1

Readiness for the Licensure Exam of the

Engineering Students

Adrian M. Tamayo
Research and Publication Center
University of Mindanao

Geffren Bernardo
College of Engineering
University of Mindanao

Rec Eguia
College of Development Management
University of Southeastern Philippines

Abstract
The paper evaluated the performance of the civil engineering
graduates of the University of Mindanao in the licensure exam for period
2009 to 2011. The paper employed maximum likelihood estimation of a
model with categorical variable.
The findings revealed that grade point average, design and
construction, mathematics and hydraulics and survey subjects predict
passing the licensure examination. A simulation of the predictor variables
determines the degree of readiness of the students for the examination.

Keywords: Licensure examination, engineering education, logistic analysis

INTRODUCTION

Licensure examination to practice profession is a regulatory mechanism of a

State. It has many functional uses. In the United States, a law was passed which used

licensure results as barometer of teaching quality of an educational institution (Mitchell,

Robinson, Plake & Knowles, 2001). It is deemed important in the career advancement in

the practice of the engineering profession (Johnston, Ahluwalia & Gwyne, 2007) as

Electroniccopy
Electronic copy available
available at:
at:https://ssrn.com/abstract=2395037
http://ssrn.com/abstract=2395037
 2

evidence of equivalent competency across countries (Kasuba & Vohra, 2004; Kelly,

2007).

Higher learning institutions employ the licensure and certification exams as

measure for practice analysis (Raymond, 2010) of the profession and eventually improve

curricula (Hedderick, 2009). The effectiveness of a curriculum is also measured using

results of the licensure examinations (Stewart, Bates & Smith, 2004). Given these,

schools usually develop remediation measures to increase number of passing practitioners

such as mock boards and remediating tests to determine readiness for licensure

examinations (Sifford & McDaniel, 2007; Simon, McGinnis & Krauss, 2013; Schmidt,

2000). This study determined factors that predict passing the licensure examination for

civil engineers.

METHOD

This section presents the research design, variables and data, and the empirical

model used to predict board outcome and simulate scores.

Research design

The paper employed prediction model which had been used by many studies to

determine corporate failure (Ohlson, 1980); measurement of future landslide hazard

(Chung & Fabbri, 1999); it is also employed in clinical trials to select prognostic factors

(Spiegelhalter, 2006); and employed in geosciences by predicting wave energy source of

random seas (Ochi, 1982). The maximum likelihood estimation of the logistic regression

generated estimates for the categorical values of passed and failed in the licensure exam

(Press & Wilson, 2012; Silvapulle, 1981)

Electroniccopy
Electronic copy available
available at:
at:https://ssrn.com/abstract=2395037
http://ssrn.com/abstract=2395037
 3

Variables and Data

The variables that were used to determine the probability outcome of the civil

engineering board exam were the grade point average and the correlation grade which is

an intervention measure offered by University of Mindanao to civil engineering students;

the students’ scores in the licensure exam for civil engineers were also gathered. Note

that the variables could be grouped to two: the scholastic performance and the licensure

exam performance. The former is the student’s academic preparation while the latter

could be considered as the target scores in order to pass the board.

The data were collected from the students’ permanent record of the students who

took the exam for covering the period 2009 to 2011 while the licensure exam

performance were taken from the State’s commission for the conduct of licensure

examination, the Professional Regulation Commission.

Model

The study employed the logistic regression and the linear regression analyses. The

empirical tool is appropriate to the study because the dependent variable takes a

dichotomous categorical value ( pass = 1; fail = 0) (Bewick, Cheek & Ball, 2005) It is

assumed that the probability of passing increases as the grade indicators increase until a

threshold for passing is reached equal to probability of 1, thus the probability of passing

the board exam is assumed to take an S − shaped function. The log-odds of the

 π 
proportion of the dependent variable takes the form log it (π ) = log e  .
1 + π 

Electronic copy available at: https://ssrn.com/abstract=2395037

 4

FINDINGS

For the last 3 years, 2009 to 2011, the average passing percentage was merely

27.85%; the highest passing percentage was in May 2009 with 37.50% while the lowest

was in November 2011 with 23.33%. In 2011, the University of Mindanao’s passing

percentage (25.40%) was way below the 38.34% national passing percentage1.

It could be observed that the passing percentage of the Civil Engineering

graduates of the University of Mindanao has been depreciating.

Of those who took the board exam, their aggregative scholastic performance was

found to be not strikingly impressive. The scholastic performance indicators in the study

were grade point average (Harris, 2006; Heupel, 1994) and the correlation subject which

is an intervention program of the institution (Crow, Handley, Morrison & Shelton, 2004).

The maximum GPA was 2.00 (90), the minimum was 3.82 (72), mean GPA was

Table 1
Number of passers and flunkers per exam period
Exam period Passers Flunkers % of passing
May 2011 17 50 25.40
November 2011 7 23 23.33
May 2010 18 51 26.09
November 2010 16 49 24.62
May 2009 30 50 37.50
November 2009 22 51 30.14
Average passing percentage 27.85

1
(http://www.coolbuster.net/2011/05/civil-engr-board-exam-results-may-2011.html)

Electronic copy available at: https://ssrn.com/abstract=2395037

 5

3.05 (80), median of 3.04(80) and mode of 3.00 (80). Note that the students’ GPA

congregated around the score of 3 (80) which could be considered as a norm grade for

civil engineering education.

Meanwhile, the correlation subject indicated a grade of 2.20 (88) as the highest,

5.00 as the lowest grade, average was 3.11 (79), median of 3.10 (79) and 3.5 (75) as the

most common grade. The correlation subject serves as review course for the graduating

civil engineering students where their knowledge on the fundamentals to advanced

engineering was evaluated with the intention of fortifying their potential to pass the

licensure exam.

Table 2
Scholastic Performance Indicators
Personal ability indicators Maximum Minimum Mean Median Mode
GPA 2.00 3.82 3.05 3.04 3.00
Correlation grade 2.20 3.50 3.11 3.10 3.50

The empirical model used to determine the causal relationship between board

exam outcome and the scholastic performance was given as:

pass
ln( ) = 6.03 − 2.43GWA * *i + 0.16Corri
fail

Nagelkerke R 2 =0.164
Cox and Snell R 2 =0.116

The model could explain 11.6% to 16.4% variation of the data. The empirical

exercise revealed that only the GPA could predict the board exam outcome, while the

correlation course proved no significant prediction ability. The negative coefficient also

indicates that a higher GPA would ensure a better chance in the exam. It is surprising to

note that the correlation course lacks the ability to determine success in the board. This is

Electronic copy available at: https://ssrn.com/abstract=2395037

 6

quite a revelation given that the very nature of the correlation course is supposed ready

the civil engineering students for the licensure exam.

Table 4 presents the simulation model to determine the readiness of the graduates

to take and pass the licensure exam for the civil engineers; also a predicted score in the

board exam was determined using the GPA and the correlation course grade. Although, it

was determined that the correlation grade did not statistically signify influence in the

board outcome yet, in the simulation model, determining a threshold correlation grade

might improve a student’s performance in the board.

Table 3
Simulation of personal ability indicators to determine probability of passing
Constant GPA Correlation Odds Predicted
Readiness average
Simulations 6.03 -2.43 0.16 score
GPA=2, Corr=2.2 416.8 -0.18 2.57 419 99.76 78.31
GPA=3.05;Corr=3.11 416.8 -0.27 3.64 420 80.00 61.42
GPA=2.5,Corr=3.5 416.8 -0.22 4.10 421 82.00 69.65
GPA=2.5,Corr=3.0 416.8 -0.18 3.51 420 99.76 70.00

Note that if a student able to obtain a GPA of 2.0 and a correlation grade of 2.2,

the student’s readiness to take the board is estimated at 99.76% and the student’s

predicted score in the board exam would be 78.31. A student with a GPA of 3.5 and

correlation grade of 3.11 would render an 80% readiness and an average score of 61.42; a

GPA of 2.5 and correlation grade of 3.0 yielded a readiness of 99.76% and a predicted

score of 70. Apparently, a GPA of 2.5 would render a safe threshold as predictor for the

board exam, and while considered as pseudo predictor, a correlation grade of 3.0 would

notch up the chance in the licensure exam.

Table 4
Civil Engineering board exam cluster subjects

Electronic copy available at: https://ssrn.com/abstract=2395037

 7

covering 2009, 2010, 2011

Subject clusters Maximum score Mean score Median score Mode score
Mathematics 100 59.76 58 50
Hydraulics and
surveying 100 63.83 64 95
Design and
construction 95 56.41 56 70

The licensure examination is widely grouped to three subjects, in this paper for

purposes of simplicity were termed as subject cluster 1, 2 and 3. Table 5 reveals the

descriptive statistics of the three subject clusters by the maximum score and the central

tendencies. Subject cluster 1 which represents mathematics had a 100 as the highest score

over the last three years, 59.76 for an average score, 58 as median and 50 as the common

score of the exam takers. Subject cluster for hydraulics and surveying had 100 as the

highest score, 63.83 as the average, 64 as median and 95 as common score. Subject

cluster for design and construction had 95 as maximum score, 56.41 as average, 56 as the

median and 70 as the common score.

The prediction model was developed to determine the influence of GPA,

correlation, subject clusters in the chances in the licensure exam, hence the empirical

model

pass
ln( ) = −149.42 − 0.24GPA + 0.71Subj1 * * * +0.58Subj 2 * * * +0.87 subj 3 * * *
fail

Nagelkerke R 2 = 0.97
Cox and Snell R 2 = 0.68

The model could explain 68% to 97% variation of the data. The correlation course

was removed from the list of the predictors due to its inability to predict board outcome.

The empirical model was a combination of the scholastic performance and the target

score of the subject cluster. It was portrayed that the GPA did not have strong statistical

Electronic copy available at: https://ssrn.com/abstract=2395037

 8

influence to predict the licensure exam result when combined with the subject clusters.

On the one hand, the entire subject clusters have strong statistical ability to predict board

outcome.

Of the three cluster subjects, the subject cluster for design and construction had

the strongest influence in the board exam; subject cluster of mathematics comes as the

second strongest, while subject cluster for hydraulics and surveying as the least

determinant.

The simulation of the model was constructed to determine licensure exam

outcome and predicted score given the values of the predictors; the GPA was included in

the simulation.

Note that a student with a GPA of 2.00, aimed to obtain a score of 100 in subject

clusters 1 and 2, and 95 for subject cluster 3 would yield a 99.84% probability of success

with a predicted score of 98.25 in the board exam; a GPA of 3.04, subject cluster 1 score
Table 5
Simulation of personal ability indicators to determine
probability of passing
Constant GPA Subj1 Subj2 Subj3
- Prob of Predicted
Simulations -149.42 0.24 0.71 0.58 0.87 success score
GPA=2; -
Subj=100;subj2=100;subj3=95 0.00 1.57 204.27 178.98 226.02 99.84 98.25
GPA=3.04;subj=50; -
subj2=95; subj3=70 0.00 2.38 102.14 170.03 166.54 99.77 70.58
GPA=3.5;subj=70; subj2=65; -
subj3=75 0.00 2.74 132.78 125.29 178.43 99.77 70.35
GPA=3.04;subj=70; -
subj2=70; subj3=70 0.00 2.38 142.99 125.29 166.54 99.77 70.05
of 50, 95 for cluster 2 and a score of 70 for cluster 3 would give a 99.77% chance of

success and a predicted score of 70.58 in the licensure exam; a GPA of 3.5, subject

cluster mathematics of 70, subject cluster of hydraulics and surveying score of 65 and

Electronic copy available at: https://ssrn.com/abstract=2395037

 9

design and construction with 75 would result to a 99.77% probability of success with a

score of 70.35; a GPA of 3.04, and a score of 70 for clusters mathematics through design

and construction would yield a chance of 99.77% of passing the board exam with a score

of 70.05.

What was clear in the simulation was the causal ability of the subject clusters to

determine success in the licensure exam. In particular, examinees must give bigger

weights on subject cluster number design and construction, followed by subject cluster

mathematics, then subject cluster of hydraulics and surveying. Threshold scores would

be: subject cluster of mathematics is 70; subject cluster hydraulics and surveying is 65

and subject cluster for design and construction is 75. The paper did not put in the outskirt

the necessity of getting good grades while within the academic institution which would

assure higher probability of success in the board exam.

CONCLUSION

The empirical model developed in this paper found out that the variables that

predicts passing in the licensure examination for civil engineers are [1] grade point

average; [2] mathematics subject; [3] hydraulics and survey subjects; and [4] design and

construction subjects. The design and construction subjects have the relatively strongest

influence to pass followed by mathematics, and then the least to show influence was

hydraulics and survey. It was found that the intervention subject known as the correlation

course did not predictive ability to determine passing in the exam.

Electronic copy available at: https://ssrn.com/abstract=2395037

 10

REFERENCES

Bewick V, Cheek L, Ball J (2005). Statistics review 14: Logistic regression. Crit Care,
9(1), 112-118.

Civil Engineer Licensure Exam Results May 2011 — Civil Engineering (CE) Board
Passers May 2011. http://www.coolbuster.net/2011/05/civil-engr-board-exam-results-
may-2011.html. Retrieved June 15, 2012.

Chung CJF, Fabbri AG (1999). Probabilistic prediction models for landslide hazard
mapping. Photogrammetric engineering and remote sensing, 65(12), 1389-1399.

Crow CS, Handley M, Shaw Morrison R, Shelton MM. (2004). Requirements and
interventions used by BSN programs to promote and predict NCLEX-RN success: A
national study. Journal of Professional Nursing, 20(3), 174-186.

Ellis, AM. HB 4071 An Act Further Amending, As Amended, Or the Civil Engineering
Law. http://www.agham.org.ph/legislation/bills/authored/hb-4071-an-act-further-
amending-republic-act-no-544-as-amended-or-the-civil-engineering-law.html. Retrieved
June 15, 2012

Harris MS (2006). Investigation of prerequisite science course performance and

cumulative grade point average as predictors of success on the National Council
Licensure Examination for Registered Nurses.

Hedderick, V M. (2009). The utilization of NCLEX predictor assessments and the effect
on NCLEX success rates in nursing programs within the state of Pennsylvania (Doctoral
dissertation, Indiana University of Pennsylvania).

Heupel, C. (1994). A model for intervention and predicting success on the national
council licensure examination for registered nurses. Journal of professional nursing,
10(1), 57-60.

Johnston DW, Thomas Ahluwalia N, Gwyn MB (2007). Improving the professional

engineering licensure process for construction engineers. Journal of Construction
Engineering and Management, 133(9), 669-677.

Electronic copy available at: https://ssrn.com/abstract=2395037

 11

Kasuba R, Vohra P (2004). International mobility and the licensing of professional

engineers. World Transactions on Engineering and Technology Education, 1, 43-46.

Kelly WE (2007). Certification and accreditation in civil engineering. Journal of

Professional Issues in Engineering Education and Practice, 133(3), 181-187.

Layton, Edwin (1986). The Revolt of the Engineers: Social Responsibility and the
American Engineering Profession. Baltimore, Maryland, USA: The Johns Hopkins
University Press. ISBN 0-8018-3287-X. (pp. 6–7)

Mitchell KJ, Robinson DZ, Plake BS, Knowles K T (Eds.). (2001). Testing teacher
candidates: The role of licensure tests in improving teacher quality. National Academies
Press.

National Council of Examiners for Engineering and Surveying. Model Law. 2009.
http://www.ncees.org/About_NCEES/Publications/Publications/Model_Law.php.
Retrieved June 15, 2012

National Soceity of Professional Engineers. Licensure by Comity. 2008.

http://www.nspe.org/Licensure/Resources/LicComity/index.html. Retrieved 2008-03-14

Ochi MK (1982). Stochastic analysis and probabilistic prediction of random seas. Adv.
Hydrosci.;(United States), 13.

Ohlson JA (1980). Financial ratios and the probabilistic prediction of bankruptcy.

Journal of accounting research, 18(1), 109-131.

Press SJ, Wilson S (1978). Choosing between logistic regression and discriminant
analysis. Journal of the American Statistical Association, 73(364), 699-705.

Raymond M R (2001). Job analysis and the specification of content for licensure and
certification examinations. Applied Measurement in Education, 14(4), 369-415.

Schmidt AE (2000). An approximation of a hierarchical logistic regression model used to

establish the predictive validity of scores on a nursing licensure exam. Educational and
Psychological Measurement, 60(3), 463-478.

Sifford S, McDANIEL DM (2007). Results of a remediation program for students at risk

for failure on the NCLEX exam. Nursing Education Perspectives, 28(1), 34-36.

Silvapulle MJ (1981). On the existence of maximum likelihood estimators for the

binomial response models. Journal of the Royal Statistical Society. Series B
(Methodological), 310-313.

Simon EB, McGinniss SP, Krauss BJ (2013). Predictor Variables for NCLEX-RN
Readiness Exam Performance. Nursing education perspectives, 34(1), 18-24.

Electronic copy available at: https://ssrn.com/abstract=2395037

 12

Spiegelhalter DJ (1986). Probabilistic prediction in patient management and clinical

trials. Statistics in medicine, 5(5), 421-433.

Stewart CM, Bates RE, Smith GE (2004). Does performance on school-administered

mock boards predict performance on a dental licensure exam?. Journal of dental
education, 68(4), 426-432.

Electronic copy available at: https://ssrn.com/abstract=2395037