Abstract:
Purpose: With improvements in computer
technologies and test implementations in the
computer environment, when advantageous points
of computer-based test implementations are
considered, it is inevitable to compare psychometric
characteristics of paper-and-pencil tests and
computer-based tests and students’ success. In
computer-based tests, individuals’ familiarity with
computers and competency in using computers,
conditions may show diversity depending on the
country or region. Therefore, this study aims to
investigate the effects of mean differences between
PP and CBT using meta-analysis concerning the
studies, including samples from Turkey and
conducted between 1993 and 2020.
Received: 25 Jan. 2020
Received in revised form: 14 Feb. 2021
Accepted: 12 Mar. 2021
DOI: 10.14689/ejer.2021.93.13
Keywords
computer-based test, paper-pencil test,
effect size, meta-analysis
Research Methods: In this meta-analysis, 37 findings were included. Cohen’s d was used as
the effect size. And also, in this study, concerning the equivalence of the PP and CBT forms, it
was investigated whether mean effect sizes differ or not according to variables like type of
computerized, education level, and subject matter. In this direction, ANOVA and Q values
were used.
Findings: As a result of the meta-analysis conducted, the general effect size was 0.042. In this
direction, it was found that the difference in test implementation methods (paper-and-pencil,
computer-based) was negligible.
Implications for Research and Practice: Results suggest that CBT can be an acceptable
alternative to traditional pencil and paper tests. In this way, results obtained are expected to
lead to educational policies and measurement implementations in the future.