Options
Estimating achievement in mathematics using computer administered and paper-and-pencil tests
Supervisor
Lam, Tit Loong
Lam, Peter Tit Loong
Abstract
This study is an attempt to estimate the change in proficiency levels of students who went through an enrichment program in Additional Mathematics. The test items used were classified by hierarchies modeled after the taxonomy, the Structure of the Observed Learning Outcome, often referred to as SOLO taxonomy. The estimation was done using a four-group pretest-treatment-posttest counter-balance design. The tests were administered in two modes, the paper-and-pencil mode and the computer mode so as to establish comparability of modes of testing.
The enrichment program was shown to be definitely effective where students progressed not only in mastery of content but also in the ability to solve higher level problems. By classifying test items into `SOLO' hierarchies, it enabled identification of types of learners among those who attempted these items. This is most desirable for making professional judgement of the next step in any instructional process.
Test models seemed to have no effect on students' group performances. However, this study had not included in its scope a study on biases that the computer mode may have on certain individuals, if any, or if significant.
Computers may have been shown as a viable testing tool. However, this study is by no means conclusive. It only goes as far as to underscore the need to determine repeatedly and empirically the equivalence of computer and paper versions of examinations. An offshoot from the use of computer technology is a discussion on the applications of the Item Response Theory, or IRT in short, for item analysis as well as for student evaluation within the polytechnic.
The enrichment program was shown to be definitely effective where students progressed not only in mastery of content but also in the ability to solve higher level problems. By classifying test items into `SOLO' hierarchies, it enabled identification of types of learners among those who attempted these items. This is most desirable for making professional judgement of the next step in any instructional process.
Test models seemed to have no effect on students' group performances. However, this study had not included in its scope a study on biases that the computer mode may have on certain individuals, if any, or if significant.
Computers may have been shown as a viable testing tool. However, this study is by no means conclusive. It only goes as far as to underscore the need to determine repeatedly and empirically the equivalence of computer and paper versions of examinations. An offshoot from the use of computer technology is a discussion on the applications of the Item Response Theory, or IRT in short, for item analysis as well as for student evaluation within the polytechnic.
Date Issued
1997
Call Number
QA11 Tha
Date Submitted
1997