Research Tips
Reliability
The word “reliability” has a more specific meaning in research than it does in everyday usage. When researchers talk about reliability, what they mean is “the extent to which an experiment, test, or any other measuring procedure yields the same results on repeated trials”.1 Researchers are concerned with reliability because if a measurement instrument, such as a questionnaire, is unreliable then there can be no confidence in the results generated.
Several different techniques can be used to measure reliability. Most rely on examination of correlations between responses. One technique involves examining correlations between answers to specific questions that are expected to result in similar responses. Another involves repeating the original research using the same subjects and the same research instruments and then checking the correlations between the two sets of results.
1 Carmines, Edward G. and Richard A. Zeller (1991), Reliability and Validity Assessment, page 11.