Assessment engine for mathematical expressions / Arsmah Ibrahim, Zainab Abu Bakar and Nuru’l-‘Izzah Othman

In this research an assessment engine that emulates human marking processes in checking the correctness of student-constructed responses in a free-response mathematics assessment setting. The assessment engine is incorporated with the computational techniques that were constructed in this research....

Full description

Saved in:
Bibliographic Details
Main Authors: Ibrahim, Arsmah, Abu Bakar, Zainab, Othman, Nuru’l-‘Izzah
Format: Research Reports
Language:English
Published: 2010
Subjects:
Online Access:https://ir.uitm.edu.my/id/eprint/32502/1/32502.pdf
https://ir.uitm.edu.my/id/eprint/32502/
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this research an assessment engine that emulates human marking processes in checking the correctness of student-constructed responses in a free-response mathematics assessment setting. The assessment engine is incorporated with the computational techniques that were constructed in this research. The techniques have succeeded in enabling the engine to implement step-by-step correctness checking of constructed responses in a scheme of worked solutions and to “award” scores that indicate the degree of correctness of each step. The domain of the mathematical problem chosen for this research involved solving linear algebraic equations in one variable. In providing the process framework for the construction the computational techniques, basic processes and methods from the field of textual information retrieval have been adapted. In providing the method for the determination of the degree of correctness, the method of exact pattern matching and approximate string matching were adapted. Nine questions on solving linear algebraic equations in one variable which are of different form and level of difficulty were used the testing of the precision of the automated marking implement by the engine. The scores indicating the degree of correctness evaluated by the assessment engine is benchmarked against the scores