Title: Predictors of Academic Performance Among Pre-service Teachers:

The Case of a Private Higher Educational Institution in Northern Philippines

ed
Author: Kenneth L. Maslang

iew
Affiliation: Department Head, Social Science and Philosophy, Saint Mary’s University
Complete Address: Ponce St., District IV, Bayombong, Nueva Vizcaya, 3700, Philippines
Email: kenchong@smu.edu.ph
Mobile: +63 9084664123

r ev
er
pe
ot
tn
rin
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530
 PREDICTORS OF ACADEMIC PERFORMANCE AMONG PRE-SERVICE TEACHERS:
THE CASE OF A PRIVATE HIGHER EDUCATIONAL INSTITUTION

ed
IN NORTHERN PHILIPPINES

ABSTRACT

iew
Predictors or prior knowledge and conditions play a vital role in the academic life of students,
particularly their first and second year in the university. Using descriptive, correlation and
comparative methods, this study explored some of the cognitive and non-cognitive factors that could
predict academic performance. A multiple regression analysis, binary logistic regression and
repeated measure ANOVA were utilized to present the predictors for academic performance,

ev
scholarship and success in general. The results showed that the best predictor of academic
achievement is the Grade 12 – general weighted average (GWA). The college examination test (CET)
is a predictive of academic scholarship for the 1st and 2nd semesters of SY2018-2019, while Grade
12 – GWA is predictive of academic scholarship in all the three semesters including the overall

r
average of these. There are significant differences in the GWAs of pre – service teachers between
their first two semesters and first semester of their second year. Therefore, Grade 12 – GWA maybe

er
considered in the admission and selection policy of the university including the criteria for the
recommendation on whether a student should be taking board or non-board degree programs.
Likewise, the school may not necessarily look into the background of the students but rather improve
on the current aspects that could make the students more prepared and flexible in taking new
pe
conditions and challenges.

Keywords: Predictors of Academic Performance, College Entrance Test, High School GWA
ot

INTRODUCTION
tn

The value of education is incomparable and greatly irreplaceable. It is known to be an all-

inclusive system that could paved the way for the success of students in the future regardless of their
socio-cultural and political affiliation as well as their economic status. The individual development of
every citizen, if it would be magnified and its ripple effect, could reflect a nation’s growth and
advancement. Taking from the declaration of the United Nations, the basic building block of every
rin

nation is its educational system. It is an investment of every person and group that could amplify to
a nation that is progressive, cultured, flourishing, just and leading to a condition where the
development could be sustained (Maslang, 2021). This is what was enshrined in the 1948 Universal
Declaration of Human Rights which emphasizes for everyone their right to be educated. However,
ep

the right to education is not an end in itself but a precursor for the future that will lead to myriad of
opportunities, an avenue for greater life and betterment of communities and nations.

It has also been established that success in education is a fruition of numerous and
Pr

multifaceted factors. One of the crucial factors is the early part of tertiary education where students
have to bring out their learnings from prior knowledge, personal circumstances, and conditions in the
community where they came from. Some variables identified as predictors of academic performance

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530
 were classified into two, these are: cognitive – pre-entry academic performance, verbal and
quantitative aptitude, entrance test scores and noncognitive – age, socioeconomic status criteria

ed
(Laus, 2021), behavioral and motivational factors (Zivcic – Becirevic, 2017).

Meanwhile, academic failure in college is a huge burden not only for students and their
families but also reflective of the portfolio of the university. It is a common wisdom that failure of a

iew
student is also a failure of the school where this student is enrolled. Not in all cases, however, since
the school also has to maintain its academic standards. At any rate, some interventions must be in
place for some universities could not afford to lose even a single student.

For some academic institutions, college entrance examination (CEE) is the sole determinant
of college admission and this was found to be a significant predictor of academic achievement

ev
especially in the early years of students’ schooling (Bai et al., 2015). Some studies, on the other
hand, found out that the CEE could not be a predictor but the general point average (GPA) is and
other environmental aspects like support from the family of students (Duckworth et al., 2019).

r
Following related studies about predictors of academic success, some prior knowledge and
conditions are known to be helpful mostly in the early years (Zivcic – Becirevic, 2017). As students

er
progressed in higher years, the predictive power of these prior knowledge and conditions diminishes.
They could still be helpful but not as strong compared to the more important factors in the higher
years like the determination to succeed, current environment and diligence of students in studying.
pe
Similarly, academic performance is something that reflects students’ hard work and
perseverance in the school. This is considered as a crucial aspect of higher education institutions in
many countries as well as it is an important criterion for assessing the quality of educational
institutions (National Commission for Academic Accreditation, 2015) as cited by Alyahyan and
Dustegor (2020). Student success is defined in many literature; Kuh et al. (2006) related that student
ot

success is an academic achievement, engagement in educationally purposeful activities,

satisfaction, acquisition of desired knowledge, skills, and post-college performance. York et al.
(2015) simplified this definition in concentrating with six aspects, namely: academic achievement,
tn

satisfaction, acquisition of skills and competencies, persistence, attainment of learning objectives,

and career success.

Whatever definition one is referring to, there is no doubt that academic success could be
affected by myriad of factors. In general, these factors include prior knowledge and conditions,
rin

current environment and the psychological upbringing or the will of students to finish and become
successful in their academic pursuit. Alyahyan and Dustegor (2020) claimed that while there are
many factors that have been investigated on the predictors of students’ academic success, prior-
academic achievement, student demographics, e-learning activity, psychological attributes,
ep

and environments are the most commonly reported factors. For them, the top 2 factors are prior-
academic achievement and student demographics as these were presented in 69% of the research
papers they have reviewed.
Pr

Meanwhile, Duckworth et al. (2019) related that cognitive factors are broadly defined as the
“ability to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly,
and learn from experience”. These refer to the intelligence of individuals as they respond to paper

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530
 and pencil test or any assessment. The way an individual write in English, solve verbal problems in
Mathematics, understand critical theories and laws in the hard Sciences, and the like are all the

ed
characteristics of cognitive factors.

On the other hand, the Walberg’s theory of education productivity identified non cognitive
factors referring to profile variables (age, sex, school background), motivation, home environment,

iew
school environment, peer group, and mass media (Laus, 2021). Some behavioral and motivational
factors such as goal orientations, time and study management environment, Big Five Personalities
(neuroticism, extraversion, conscientiousness, agreeableness, and openness), were also included
by Zivcic – Becirevic (2017) as non-cognitive factors.

Models of predictive validity had been found out to have produced varying results based on

ev
the prior knowledge and profile characteristics or conditions of students entering in high school or in
college. For the cognitive factors, there were varying results of previous studies. The study of
Westrick et al. (2015) claimed that High school grade point average (GPA) and admission test scores
are individually valid predictors of undergraduate academic performance, moreover, if these will be

r
treated together they provide a more accurate prediction of future academic performance of students.

er
In China, Bai et al. (2015), with the use of multiple linear regression and inclusion of cognitive
factors and profile variables such as gender, birth year and month, ethnicity, the province from which
they were admitted, and whether the students were from a rural or urban area, they revealed that
college entrance exam is a significant predictor of undergraduate GPAs for all 4 years. They also
pe
proved that with CEE and all else controlled for, female students perform better in college.
Furthermore, aside from the CEE, high school performance measured by the level, types of award
and whether a student has received any award in high school also predict significantly the academic
performance of students in college.
ot

In the studies of Cadao-Esperal (2016), Takeley (2017) and Ferrão and Almeida (2019), they
emphasized on the role of entrance exam scores on the first year academic performance of students.
Their findings could also reflect the idea of historical baggage of students as presented by Alyahyan
tn

and Dustegor (2020). For Zivcic – Becirevic (2017), the predictors of university students’ academic
achievement include both cognitive and non-cognitive aspects. The study also proved that initial
academic adjustment to college is predictive of future academic achievement.

It is then useful to help students become better prepared for the college requirements through
rin

the organization of pre-college seminars, and to develop screening methods for early identification
of at-risk students with poor initial adjustment.

In the Philippines, Laus (2021) revealed that cognitive factors are predictors of academic
ep

success. The predictor that has the highest impact in the first year of students in high school is the
general point average of students in their Grade 6 level. Admission or entrance test is also a predictor
but with lesser predictive power. In particular, scores in Mathematics, English and AP/Fil are potential
predictors of student’s academic performance in Grade 7 while scores in Science and OLSAT are
Pr

not.

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530
 On the other hand, there were also studies which claimed that prior knowledge and conditions
do not necessarily predict academic achievement; thus, the prevailing conditions and the

ed
perseverance of student to learn could be the factors of academic success. There may be students
with no honors or awards in high school and very low entrance exam but could have high grades in
his or her subjects in college.

iew
This study then highlights several factors that are perceived to be predictors of academic
performance in college. These are relevant particularly for the policies and practices of universities
in selection and admission of students. Once admitted, universities have to perform some
interventions in the adjustment period – the first and second year of students in the university.
Likewise, this study provides some bases for monitoring students’ academic performance during
their stay in the academe. Specifically, this study endeavored to profile the respondents in terms of

ev
sex, course, ethnicity, type of High School, type of municipality, Grade 12 GWA, College Entrance
Test, and GWA in the First and Second Semester of SY 2018 – 2019 and First Semester of SY 2019
– 2020. Further, it also sought to find out whether the college entrance examination and Grade 12
general weighted average are predictors of academic performance. It also determined if the cognitive

r
and non-cognitive variables predictive of academic scholarship. Finally, it surfaced whether there is
a significant difference in the GWA of pre-service students in three succeeding semesters, namely,

er
1st sem SY2018-2019, 2nd sem SY2018-2019, and 1st sem SY2019-2022.
pe
METHODOLOGY

The study employed a quantitative approach using the descriptive – correlational and
comparative method. The descriptive part showed the profile of the respondents considered as:
predictors which were grouped as cognitive variables (Grade 12 GWA and college entrance test) and
ot

non – cognitive (sex, course, ethnicity, type of high school, type of municipality), and the dependent
variable (GWAs in first and second semester of SY 2018 – 2019 and first semester of SY 2019 –
2020). The correlational part covered the relations between the predictors and dependent variables
tn

in multiple linear and logistic regressions analyses. Lastly, the comparative aspect dealt with the
comparison of GWAs in the three succeeding semesters through repeated measures ANOVA.

The study was conducted in Saint Mary’s University, College Department. A private Catholic
HEI in Bayombong, Nueva Vizcaya supervised by the Congregatio Immaculati Cordis
rin

Mariae (CICM). The vision statement relates that SMU is a premier CICM Catholic educational
institution drawn into communion by the Wisdom of God, dedicated to forming persons exemplifying
excellence, innovation and Christ’s mission. Its motto in Latin is Sapientia a Deo or Wisdom from
God in English. It accommodates students mostly in Nueva Vizcaya but some feeder schools include
ep

the provinces of Quirino, Ifugao, Isabela and Cagayan. SMU is one of the top performing
schools/university in Region II (Maslang et al., 2020).

A total of 73 pre-service teachers were included in the study. There were 15 Bachelor of
Pr

Elementary Education (BEED), 17 Bachelor in Physical Education (BPE) and 41 Bachelor of

Secondary Education (BSED) students. This was a population study of third- and fourth-year
students enrolled in BEED and BSED program. Other profile variables were included such as

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530
 ethnicity, type of High School, type of municipality and Grade 12 GWA. The College Entrance Test
Score was requested from the Guidance and Testing Office, and the General Weighted Average in

ed
First and Second Semester of SY 2018 – 2019 and First Semester of SY 2019 – 2020 were requested
from the University Registrar.

iew
Table 1.
Profile of Respondents (N=73)
Profile Variables f %
Sex
Male 22 30.1
Female 51 69.9

ev
Course
BEED 15 20.5
BPE 17 23.3
BSED 41 56.2

r
Type of High School
Public 32 43.8
Private 41 56.2
Type of Municipality
Rural
Urban
Ethnicity
44
29
er 60.3
39.7
pe
Ilocano 21 28.8
Tagalog 18 24.7
Gaddang 10 13.7
Tuwali 16 21.9
Kalanguya 8 11.0
College Entrance Test Category
ot

Non-Board 27 37.0
Board Course 46 63.0
tn

For the treatment of data, frequencies and percents were used to present the profile variables
of the respondents. Multiple linear regression was used in the analysis for the college entrance test
and high school general weighted average as predictors of academic performance. Binary logistic
regression was utilized for the cognitive and non-cognitive variables as predictors of academic
scholarship. The GWAs were recoded with 0 for the non-academic scholars (89 and below) and 1
rin

for the academic scholars (90 and above). Lastly, repeated measure ANOVA was used in the
analysis of significant difference in the GWA of pre-service students in three succeeding semesters,
namely, 1st sem SY2018-2019, 2nd sem SY2018-2019, 1st sem SY2019-2022.
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530
 RESULTS AND DISCUSSION

ed
Predictors of Academic Performance

Two of the profile variables which were linear were taken as predictors of academic
performance, these were college entrance test (CET) and Grade 12 general weighted average

iew
(GWA). Table 2 showed the model summary of each of the GWA in three semesters and Table 3 for
the predictors and average of all the GWAs.

Table 2.
Model Summary for Predictors and GWA of 1st Sem, SY2018 – 19
Model Summaryb

ev
Model R R Square Adjusted R Square Std. Error of the Estimate Durbin-Watson
1 .624a .390 .373 2.76568 1.699
a. Predictors: (Constant), Grade 12 – GWA, CET
b. Dependent Variable: GWA 1st Sem SY2018-2019

r
Model Summary for Predictors and GWA of 2nd Sem, SY2018 – 19
Model Summaryb
Model
1
R
.686a .470 .455
a. Predictors: (Constant), Grade 12 – GWA, CET
er
R Square Adjusted R Square Std. Error of the Estimate
2.53756
Durbin-Watson
1.560
pe
b. Dependent Variable: GWA 2nd Sem SY2018-2019

Model Summary for Predictors and GWA of 1st Sem, SY2019 – 20

Model Summaryb
Model R R Square Adjusted R Square Std. Error of the Estimate Durbin-Watson
1 .845 a .713 .705 1.79119 1.875
ot

a. Predictors: (Constant), Grade 12 – GWA, CET

b. Dependent Variable: GWA 1st Sem SY2019-2020

From the models generated for each of the GWA, the 1st semester of SY 2019 – 2020 has the
tn

highest R coefficient. The regression analysis of the model for this has an R coefficient of .845 and
R squared of .713 with a standard error of the estimate at 1.79119 and DW value at 1.875. These
figures reflect that the predicted variable (the GWA of students in the 1st semester of SY 2019 – 2020)
will deviate from the true value by 1.79119 and the predictors could explain a total of 71.3% of the
rin

variability of the students academic performance in 1st semester of SY 2019 – 2020. Since DW is
below 2, this entails positive autocorrelation between the predictors (CEE and Grade 12 GWA) and
the dependent variable (GWA of students in the 1st semester of SY 2019 – 2020).
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530
 Table 3.
Model Summary for Predictors and All GWAs (1st Sem and 2nd SY2018 – 19 and 1st Sem SY2018 – 19)

ed
Model Summaryb
Model R R Square Adjusted R Square Std. Error of the Estimate Durbin-Watson
1 .868a .753 .746 1.55357 1.780
a. Predictors: (Constant), Grade12_GWA, CET

iew
b. Dependent Variable: All_GWA_Mean

The model summary of the predictors and all GWAs in Table 3 has an R coefficient of .868
and R squared of .753 with a standard error of the estimate at 1.55357 and DW value at 1.780. These
figures reflect that the predicted variable (overall GWA of students in the three semesters) will deviate
from the true value by 1.55357 and the predictors could explain a total of 86.8% of the variability of

ev
the students overall academic performance in the three semesters 1st semester of SY 2019 – 2020.
Since DW is below 2, this entails positive autocorrelation between the predictors (CEE and Grade
12 GWA) and the dependent variable (overall GWA of students in the three semesters).

r
For the coefficients of predictors, Table 4 showed that the computed p – value (.0005) of
Grade 12 GWA in the final model is similar to the p-value of the original one. This confirms the earlier
conclusion to reject the null hypothesis and so the Grade 12 – GWA is the best predictor of academic
er
achievement. This provided further that for every one-point increase in the Grade 12 – GWA, the
overall academic performance in the three semesters will increase by .781.
pe
Table 4.
Coefficients for Predictors and GWA of 1st Sem, SY2019 – 20
Coefficientsa
Unstandardized
Coefficients Standardized Coefficients
Model B Std. Error Beta t Sig.
ot

1 (Constant) 19.590 4.794 4.086 .000

CET .028 .017 .105 1.639 .106
Grade12_GW
.745 .058 .823 12.865 .000
tn

A
2 (Constant) 19.096 4.841 3.945 .000
Grade12_GW
.781 .054 .862 14.333 .000
A
a. Dependent Variable: All_GWA_Mean
rin

This result on the multiple linear regression analysis could be supported by the findings in the
reviewed studies that High school grade point average (GPA) and admission test scores are
predictors of academic performance (Westrick et al., 2015; Bai et al., 2015; Zivcic – Becirevic, 2017)
ep

and Grade 6 GPA as strong predictor of academic performance in Grade 7 (Laus, 2021). With these,
the Grade 12 – GWA can be a good inclusion in the consideration whenever the school will come up
with ranking or coming up with homogenous or heterogenous sections. As discussed by Alyahyan
and Dustegor (2020) in their concept of historical baggage, grades in the secondary school are the
Pr

most common factors used to predict student performance in the tertiary level.

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530
 Before the pandemic, one policy of the university is the categorization of those who will be
advised for board and non-board courses. A score of 90 and above in the entrance exam will merit

ed
students to enroll in the courses with board exam while non-board courses will be recommended for
students who will have a score of 89 and below. The administrators then may want to include the
Grade 12 – GWA as part of their criteria whether a student will be endorsed to the board or non-board
courses.

iew
Cognitive and Non-cognitive Variables as Predictive of Academic Scholarship

For comparison and more useful binary regression function, the overall academic
performance of the pre-service teachers in three semesters were entered with all the identified
predictors. The results were shown in Table 5.

ev
Table 5.
Binary Logistic Regression Results for the Overall Academic Performance in Three Semesters

r
Hosmer and Lemeshow Test
Step Chi-square df Sig.
1 1.065
er
Classification Tablea
8

Predicted
.998
pe
All_GWA_Mean
Non-academic Academic Percentage
Observed Scholar Scholar Correct
Step All_GWA_Me Non-academic
45 3 93.8
1 an Scholar
Academic Scholar 4 21 84.0
ot

Overall Percentage 90.4

a. The cut value is .500
tn

Table 5 showed the Binary Logistic Regression results for the overall academic performance
in the three semesters. Just like the three semesters, the Hosmer and Lemeshow test for goodness
of fit of the overall academic performance reflected that the model is not statistically significant (P =
.998) indicating that it is also a good fit in predicting academic performance.
rin

The classification table reflected that the model had correctly classified 90.4% of the students
who are academic scholars from non-academic scholars. The model had correctly predicted 45
students who are non-academic scholars out of the total of 48 making the accuracy of 93.8 % and it
had correctly predicted 4 students who are academic scholars out of the total of 21 making the
ep

accuracy of 84%. With a little lower values than the earlier three semesters, the independent
variables included in the study, in general, are good predictors of academic performance of students
in college.
Pr

Meanwhile, Table 6 showed the logistic regression predicting the likelihood of becoming an
academic or non-Academic scholar for the overall academic performance in three semesters based
on profile variables. Just like the third semester, only the Grade 12 – GWA with Wald (7.285, df=1), p

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530
 – value = .007 can significantly predict whether the pre-service teachers could become an academic
scholar or not. It showed further that for every one unit increase in Grade 12 – GWA, there is 7.285

ed
odds of becoming an academic scholar. The students’ sex, course, ethnicity high school graduated
from, CEE, category and ethnicity are not significant regressors of becoming an academic scholar.
This implies that except for the Grade 12 – GWA, the binary categories of all the independent
variables have the same in their likelihood of becoming an academic scholar considering the overall

iew
academic performance in the three semesters.

Table 6.
Logistic Regression Predicting Likelihood of Becoming an Academic or Non-Academic Scholar for the
Overall Academic Performance in three Semesters based on Profile Variables
Variables in the Equation

ev
Odds Ratio 95% C.I.for EXP(B)
B S.E. Wald df Sig. Exp(B) Lower Upper
Step Gender -.086 1.366 .004 1 .950 .917 .063 13.349
1a Course 5.606 3.249 2.977 1 .084 272.051 .467 158597.192

r
Highschool -.537 1.011 .282 1 .596 .585 .081 4.243
Mun -2.056 1.194 2.968 1 .085 .128 .012 1.327
CET .005 .069 .005 1 .941 1.005 .879 1.150
Ilocano
Tagalog
Gaddang
BEED
3.257
-3.285
1.945
9.444
1.810 3.239 1 .072
1.845 3.171 1 .075
1.627 1.429 1 .232
er 25.981
.037
6.991
.748
.001
.288
6.152 2.356 1 .125 12629.045 .073 2177218129.343
902.086
1.392
169.581
pe
Grade12_GW
1.986 .736 7.285 1 .007 7.285 1.723 30.810
A
Constant -192.496 69.290 7.718 1 .005 .000
a. Variable(s) entered on step 1: Gender, Course, Highschool, Mun, CEE, Ilocano, Tagalog,
Gaddang, BEED, Grade12_GWA.
ot

Comparison of GWAs of Pre-Service Teachers

tn

To strengthen the value of the predictors of academic performance, it could be helpful to

present where the pre-service students perform the most in the three succeeding semesters being
observed. Table 7 presented the descriptive statistics, and multivariate tests.
rin

Table 7.
Descriptive Statistics and Multivariate Tests of the GWAs
Descriptive Statistics
Mean Std. Deviation N
ep

1. GWA 1st Sem SY2018-2019 87.6167 3.49138 73

2. GWA 2nd Sem SY2018-2019 87.3534 3.43734 73
3. GWA 1st Sem SY2019-2020 89.9162 3.29873 73
Multivariate Testsa
Hypothesis Partial Eta
Pr

Effect Value F df Error df Sig. Squared

GWA Pillai's Trace .590 51.070b 2.000 71.000 .000 .590
Wilks' Lambda .410 51.070b 2.000 71.000 .000 .590

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530
 Hotelling's Trace 1.439 51.070b 2.000 71.000 .000 .590
Roy's Largest Root 1.439 51.070b 2.000 71.000 .000 .590

ed
a. Design: Intercept Within Subjects Design: semester
b. Exact statistic

The computed p – values of the multivariate tests are all below .05 indicating a significant

iew
difference. The decision for this problem is that the null hypothesis is rejected which means that there
is a significant difference in the GWA of Teacher Education students in three succeeding semesters,
namely, 1st sem SY2018-2019, 2nd sem SY2018-2019, 1st sem SY2019-2022. To know more about
this significant difference and where the difference occurred, Mauchly’s test of sphericity and pairwise
comparison were performed.

ev
Table 8.
Mauchly’s test of sphericity and Tests of Within-Subjects Effects
Mauchly's Test of Sphericitya
Measure: GWA

r
Epsilonb
Within Subjects Mauchly' Approx. Chi- Greenhouse- Lower-
Effect sW Square df Sig. Geisser Huynh-Feldt bound
Semester .946 3.953 2 .139
er .949 .974 .500
Tests the null hypothesis that the error covariance matrix of the orthonormalized transformed
dependent variables is proportional to an identity matrix.
pe
a. Design: Intercept
Within Subjects Design: Semester
b. May be used to adjust the degrees of freedom for the averaged tests of significance.
Corrected tests are displayed in the Tests of Within-Subjects Effects table.
Tests of Within-Subjects Effects
Measure: GWA
ot

Partial
Type III Eta
Sum of Mean Square
Source Squares df Square F Sig. d
tn

Semester Sphericity Assumed 290.161 2 145.081 63.837 .000 .470

Greenhouse-Geisser 290.161 1.897 152.936 63.837 .000 .470
Huynh-Feldt 290.161 1.947 149.019 63.837 .000 .470
Lower-bound 290.161 1.000 290.161 63.837 .000 .470
Error Sphericity Assumed 327.267 144 2.273
rin

(Sem) Greenhouse-Geisser 327.267 136.603 2.396

Huynh-Feldt 327.267 140.195 2.334
Lower-bound 327.267 72.000 4.545
ep

Based on the Mauchly Test of Sphericity, W = 0.946, p = .139 > .05, the assumption of
sphericity is met. Thus, the row on sphericity assumed was considered, so F = 63.837, df = 2, p <
0.0005 and since p < 0.0005 < .05, this validates the conclusion that there is a significant difference
in the GWAs of students in the three succeeding semesters, namely, 1st sem SY2018-2019, 2nd sem
SY2018-2019, 1st sem SY2019-2022.
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530
 Table 9.
Pairwise Comparison

ed
Pairwise Comparisons
Measure: gwa
95% Confidence Interval for
(J) Mean Differenceb

iew
(I) GWA GWA Difference (I-J) Std. Error Sig.b Lower Bound Upper Bound
1 2 .263 .219 .701 -.274 .801
3 -2.299 * .258 .000 -2.933 -1.666
2 1 -.263 .219 .701 -.801 .274
3 -2.563 * .268 .000 -3.220 -1.905
3 1 2.299* .258 .000 1.666 2.933

ev
2 2.563 * .268 .000 1.905 3.220
Based on estimated marginal means
*. The mean difference is significant at the .05 level.
b. Adjustment for multiple comparisons: Bonferroni.

r
The pairwise comparison shows further that the significant differences exist between GWA
1st
Sem SY2018-2019 and GWA 1st Sem SY2019-2020 and between GWA 2nd Sem SY2018-2019

er
and GWA 1st Sem SY2019-2020. For the extended inequality, the following could be formulated:

1=2, 1< 3, and 2<3 or in ascending order based on the means 2=1<3
pe
The above presentation means that:
1. there is no significant difference between GWA 1st Sem SY2018-2019 and GWA 2nd Sem
SY2018-2019;
2. GWA 1st Sem SY2018-2019 is less than GWA 1st Sem SY2019-2020; and
3. GWA 2nd Sem SY2018-2019 is less than GWA 1st Sem SY2019-2020.
ot

These findings imply that the pre-service teachers had the highest GWA in the 1st Sem
SY2019-2020 and it is uncertain whether they perform better in the 1st Sem SY2018-2019 or 2nd Sem
of the same year. One possible and obvious explanation here is that since the first year of students
tn

is consisted of myriad of conditions where they are still adjusting from high school to university life,
a lot of concerns, skepticism and confusions are being faced by these students in their new
environment. Therefore, their academic performance is greatly affected. As they settled in the second
year and since they survived their first year, it is expected that they will perform better, hence their
rin

higher GWA in their second year.

This implies further that predictors or students’ prior knowledge and conditions are very
helpful for students in their first or second year in college but their impact is limited and could not
ep

account consistently in different conditions. In the study of Ferrão and Almeida (2019), they noted
that aptitude test, entrance exams and high school GPA are predictors of academic success in
college but these could only account for about 25% on the variance of mean scores during the first
year of students in the university. Other factors, particularly the prevailing conditions actually account
for the greater fraction.
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530
 Summary of findings

ed
This study dealt with 73 pre-service teachers, one third of these were male and almost two
thirds were female. For the course, nearly one fourth each belongs to BEED and BPE while a little
more than fifty percent were composed of BSED students. With regard to the type of High School,
most of the students came from the private school with 56.2%. For the type of municipality, 60.3%

iew
comes from the rural areas and 39.7% were from urban areas. For the ethnicity, nearly one third
belong to the Ilocano group, followed by the Tagalog with 24.7%, Tuwali at 21.9%, Gaddang at 13.7%
and Kalanguya at 11%. Lastly, on the College Entrance Exam category, 63% got scores classified
as board course and 37% for the non-board courses.

Three models were generated, and among these, model 3 has the highest R coefficient. The

ev
computed p – values of the three ANOVA tables for Predictors and GWAs were all less then .05. The
CET had a p – value less than 0.05 only in the second semester of SY2018 – 2019 while the Grade
12 – GWA had p – values less than .05 in all the three semesters. The summary model of the overall
academic performance showed that the computed p – value (.0005) of Grade 12 GWA in the final

r
model is similar to the p-value of the original one.

The CET had p – values less than .05 only in the second semester of SY 2018 – 2019 while

er
the Grade 12 – GWA had p – values less than .05 in all the three semesters including the overall
peformance. All the other variables had p – values greater than .05 in all the three semesters. The p
– values of the multivariate tests are all below .05. The Mauchly Test of Sphericity had W = 0.946
pe
and p = .139. The extended inequality formulated was 1=2, 1< 3, and 2<3 or in ascending order
based on the means 2≤1<3.

Conclusions

The study was dominated by female pre-service teachers. Most of them: belong to the BSED
ot

program; from rural areas and private schools; Ilocano; and got scores belonging to the Board
Course. The best predictor of academic achievement in three semesters, namely: 1st sem SY2018-
2019; 2nd sem SY2018-2019; and 1st sem SY2019-2022, and overall is the Grade 12 – GWA. CET is
tn

a predictive of academic scholarship for the 1st and 2nd semesters of SY2018-2019, while Grade 12
– GWA is predictive of academic scholarship in all the three semesters including the overall average
of these. Finally, there is a significant difference in the GWAs of pre – service teachers between 1st
Sem SY2018-2019 and 1st Sem SY2019-2020 and between 2nd Sem SY2018-2019 and 1st Sem
SY2019-2020.
rin

Recommendations

Other predictors may be included in future researches like from their prior conditions are
ep

motivational and psychological factors (environment and support systems) and current conditions
such as teacher factor, peer influence, academic curriculum and school facilities. Grade 12 – GWA
maybe considered in the criteria for the recommendation on whether a student should be taking
board or non-board degree programs. The school may not necessarily look in to the background of
the students but rather improve on the current aspects that could make the students more prepared,
Pr

resilient and flexible in taking new conditions and challenges. Lastly, It is useful to help students
become better prepared for the college requirements through the organization of pre-college

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530
 seminars, and to develop screening methods for early identification of at-risk students with poor initial
adjustment. This strengthens the responsibility of universities to provide support services to help

ed
students' overcome possible barriers in achieving their academic goals.

Declaration of Conflicting Interests

iew
The author declared no potential conflicts of interest with respect to the research, authorship, and/or
publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this
article.

ev
References

r
Alyahyan, E. & Dustegor, D. (2020). Predicting academic success in higher education: Literature
review and best practices. International Journal of Educational Technology in Higher
Education, 17 (3).
er
Bai, C., Chi, W. & Qian, X. (2015). Do college entrance examination scores predict undergraduate
GPAs? A tale of two universities. China Economic Review, 30, 632-647.
pe
Cadao-Esperal, D. L. (2016). Correlates of academic performance of College of Arts and Sciences
students: Basis for intervention. Tilamsik,8(2). http://ejournals.ph/ form/cite.php?id=9985.

Coovert, C. (2017). Understanding how incoming first year college students develop academic
performance expectations: A qualitative case study. https://digital
commons.lsu.edu/cgi/viewcontent.cgi?article=5323&context=gradschool_dissertations
ot

Davidovitch, N. & Soen, I.D. (2015). Predicting academic success using admission profiles. Journal
of International Education Research – Third Quarter, 11(3), 125-142.
tn

Duckworth, A., Quirk, A., Gallop, R. & Matthews (2019). Cognitive and non-cognitive predictors of
success. Psychological and Cognitive Sciences, 116 (47). https://
doi.org/10.1073/pnas.1910510116

Ferrão, M. & Almeida, L. (2019). Differential effect of university entrance score on first-year students’
rin

academic performance in Portugal. Assessment & Evaluation in Higher Education, 44 (4),

10-622.

Kuh, G. D., Kinzie, J., Buckley, J. A., Bridges, B. K., & Hayek, J. C. (2006). What matters to student
success: A review of the literature commissioned report for the National Symposium on
ep

postsecondary student success: Spearheading a dialog on student success.

https://nces.ed.gov/npec/pdf/Kuh_Team_ Report.pdf

Laus, M. (2021). Admission profiles as predictors of academic performance. Asia Pacific Journal of
Multidisciplinary Research, 9 (1), 39 – 47.
Pr

Lopez – Garrido, G. (2020). Self-efficacy theory. https://www.simplypsychology.org/ self-efficacy.html

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530
 Maslang, K., Baguilat, I., Mania, E., Damayon, S. & Dacles, D. (2020). Student services awareness
and satisfaction in a private higher education institution. American Journal of Educational

ed
Research, 2021, Vol. 9, No. 12, 708 – 719.

Maslang, K., Baguilat, I., Mania, E., Damayon, S., & Dacles, D. (2021). Student services awareness
and satisfaction in a private higher education institution amid the pandemic. American Journal
of Educational Research, 9 (12), 708-719.

iew
Ocbian, M.M., Gamba, M. P. & Ricafort, J D. (2015). Admission test as predictor of performance of
students in the english subject. JPAIR Institutional Research,
6(1), 34-45. https://doi: 10.7719/irj.v6i1.361

Westrick, P., Le,H., Robbins, S., Radunzel, J. & Schmidt, F. (2015). College performance and

ev
retention: A meta- analysis of the predictive validities of ACT scores, high school grades, and
SES. Educational Assessment, 2(1), 23–45. https://doi.org/10. 1080/10627197.2015.997614

Sulphey, M. M., Al-Kahtani, N.S. & Syed, A.M. (2018). Relationship between admission grades and
academic achievement, entrepreneurship, and sustainability issues. Entrepreneurship and

r
Sustainability Center, 5 (3), 648 – 658.

Takele, S. (2017). Validity strength of college entrance assessment score and high school academic

67-75. er
records in predicting college academic performance. Journal of Education and Practice, 8(7),

Zivcic – Becirevic, I., Dorcic, M. & Smojver – Azic, S. (2017). Predictors of university students’
pe
academic achievement: A prospective study. Research Gate Publication, 457 – 476. DOI:
10.5559/di.26.4.01.
ot
tn
rin
ep
Pr

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4325530