Asia Market Research Home > Market Research Terminology Index > Reliability



Apart from it's intuitive meaning, theoretical reliability refers to the extent that a measure of a concept, whether it be a product attribute or a concept such as ad recall or brand recognition, would deliver the exact same results no matter how many times it is was applied to random members of the same target group. In the academic setting, reliability estimates of a question or set of questions that together are posited to be a measure of a certain concept can take several forms including test-retest and split-half reliability testing.

In the commercial sphere and in practical application, these techniques can also be used when appropriate, but more often reliability is addressed by making sure research and investigative questions are consistent in meaning across all members of a targeted group, delivered objectively and pre-tested as such.

Reliability can be differentiated from Validity which measures the extent to which the research is really measuring what it says it is.

A key question in research is the reliability of results. After all to a large extent that is what a client is looking for - independent, objective results and analysis that reflects reality. Most reputable agencies will provide raw data of all research findings, plus a clear precise description of the research methodology from sample selection, screening, questionnaires and other research aids used. The principle is that another market research agency should be able to reproduce the methodology exactly. If the research was reliable (and given that attitudes have not changed in the meantime!) the results should be substantively similar.

AMR Recommends...
Recommended by Asia Market Research dot com : Sponsorship info

Market Research Terminology | Articles | Columns | Country Profiles
Asian Market Research News Headlines | Search

Asia Market Research