E-Test – from Writing Test Items to Statistical Validation

Abstract

Education, assessment, and learning are interconnected and influential components of the learning process. As technology continues to advance, electronic assessment methods have evolved to reliably evaluate knowledge and skills across the cognitive scale. While online test examinations offer clear benefits, such as convenience and objectivity, there are challenges in creating and programming the tests in electronic systems. Developing a valid and reliable electronic test requires specific knowledge and skills.

This paper aims to provide a comprehensive overview of the process involved in creating a valid and reliable electronic didactic test. It summarizes the requirements and recommendations for crafting effective test questions, and explores the development of a test specification, which involves distributing questions based on learning units and cognitive levels. Additionally, the paper discusses the contrast groups method for establishing the cut-off point.

It is important to view the test as a comprehensive assessment of learning objectives at various cognitive levels, rather than a mechanical collection of exam questions. The qualities of the educational test, particularly its objectivity, are determined by statistical indicators. The paper also delves into various methods for determining the test's validity and reliability.