Abstract:
Assessment of academic performance in whatever form, remains the only practical
and veritable means of gauging students' learning outcomes. Assessment as it is used
today using numerical and statistical scores have many shortcomings like: uneven
scores awarded by different examiners, test area covering a sample of the syllabus,
tests having measurement error, examinations being criterion referenced and not norm
referenced and done within a time limit. The aim of the study was to investigate how
comments are given by lecturers at Kaiboi Technical Training Institute with a view of
developing a system that uses linguistic and numerical variables to measure students‟
performance. The objectives of the study were to: identify the current academic
performance comments given by lecturers to classifying students‟ academic
performance so as to model the data to develop a membership function, use fuzzy
logic to model the membership function generated to get student‟s numeric grade,
model fuzzy sets using a suitable membership function to determine a student‟s grade
in linguistic variables, develop a system that uses fuzzy logic to measure student‟s
performance. The study was informed by fuzzy logic theoretical framework which is a
multivalued logic that allows intermediate values to be defined between conventional
evaluations like true/false, yes/no, high/low, good/bad. The study used case study
research design. A document review was used to get comments and numerical scores
from students‟ records, the data was modelled to a Trapezoidal membership function
which was used to get degree of membership and crisp values. Prototyping
methodology was used to model and develop the system. Previous students‟ records
were used as training data. The system developed calculates degree of membership
that incorporates marks and comments in students‟ final grade and gives comments
based on the student‟s performance. The output from the system showed variation in
the students marks when comments are incorporated. The study has incorporated
comments in measuring students‟ performance and is therefore an addition to the
current knowledge on students‟ performance measurement