Research Reliability and Validity
Part 2. Marketing Research
Published Online: 15 DEC 2010
Copyright © 2011 John Wiley & Sons, Ltd. All rights reserved.
Wiley International Encyclopedia of Marketing
How to Cite
Du, R. Y. 2010. Research Reliability and Validity. Wiley International Encyclopedia of Marketing. 2.
- Published Online: 15 DEC 2010
This article discusses two important and interrelated concepts in marketing research: reliability and validity. A reliable research instrument should demonstrate consistency, that is, leading to the same observation/conclusion each time it is used with the same subjects under the same condition. There are many operational approaches to assess reliability, including internal consistency (e.g., average interitem correlation, average item-total correlation, split-half, Cronbach's alpha), parallel-forms, test–retest, and interrater, all of which involve some sort of repetition. A valid research instrument should capture what it is supposed to examine. Major types of validity include face, content, criterion (concurrent and predictive), and construct (convergent, discriminant and nomological). Assessing reliability is essentially about gauging the size of random error. Assessing validity is essentially about gauging the existence of systematic error. Perfect reliability is achieved when there is no random error. Perfect validity is achieved when there is neither random nor systematic error. In practice, it is often easier to establish reliability than validity. However, it is important to keep in mind that reliability is a necessary, but not sufficient, condition for validity.
- internal-consistency reliability;
- parallel-forms reliability;
- test–retest reliability;
- interrater reliability;
- face validity;
- content validity;
- criterion validity;
- construct validity