Student Performance Predictive Model Mitigating Students' Performance Gap In Outcome-Based Education Systems Using Mathematical Model A Case Study

Outcomc-Based Education (OBE) model is a recurring modern means for education
reforrn - a process of improving public education. It embodies the idea that best
educational practice is to determine the end goals, or "outcomes", before the
strategies, processes, techniques, and other nreans can be put into place to achieve
them. While applications of OBE model have been continuously expanding and
improving, "performanoe gap" - the gap between what students can do and what they
are expected to do - still hinders its potential benefits. Mitigating this gap is among
priority tasks of educators to achieve long-term goals of educational reform; and
developing student performance predictive models is one way to approach this
problem.
Most previous studies had targeted big scope of a long-term prediction and most had
used various range of educational settings as their inputs, including students'
demographic profiles and behavioral contents. They had applied diff[erent techniques
in order to predict students' academic performance; however, due to the nattrre of
these inputs, all had adopted complex data mining models. This project, instead, was
purposely narrowed down to short-term programming cowses at Universiti
Teknologi PETRONAS (UTP), Malaysia Main purpose was to design a finctioning
short-term predictive model which continuously assised lecturers to analpe patterns
and to accurately predict students' upcoming perforrnance and final rezuh in order to
provide timely intervention and adjustment. The writer introduced a unique approach
by focusing on a simplified set of rnputs including (1) students' courtework
breakdown and (2) users' dynamically subjective inputs. Instead of conplex data
mining modcls, a straightforward mathematical model was developed and y65 highly
customized to best utilize those inputs, which resulted in a high level of accuracy for
predictive outputs. A fully developed system from the testrng protot)?e promises to
s€rve as a relatively convenient tool for UTP lecturers to rnilize simple yet richly
inforrnative coursework data into predicting students' performance, then mitigating
the performalrce gap and ultimately achieving set objectives of UTP's OBE systenr.