Simplified Reliable Online Essay Test Marking for Massive Open Online Course (MOOC) using Rasch Model Analysis

Abstract

Manual practice in formal examination does not assess accurate measure
of a student’s ability, as it merely counts the score of every question to be
considered for the student’s grade. There are many educators who have
used raw score as a form of measurement for a student’s ability, but it never
truly measures the right measurement. The raw score should be converted
into the right linear metrics for ability measurement. This procedure
contains measuring score of accurate student’s ability in LOGIT unit,
providing of student’s result profile, and measuring reliability of the test
set and the student’s answers. The procedure is designed for massive open
online learning and paperless essay-based test which is more difficult to
be analysed. This procedure converts the student’s answer into rubrical
ratio-based scale to be more accurately measured. It is definitely better than
the common practice of merely analysis on raw marks for each question.
It would show true student’s performance of cognitive performance (test)
which represents the true student’s ability (in LOGIT unit), in order to
accurately measure the right outcome. This new paradigm of assessment
is fit to be applied for massive numbers on online students. It uses Rasch
model which offers reliable solution in producing accurate ability marks
for students, together with scientific reliability score for student’s answer.