Home Instructors Journals ContactUs 

JOURNAL OF RESEARCH IN NATIONAL DEVELOPMENT VOLUME 7 NO 1, JUNE, 2009
THE PREDICTIVE VALIDITY OF WEST AFRICAN
SENIOR SECONDARY CERTIFICATE EXAMINATION (WASSCE) FOR ACADEMIC PERFORMANCE IN THE
UNIVERSITY
Juliet
Ogadinma Ajuonuma
Faculty
of Education, Imo State University, Owerri
and
N.D.
Mkpa
Pre
Degree Programme Unit, Abia State University, Uturu
Abstract
The
study was designed to investigate the Predictive Validity of WASSCE and first and
second year Grade Point Averages (GPAs) in Imo and Abia state universities. Students
in departments of mathematics, physics, chemistry, biology, zoology, microbiology
and statistics of the above universities whose records were available for 1997/98
to 2000/01 sessions were purposively selected to constitute the sample. The relationship
between the WASSCE scores and first and second years grade point averages in the
above mentioned subjects were determined using regression model and ttest. Result
shows that there is significant relationship between students’ grade points in WASSCE;
and first year and second year GPAs of students in both universities. The significance
showed that students who did well in WASSCE also did well in both universities.
It was therefore recommended that the said universities should continue to use the
services of West African Examination Council for placement of students.
Key
words: Predictive; validity; performance;
academic.
Introduction
Validity of a test refers to the extent to
which the test measures what it is supposed to measure.
A test is said to have predictive validity if an individual’s score in the
test can be used to predict his future performance (Nwana in Ajuonuma, 2004). Predictive Validity is therefore basically
concerned with how well a test can predict future behaviour.
Among the tribunal’s findings were that WAEC
had excessive load, having taken up the conduct of nearly all the public examination
in the country and this had in turn affected drastically the effective conduct of
WASSCE (Nkwocha, 2002) A renowned measurement
and evaluation expert, O.C. Nwana at the 15^{th} conference proceedings
of Association of Educational Assessment in Africa held in 1981 at Kaduna in Nigeria
noted that the credibility of Nigerian certificates (WASSCE inclusive) has started
facing serious scrutiny in many foreign countries because of the rate of examination
malpractices in the country. Based
on the above, the credibility of WASSCE results became in doubt.
Since WASSCE results appear to have lost credibility, it implies that universities
may not have been admitting the right candidates (since they use WASSCE results
as part of requisite qualifications
for admission) and also that WASSCE results may not have predictive validity. The wonder then is whether the fears
stated above are founded? Could it be possible that WASSCE results do reasonably
predict the performance of first and second years universities students?
The problem of the study therefore is to find out whether the WASSCE results
do reasonably predict first and second years performances at university level with
particular reference to IMSU and ABSU.
Objectives
of the Study
The objectives of this study were
i.
To ascertain the pattern of relationship between
the WASSCE and first year cumulative grade point averages in mathematics, physics,
chemistry, botany, zoology, statistics and microbiology of students in the relevant
universities.
ii.
To ascertain the pattern of relationship between the
WASSCE and the second year cumulative grade point averages in mathematics, physics,
chemistry, botany, statistics and microbiology of students in the universities.
iii.
To examine the extent to which the independent variable
(WASSCE) contribute to the prediction of the dependent variable (first and second
years GPAs).
iv.
To determine whether
the coefficient of regressions are significant or not.
Research
Questions
1.
What is the simple linear regression coefficient between
0’level examination grade and first year GPA?
2.
What is the simple linear regression coefficient between
0’level examination grade and second year GPA?
Research
Hypotheses
1.
The simple linear regression coefficient between 0’level
certificates grades and first year GPA in the sciences in Imo and Abia state universities
is not significant (P<.05).
2.
The simple linear regression coefficient between 0’level
certificates grades and second year GPA in the sciences in Imo and Abia state universities
is not significant (P<0.05).
Methods
and Procedures
Designs
This is a correlational survey of the extent
of relationship between WASSCE and performances of students in their first and second
years in Imo and Abia state universities. It involves establishing the extent of
relationship between the three variables of interest: the WASSCE, first year GPA
and second year GPA of students’ performances in Imo and Abia state universities. Generally, correlational studies are
undertaken so as to ascertain how far scores or frequencies of observation on two
or more variables go together in some trend or are independent.
In this study, the simple linear regression model was adopted.
Population
The population comprised seven thousand four
hundred and sixty (7,460) students of faculty of science who were admitted in 1997/98–2000/2001
through the university matriculation examination and who completed their first and
second years in the two universities.
Sample
and Sampling Technique
The sample for the study comprised two thousand,
three hundred and eighty (2,380) students purposively sampled from the population.
Description
of the Sample Based on Pattern of Admission
Table 1:
Admission into Faculty of Science, Imo State University Statistics (1997  2001).
First Year
Admission
Departments
Year
Maths
Physics
Chemistry
Botany
Zoology
Statistics
Micro Total
19971998
5
12
28
47
20
8
53 173
19981999
6
18
14
16
26
8
55 143
19992000
13
27
12
18
22
6
52
150
20002001
17
12
6
2
36
8
46 127
TOTAL
41
69
60
83
104
30
206
593
Source: Imo State University
Second Year
Admission
Departments
Year
Maths
Physics
Chemistry
Botany
Zoology
Statistics
Micro
Total
19971998
5
12
28
47
20
8
53 173
19981999
6
18
14
16
26
8
55 143
19992000
13
27
12
18
22
6
52 150
20002001
17
12
6
2
36
8
46 127
TOTAL
41
69
60
83
104
30
206
593
Table 2: Admission into Faculty of Science. Abia State University Admission Statistics
(1997  2001).
First Year
Admission
Departments
Year
Maths
Physics
Chemistry
Botany
Zoology
Statistics
Micro
Total
19971998
4
12
14
14
22
6
49 127
19981999
4
10
12
12
30
4
56 146
19992000
5
4
10
10
24
7
38 189
20002001
6
27
16
16
36
15
55 175
TOTAL
19
53
52
52
112
32
218 597
Second Year
dmission
Departments
Year
Maths
Physics
Chemistry
Botany
Zoology
Statistics
Micro
Total
19971998
4
12
14
20
22
6
49 127
19981999
4
10
12
30
30
4
56 146
19992000
5
4
10
27
24
7
38 189
20002001
6
27
16
20
36
15
55 175
TOTAL 19 53
52
97 112
32
218
597
Source:Abia State University
Instrument
for Data Collection
The instrument for data collection was a
proforma that enabled the researcher copy from the students’ records.
Method
of Data Collection
Grade points of students in WASSCE and first
and second years grade point averages constituted the data that were used for the
study and they were collected from students’ past records in the registry departments,
Imo and Abia state universities. The
researcher also used the university supplied grade point average computation table
to compute the Grade Point Averages where they were not computed.
Data
Analysis
The data were analyzed after abstracting
information from students’ records.
Simple linear regression models involving three variables (two dependent and one
independent variables) were used to carry out the analyses.
The independent variable is the WASSCE scores and the two dependent variables
are the first year GPA (Y_{1}) and second year GPA (Y_{2}).
The t – test was used to test if the linear regression
coefficients were significant.
Results
Research Question 1:
What is the simple linear regression coefficient between 0’level examination grade
and first year GPA?
Table 3: Regression coefficient (a_{1})
and coefficient of determination (r^{2}) between WASSCE results and first
year grade point average.
University
Sample
Regression Coeff. (a_{1})
Coeff of det. (r^{2})
IMSU
593
0.46
0.11
ABSU
579
0.82
0.58
From table 3, the simple regression coefficient
of WASSCE scores and first year GPA is 0.46 in IMSU and 0.82 in ABSU.
The coefficients of determination in IMSU and ABSU are 0.11 and 0.58 respectively. The values obtained imply that the regression
is moderate in IMSU and high in ABSU, this suggesting moderate relationship in IMSU
and a strong relationship in ABSU.
The estimated regression equations are
IMSU:
Y_{1}
= 0.32 +
0.46 x_{1}
ABSU:
Y_{1} = 2.08
+
0.82 x_{1}
(Y_{1} = Year 1 GPA, X_{1}
= West African Senior School Certificate Examination Results).
Research Question 2
What is the simple linear regression coefficient
between 0’level examination grade and second year GPA?
Table 4: Regression Coefficient (a_{1})
and Coefficient of Determination (r^{2}) between WASSCE Scores and Second
Year Grade Point Average.
University
Sample Regression
Coeff. (a_{1})
Coeff. Det. (r^{2})
IMSU
593
0.77
0.33
ABSU
597
0.71
0.24
The results of analysis of data, from table
4, show that the regression coefficient between 0’level examination grade and second
year GPA is 0.77 in IMSU and 0.71 in ABSU.
The coefficients of determination are 0.33 and 0.24 in IMSU and ABSU respectively. The values in both universities are
high, showing high or strong relationship between 0/L grade and second year GPA
in the two universities. The estimated
regression equations are:
IMSU:
Y_{2}
= 1.85 +
0.77 x_{1}
ABSU:
Y_{2} = 1.38
+
0.71 x_{1}
(Y_{2} = Year 2 GPA)
Answers
to Research Hypotheses based on the Results Analysis of Data
Hypothesis 1:
The simple linear regression coefficient between grades 0’level certificates and
first year GPA in sciences in the Imo and Abia state universities is not significant
(P< 0.05).
From the computer print out, the computer
given probability level of t is 0.00 in IMSU and 0.00 in ABSU.
The set probability level of t is 0.05.
The null hypothesis is therefore rejected since the set probability level
of t is greater than the computer given probability level of t.
This implies that there is a significant linear relationship between 0/L
examination grades (x_{1}) and first year GPA (Y_{1}) in both universities.
Hypothesis 2:
The simple linear regression coefficient between grades 0’level certificates and
second year GPA in the sciences in Imo and Abia state universities is not significant
(P < 0.05).
Null hypothesis is also rejected in the two
cases here (IMSU and ABSU). This is
because from the computer print out (showing analyzed data of IMSU) the set probability
value of t (0.05) is greater than the computer given value of t (0.00).
In addition from computer print out (showing analyzed data of ABSU) the set
probability value of t (0.05) is also greater than the computer given value of t
which is 0.00. This therefore implies
that the relationship between 0’level examination grade and second year GPA is quite
significant in the two universities.
Discussion
in Relation to Regression Coefficient Between 0/L Examination Grade and First Year
GPA
From table 3, the simple linear regression
coefficient between 0’level examination grade and first year grade point average
is 0.46 in IMSU and 0.82 in ABSU. These values obtained are moderate in IMSU and
high in ABSU showing apparently moderate to high relationship between the two variables
in both universities. The computed
coefficient of determination in IMSU and ABSU are 0.11 and 0.58 respectively. These show that 11 percent and 58 percent
of the total variation in the GPA in IMSU and ABSU are explained by variations in
0/L examination grade while 89 percent and 42 percent of the variations in GPA in
IMSU and ABSU are attributable to the influence of other factors not explained by
the regression function.
These relationships were shown as significant
in both universities in view of the t – test.
Here, the set probability level of t which is 0.05 is greater than the computer
given probability level of t which is 0.00 in the two universities suggesting therefore,
that 0/L examination grade significantly determined performance in the first year. This implies that most of the students
who performed very well in 0/L examination equally did well in first year examination
in the two universities. That is to
say, that passing the 0/L examination very well is a guarantee that the individual
will pass very well in the first year examination.
This agrees with the findings of Bakari in Ajuonuma 2004 where he found a
significant relationship between 0/L examination grade and their first year academic
performance for school of humanities in Bayero University, Kano.
The reason, according to the researcher, for this significant relationship
may be the similarity in content and method between the 0/L examination and first
year examination. This is supported
by Thorndike’s (1969:558) theory of transfer of learning.
He stated that:
Transfer could occur from one learning situation
to another having common functions and elements.
As the similarities increase between the different situations, the amount
of transfer increase.
In fact, the idea suggests that a student’s
performance in one examination may affect the other provided there is similarity
of content which will make for transfer of learning on the part of the student. The knowledge and skills they acquire
in 0/L are prerequisite knowledge and skills for academic work in the first year
in the university.
Discussion
in Respect of Regression Coefficient between 0/L Examination Grade and Second year
GPA
The results from the data analysis clearing
reveals in table 4 that the regression coefficients between 0/L examination grade
and second year GPA are 0.77 and 0.71
in IMSU and ABSU respectively. These
values are high and show positive and strong relationship between the two variables
in the two universities. The coefficients
of determination from the same table, are 0.32 and 0.24 in IMSU and ABSU respectively. These mean that 32 percent and 24 percent
of the total variations in the GPA in both universities are explained by variations
in the 0/L examination grade while 68 percent and 78 percent of the variations in
GPA in the two universities are attributable to the influence of other factors not
explained by the regression function.
From the computer print out, the computer
given probability levels of it are 0.00 in both universities and this value is less
than the set probability level of t which is 0.05.
The null hypothesis was therefore rejected and the alternative accepted which
implies that the relationship between 0/L exam grades and second year GPA is quite
significant. This is in line with the
findings of Soyibo (2004) and Somola (2003) who found a significant relationship
between 0’level examination grade and second year grade print average at university
of Ilorin.
Implications
of the Research Findings
From the results obtained, those who do well
in WASSCE tend to do well in Imo and Abia state universities.
One can, therefore use WASSCE to screen and admit students into Imo and Abia
state universities. The results of this study indicate fruitfulness in the curriculum
planning. Planners of senior secondary
school and Imo and Abia State universities curriculum should ensure that there is
some link.
Conclusion
Despite the views of people against the findings
of this study, the result of this research shows that 1996/97 – 2000/01 university
selection examinations in IMSU and ABSU were effective in that it succeeded in identifying
in that level of accuracy, students who would perform well in first and second years
sessional examinations in Imo and Abia state universities programmes.
This inference is drawn based on the fact that the regression coefficient
obtained were high and the simple regression coefficient when subjected to statistical
test using t – test showed high level of significance.
Recommendations
On the basis of the findings of this research,
the researcher makes the following recommendations:
1.
Universities should continue to use WASSCE results for
the placement of students into Imo and Abia state universities.
2.
The Government should try and curb malpractice to ensure
validity and reliability of the scores obtained in WASSCE so that candidates who
should have scored low grades end up scoring higher grades with low performance
in the universities.
3.
Civility of the WAEC should be enhanced to ensure that
the circumstance in which the testing has taken place reflects a normal teaching
/ learning atmosphere in which assessment is integral to, not external to learning
with aliens.
4.
WAEC should ensure credibility of the results to the
extent that the testees should regard the results obtained from the test as reflecting
actual competence in performance and excellence.
References
Adekunle, C. (2003).
Crisis and Problems in Education in Nigeria. Ibadan: University Press.
Ajuonuma, J. O. (2004).
Predictive Validity of Prerequisite Qualifications for Academic Performance in
the Universities. Unpublished Degree Thesis, Abia State University Uturu.
Nwana, O.C. (2000). Evaluation of school
and College Examinations in the Nigerian Education System.In Fafunwa, A.B.,(ed)
Educational Foundation. Ibadan: Wemilore
Press (Nig) Ltd.
Nkwocha, P.C. (2002). A study of the Psychosocial
Correlates of Examination Malpractice among Adolescents in Imo State. An Unpublished
Ph.D Thesis, Abia State University Uturu.
Somala, J. L. (2003).
The Predictive Validity of Prerequisite Qualifications and Second Year University
Performance. Unpublished Degree Thesis,
University of Lagos.
Soyibo, M. N (2004).
The study of the Relationship between Entry Qualifications and Academic Performance
in the University. An Unpublished Masters
Degree Thesis, University of Lagos.
Thorndike R. L. and Hagen, E. P. (1969).
Measurement
for Social Reconstruction in Education.
New York: Macmillan.

