Professional Documents
Culture Documents
net/publication/308169118
CITATIONS READS
2 5,848
2 authors, including:
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Oluwafemi Samson Balogun on 28 September 2016.
56
Covenant Journal of Informatics and Communication Technology (CJICT) Vol. 4, No. 1, June, 2016
admission processes are conducted above the upper control limit, which
without duplication of admission. indicate the positive shift in student
The information used for this CGPA was identified. Then, a
research was obtained by the solution to correct students’ poor
method of transcription from performance and was suggested. If
records and they are all secondary the charting statistics for all the
type of data. Data were collected semester fall within the control
from eight departments in Faculty of limits, the student has maintained
Science: Biochemistry, Chemistry, the desire target GPA value.
Physics, Geology, Computer According to (Balogun et al., 2014)
Science, Mathematics, Statistics and the main focus of their research is to
Microbiology. The data include develop models that can be used to
students’ performance measured by study the trend of graduate
their GPA and CGPA as emigration in Nigeria using log-
appropriate, in both their first and linear modeling based on the results
final year. from of likelihood ratio (G2),
The aim of this research is to use Akaike Information Criteria (AIC) ,
Cohen’s Kappa to study students’ Saturated model has a perfect fit for
performance of some departments in modeling graduate emigration in
the Faculty of Science, University of Nigeria. This implies all the three
Ilorin and objectives are: to know factors involved (discipline, year
the strength of agreement that exist and sex) has to be included in the
between student grade point in both model in order to have an
their first and final year, to know the appropriate result.
proportion of students that Since its introduction Kappa
maintained their CGPA (i.e. those statistics, several authors have
that maintained what they started applied the concept in different
with), to know the proportion of field, for instance; (Zeeshan et al.,
students that dropped from the class 2015) carried out an initial audit for
of grade point they started with and evaluating the case notes for each
to know the proportion of students team against the TONK score. In
that improved on their performance. order to evaluate the producibility of
this score, the Cohen’s kappa was
In a study conducted by (Akinrefon
used and substantial agreement was
& Balogun, 2014), control chart was
noted. The article by (Viera &
used to monitor students’
Garret, 2015) provided a basic
performances, the causes underlying
overview of Kappa statistics as one
the charting statistics that are less
measure of inter-observer
than the lower control limits were
agreement. They concluded that
identified which indicate a negative
“Kappa is affected by prevalence
shift in students CGPA. Also, the
but nonetheless kappa can provide
reason for charting statistics falling
more information than a simple
57
Covenant Journal of Informatics and Communication Technology (CJICT) Vol. 4, No. 1, June, 2016
58
Covenant Journal of Informatics and Communication Technology (CJICT) Vol. 4, No. 1, June, 2016
59
Covenant Journal of Informatics and Communication Technology (CJICT) Vol. 4, No. 1, June, 2016
n 1 e
2
marginal distributions becomes
more nearly equivalent. The effect
of rater bias on kappa has been Assuming that
investigated by (Feinstein &
Ciccheti 1990) and (Bryt et al., K
……………………. (5)
1993). Var0 K
Early approaches to this problem
have focused on the observed Follows a normal distribution, we
proportion of agreement; see can test the hypothesis of the chance
(Goodman & Kruskal, 1954), this
60
Covenant Journal of Informatics and Communication Technology (CJICT) Vol. 4, No. 1, June, 2016
61
Covenant Journal of Informatics and Communication Technology (CJICT) Vol. 4, No. 1, June, 2016
maintain their grade point, 44.89% their grade point, 31.46% of the
of the students improved while students improved while 5.56%
6.12dropped from the class of grade dropped from the class of grade
point they started with. point they started with.
For Mathematics Department, 36% 4.2.1 Summary
of the students were able to maintain From the above interpretation, we
their grade point, 64% of the could see that 54.42% of the
students improved while none students were able to maintain their
dropped from the class of grade CGPA that they started with,
point they started with. 34.39% of the students improved
For Geology Department, 46.15% of and 11.18% of the students dropped
the students were able to maintain from the class of grade point they
their grade point, 49.23% of the started with. Also, the strength of
students improved while 4.62% agreement between the first and the
dropped from the class of grade final year result is on the 0.40%.
point they started with.
4.3 Conclusion
For Computer Science Department,
It can be observed that Mathematics
51.49% of the students were able to
department has the highest number
maintain their grade point, 1.79% of
of students that improved on their
the students improved while 46.71%
performance, Statistics department
dropped from the class of grade
had the highest number of students
point they started with.
that maintained their grade point
For Biochemistry Department,
and Computer Science department
60.17% of the students were able to
had the highest number of students
maintain their grade point, 29.66%
that dropped from the grade point
of the students improved while
they started with. Also, the strength
10.17% dropped from the class of
of agreement that exist between the
grade point they started with.
first and the final year result is on
For Chemistry Department, 62.96% Average, that is, “fair”.
of the students were able to maintain
62
Covenant Journal of Informatics and Communication Technology (CJICT) Vol. 4, No. 1, June, 2016
63
Covenant Journal of Informatics and Communication Technology (CJICT) Vol. 4, No. 1, June, 2016
Appendix
Table 1: Example of Raters
Subject Rater 1 Rater 2 Rater 3
1 5 6 2
2 3 4 2
3 1 2 1
64
Covenant Journal of Informatics and Communication Technology (CJICT) Vol. 4, No. 1, June, 2016
65
View publication stats