented with inductive teaching methods

Electronic copy available at: https://ssrn.com/abstract=2560685

(Prince & Felder, 2006; 2007). These innovations and enhancements are in service
of making the graduates acquire and demonstrate the hard engineering skills
(Shuman, Besterfield-Sacre & McGourty, 2005) and the accompanying professional
skills (Shuman et al., 2004), in the same manner students are being exposed to
design content in the courses of each year (Wilczynski & Douglas, 1995).

Moreover, engineering school are applying for recognitions of accrediting bodies

(Patil & Codner, 2007)

In the University of Mindanao, as a manifest of commitment to the development of

the students taking engineering courses, a subject was added to the curriculum. This
subject is known to electrical engineering course as EE 500. This course is a form of
enhancement which will fill the gap in learning. An important context of filling the
gap is relatedness of theory with interest to environment (Dawes & Rasmussen,
2006). The design of the course is to develop an environment of the board
examination in a classroom setting, which would make engineering subjects
appealing to students, is through partnership (Sorby, 2006). This will allow the
students to prepare for the state-sponsored licensure examination. In this manner,
students can develop self-assessment framework relative to taking the examination,
thereby developing review strategies. In the same degree, the faculty teaching the
course will also gain vantage in understanding students’ readiness for the
examination.
The course was introduced in 2004 and had been operational since then, but
how was the course relating to the board outcome?

OBJECTIVES

This study was conducted to determine the effectiveness of the additional subject to
the curriculum, the EE 500 to the performance of the University of Mindanao in the
state-sponsored licensure examination. Likewise, the study aims to determine the
strength and weaknesses of the electrical engineering program of the institution
relative to the licensure examination.

METHOD

The study employed a causal design for research which is considered as appropriate
approach of the study (Fraenkl & Wallen, 1993). The method is appropriate because
it employs variables that would determine licensure outcome for electrical
engineering examination (Tamayo, Gevera & Aguilar, 2014; Tamayo, 2014). The
predictors used to explain board results were the GPA (grade point average) and the
correlation grade of the students which represented the academic readiness of the
engineering graduates of the University of Mindanao who would take the licensure

Electronic copy available at: https://ssrn.com/abstract=2560685

examination. In addition, the subject clusters were discriminated to determine
relative influence on the licensure examination.
The paper used combinations of statistical techniques that employed
ordinary least squares estimation to predict scores in the board exams, and the
maximum likelihood estimation to obtain the maximum scores that would
determine a likelihood of landing a successful board exam, such that logistic
regression and the linear regression were the ready beck-and-call techniques
(Tamayo, 2012f).
The empirical tools were appropriate because the interest variable assumed
a binary value ( pass  1; fail  0) (Tamayo & Quinto, 2014; Tamayo & Bernardo,
2014). Further assumed that the probability of passing increases as the grade
indicators (in the study the grade/ratings for GPA, correlation and subject clusters)
increase until a threshold for passing is reached equal to probability of 1. The log-
odds of the proportion of the dependent variable takes the form
 pass 
log     1GPA   2 Corr   where the error,  , is assumed to be
 fail 
normally distributed and independent from each other.
The discriminant analysis model takes the form
 ( x u k )  k ( x u k )
1  1
1
f k ( x)  e 2 for a density of X, given every class k in a
(2 ) p / 2 | k |
1 / 2

Gaussian distribution.

RESULTS

The results of the estimation and the corresponding discussions were

presented in this section.

Table 1 shows the performance of the electrical engineering program in the

licensure examinations from 2009 to 2011. Over the last three years, an observed
decline in the performance in the licensure examinations occurred. Although in
general, 4 out of 10 takers are passing the examination
The board performance for both year-end and midyear examinations of the
electrical engineering were observed to slowdown. It is, however, alarming that a
rapid ditch in the performance was observed in the year-end performance.
Table 1. Passers, flunkers, 2009-2011
Period Passers Flunkers % of passing
May 2009 10 15 40.00
September 2009 10 12 45.45
May 2010 5 13 27.78
September 2010 12 16 42.86
April 2011 4 15 21.05
September 2011 9 13 40.91

Electronic copy available at: https://ssrn.com/abstract=2560685

 Total/ average
50 84 36.34
passing %

The academic performance of the students of the program who eventually

took the board revealed a different picture. The GPA and the correlation grades
were used as proxy for the academic performance.
Table 2 portrayed an average academic ability of the students. The highest
GPA was 2.00 (90), the lowest was 3.87 (72), and the average grade was 2.94 (81)
with a standard deviation of 0.42.
The highest grade for the correlation course was found to be relatively lower
than the GPA. The highest was 2.30 (87), lowest was 3.60 (74) with an average
grade of 3.18 (79) with a standard deviation of 0.32.
This averaging of the values provides a glimpse of the academic performance
on the whole, but how about their relative performance?

Table 2. Min, max, average grades

Academic performance
Min Max Average Stdev
GPA 3.87 2.00 2.94 0.42
Correlation 3.60 2.30 3.18 0.32

Table 3 presents the most common GPA in pursuit of the understanding the
performance of the students over the last three years, 2009-2011. Two common
GPAs emerged: a GPA of 3.0 taking 16.4 % of the total number of students and 3.38
with 5.2%. Others were oscillating with lower than 2 percent ascription.

Table 3.Most common GPA

GPA % to total
3 16.4
3.38 5.2
Table 4 contains the most common correlation grade values. Three common
correlation grades emerged: a grade of with 24.63% of the total; 2.9 with 8.96% to
total and 3.0 with 6.72% of students relative to total.
Table 4. Most common Correlation
Correlation % to total
3.5 24.63
2.9 8.96
3.0 6.72
Quite interestingly, the correlation stood taller than the GPA in the academic
performance barometer. Will this relative advantage of the correlation transmitted
into becoming a better predictor of the board exams?
An empirical test was conducted to determine how significant the GPA and
the correlation course are to predict board outcome. The maximum likelihood

Electronic copy available at: https://ssrn.com/abstract=2560685

estimation was employed to determine causality of the two variables. Surprisingly,
GPA is a predictor; correlation is not. The float-up of the correlation seemed not able
to dent on the board outcome, the grades that were given as assessment of the
readiness of the students did not weave with the board outcome.

Table 5. Empirical model to determine board using GPA and Correlation

Variables B S.E. Wald Sig. Exp(B)
GPA -2.37 0.66 13.02 0.00 0.09
Correlation -0.42 0.70 0.37 0.54 0.66
Constant 8.32 3.21 6.73 0.01 4086.49
Cox and Snell R Nagelkerke R
2LL=119.24 Square=0.15 Square=0.22

A simulation was conducted to determine threshold GPA and correlation

grade that would render higher probability of passing.
It was determined that a GPA of 3.38 and a correlation grade of 2.9 would
render a 28.73% chances in the board for a predicted grade of 62.21. On the other
hand, a GPA of 2.0 and correlation of 2.3 renders very high chances of passing the
licensure examination at 93.165 with a predicted grade of 74.16 which is above the
rating set by the law (RA 7920, section 19 item 3). While a GPA of 2.5 and
correlation of 3.0 yielded a probability of 76% with a predicted passing grade of
70.17.
Thus for the electrical engineering program, the threshold GPA would be 2.5
and a correlation grade of 3.0.

Table 6. Determining probability of passing and predicted grade

Constant GPA Correlation Probability Predicted
Simulation 8.32 -2.37 -0.42 of passing grade
GPA=3;Corr=3.5 8.32 -7.11 -1.47 43.54 65.96
GPA=3.38;Corr=2.9 8.32 -8.01 -1.22 28.73 62.21
GPA=2.94;Corr=3.18 8.32 -6.97 -1.34 50.42 66.32
-
GPA=2.5; Corr=3.0 8.32 5.925 -1.26 75.67 70.17
GPA=2;Corr=2.3 8.32 -4.74 -0.97 93.16 74.26

With this, it is also important to juxtapose the performance of the program in

the subject clusters of the board. The requirement to pass the licensure
examination is to obtain grade not below 50% in any subject and a general rating of
70% to be admitted to the roll of the registered electrical engineers. Thus, a student
must not have below 50% rating in mathematics, engineering sciences and
professional subjects.
Mathematics cluster includes algebra, trigonometry, analytic geometry,
differential calculus, integral calculus, differential equations, complex numbers,
probability and statistics, advanced engineering mathematics including matrices,

Electronic copy available at: https://ssrn.com/abstract=2560685

power series, Fourier analysis, Laplace transformations. The law prescribed a 25%
weight for this subject cluster.
The other cluster is engineering sciences and allied subjects that includes
general chemistry, college physics, computer fundamentals and programming,
engineering materials, engineering mechanics, fluid mechanics, strength of
materials, thermodynamics, electrical engineering law, engineering economics,
engineering management, contracts and specifications, code of professional ethics,
Philippine electrical code 1 and 2. The cluster is given 30% weight.
The final cluster is electrical engineering professional subject that includes
electrical circuits, electronic theory and circuits energy conversion, power
transmission and distribution, instrumentation and measurement, circuit and line
protection, control systems, principles of communication, electrical machines,
electrical equipment, components and devices, electric systems, power plant
electronic power equipment, illumination, building wiring. This subject cluster
bears 45% weight of the total assessment of the examination.
As presented in table 7, the highest grade recorded for the subject clusters
were 89 for mathematics and professional subjects, a 2-notch lower fro engineering
sciences at 87. Looking at their average performance revealed that there were only
very slight differences in the average values of the clusters. Average grade for
mathematics was 62.59 with a standard deviation of 15.79, the highest among the
cluster. Engineering sciences registered an average grade of 61.78 with a 12.51
standard deviation, and the professional subject with 61.52 with a14.43 standard
deviation, the lowest among the clusters.

Table 7. Min, max, average

Subject clusters Min Max Average Stdev
Mathematics 21 89 62.59 15.76
Engineering sciences
25 87 61.78 12.51
and allied subjects
Electrical engineering
9 89 61.52 14.43
professional subjects

While table 7 revealed soothing figures, a greater number of the students

were also showing a good grasp of the board examinations. Table 8 presents a
uniformed, common rating for the subject cluster. A rating of 70 became a norm
grade of the students taking the examinations. Putting the average grade and the
common grade as adjacent barometers would allow for a grade of 65 as a threshold
grade (using a 5% margin of error).

Table 8. Most common grade of the clusters

Cluster Grade % to total
Mathematics 70 9
Engineering sciences 70 11.9

Electronic copy available at: https://ssrn.com/abstract=2560685

 Professional subjects 70 11.9

Finally, refining the clusters’ influence on the board outcome using the linear
disciminant analysis technique placed the engineering sciences as the dominant
cluster, followed by engineering sciences and the least is the cluster of mathematics.
The canonical coefficients were 0.52 for the engineering sciences, 0.50 for
professional subjects and 0.33 for the mathematics signifying degree of influence on
the board outcome. A higher value implies higher influence, better influence.

Table 9. Canonical discriminant function

Subject Coefficient
Mathematics 0.33
Engineering sciences and
allied subjects 0.52
Electrical engineering
professional subjects 0.50

There are two approaches to strategize preparation for the board

examinations given the foregoing information. Basically, a better GPA would render
higher chances of passing. This must be considered as the first among all readiness’
barometer. Second, employment of rating of 65 will tell how many are ready for the
board and how many of the students need to work hard in the review. Students with
rating of 65 must be subjected to more robust, serious and focus review.
As a general approach in the review, focus on professional subjects, then by
engineering sciences and the mathematics, in that order, would appropriate
leverage to the would-be takers in an efficient mode. However, given the
circumstances of the students and the professional evaluation of the teacher of the
review, schemes may be developed cognizant of the information provided.

CONCLUSIONS

Any higher learning institutions (HLIs) need to provide mechanisms and systems
that will ensure students to be ready to the industry that they will choose. The
students must be equipped by the institutions competencies, communications skills
on top of their ability to qualify to practice of their profession. On this, HLEs need to
develop systems that will allow students to pass the licensure examination.

Meantime, the ability of the students cannot and must not remain as theoretical
inertia. There is the expectations that when they are introduced to the real world of
work, they will be able to put to real actions the theories that they acquired all
through their learning years.

REFERENCES

Electronic copy available at: https://ssrn.com/abstract=2560685

Borrego, M., & Bernhard, J. (2011). The emergence of engineering education
research as an internationally connected field of inquiry. Journal of Engineering
Education, 100(1), 14-47.

Case, J. M., & Light, G. (2011). Emerging research methodologies in engineering

education research. Journal of Engineering Education, 100(1), 186-210.

Dawes L, Rasmussen G. Activity and engagement – keys in connecting engineering

with secondary school students. 2006. Technical Paper. Vol 13 No 1. Australasian
Journal of Engineering Education. Accessed November 7, 2013. Available:
http://www.engineersmedia.com.au/journals/aaee/pdf/AJEE_13_1_Dawes.pdf

Fraenkel, J. R., & Wallen, N. E. (1993). How to design and evaluate research in
education (Vol. 7). New York: McGraw-Hill.

KELLAM, N., WALTHER, J., COSTANTINO, T., & CRAMOND, B. (2013). integrating the
engineering Curriculum through the Synthesis and Design Studio. Advances in
Engineering Education, 3(3)

Lucena, J., Downey, G., Jesiek, B., & Elber, S. (2008). Competencies Beyond Countries:
The Re‐Organization of Engineering Education in the United States, Europe, and
Latin America. Journal of Engineering Education, 97(4), 433-447.

Patil, A., & Codner, G. (2007). Accreditation of engineering education: review,

observations and proposal for global accreditation. European journal of engineering
education, 32(6), 639-651.

Perrenet, J. C., Bouhuijs, P. A. J., & Smits, J. G. M. M. (2000). The suitability of

problem-based learning for engineering education: theory and practice.Teaching in
higher education, 5(3), 345-358.

Prados, J. W., Peterson, G. D., & Lattuca, L. R. (2005). Quality assurance of

engineering education through accreditation: The impact of Engineering Criteria
2000 and its global influence. Journal of Engineering Education, 94(1), 165-184.

Prince, M. J., & Felder, R. M. (2006). Inductive teaching and learning methods:
Definitions, comparisons, and research bases. Journal of engineering
education, 95(2), 123-138.

Prince, M., & Felder, R. (2007). The many faces of inductive teaching and
learning. Journal of College Science Teaching, 36(5), 14.

Electronic copy available at: https://ssrn.com/abstract=2560685

Shuman, L. J., Sindelar, M. F., Besterfield-sacre, M., Wolfe, H., Pinkus, R. L., Miller, R.
L., ... & Mitcham, C. Can Our Students Recognize and Resolve Ethical Dilemmas.
In CD) Proceedings, 2004 American Society for Engineering Education Conference.

Smith, K. A., Sheppard, S. D., Johnson, D. W., & Johnson, R. T. (2005). Pedagogies of
engagement: Classroom‐based practices. Journal of Engineering Education, 94(1),
87-101.

Sorby SA. Developing 3D spatial skills for engineering students. 2006. Technical
Paper. Vol 13 No 1. Australasian Journal of Engineering Education. Accessed
November 7, 2013. Available:
http://www.engineersmedia.com.au/journals/aaee/pdf/AJEE_13_1_Sorby.pdf

Tamayo, A., Gevera, E., & Aguilar, L. E. (2014). A Probabilistic Estimation of Passing
the Licensure Examination for Accountants. Available at SSRN 2394997.

Tamayo, A. (2014). Determining Predictors of Chemical Engineering Licensure

Examination. Available at SSRN.

Tamayo, A., & Quinto, I. (2014). Making Good Design Licensure Exam: An
Architecture Licensure Examination Study. Available at SSRN 2425489.

Tamayo, A., & Bernardo, G. (2014). Readiness for the Licensure Exam of the
Engineering Students. Available at SSRN 2395037.

Wilczynski, V., & Douglas, S. M. (1995). Integrating design across the engineering
curriculum: A report from the trenches. Journal of Engineering Education, 84(3),
235-240.

Electronic copy available at: https://ssrn.com/abstract=2560685