General information about tests

The term "test" has several meanings: In relation to the evaluation of a learning program, tests can be understood as more or less standardised procedures for the measurement of the behaviour or performance of people. A test provides quantitative data, for example, a score in a quiz.

A test consists of various items which can be questions with open, semi-open or closed answers (cf. information on Questionnaires page).

Tests are developed either ad hoc or based on a complicated procedure. The former are called informal tests. They are non-standardised tests to which quality criteria (objectivity, reliability and validity) are not relevant. They are developed by the evaluators ad hoc to meet the requirements of the respective evaluation goals. Standardised tests, on the other hand, should meet the quality criteria (objectivity, reliability and validity). There are two types of standardised tests which are based on evaluating the test performance at different reference points:

• Norm-oriented tests: In norm-oriented tests, the individual test performance is compared with the average performance of a reference group.
• Criterion-oriented tests: Criterion-oriented tests are used to measure individual performance against a predefined criterion (e.g. a certain score). These tests are used, for example, to measure differences in learning achievement after an educational program.
  

Situational tests are another form of test, these involve sample work, role plays, case studies, group discussions, planning games, and simulations. These tests provide data on procedures regarding the accomplishment  of practical tasks.

The decision in favour of a certain type of test depends on the goals and conditions of the test and on financial, staff and time resources (cf. Tergan, 2000)

Required resources
The development of standardised tests and the conducting and analysis of  situational tests in particular require considerable resources.

Advantages
One advantage of tests is the comparability of results (between people, measurement times, etc.).

Disadvantages
Broad knowledge in testing theory is necessary to develop a standardised test (cf. e.g. Bortz & Döring, 2003).