Skip to main content icon/video/no-internet

Validity is a central concept in social science research. At the broadest level, validity refers to the extent to which a claim, result, inference, or argument is well founded. In the social sciences, the term validity is often (but not exclusively) used in reference to educational and psychological measurement and assessment, where it is frequently referred to as the most fundamental consideration in developing and evaluating tests. However, despite wide agreement regarding its importance, there is no single conception of validity universally accepted in the scholarly and professional communities, and there remains considerable controversy surrounding the definition of validity and many related concepts and terms.

This entry introduces the topic of validity, concentrating on its applications in measurement and assessment. The entry begins with an overview of basic concepts and terminology, followed by sections describing perspectives on validity and validation roughly following a historical progression. Early perspectives on validity are described first, including the concepts of criterion- and content-related forms of validity. The next section discusses construct validity, as first introduced by Lee J. Cronbach and Paul E. Meehl in the mid-1950s, and the idea of nomological networks. Following this are discussions of the unified perspective on validity due primarily to Samuel S. Messick and the interpretive argument-based approach to validation due primarily to Michael T. Kane. The final section discusses a causal perspective on validity due primarily to Denny Borsboom.

Although this entry does not aim to provide a thorough historical overview of thinking about validity, this historical presentation may nonetheless be helpful in contextualizing the origins of many common ways of thinking about validity, especially insofar as each of these perspectives remain influential, to varying degrees, in different areas of contemporary social scientific scholarship. This entry does not aim to provide an introduction to how validation activities do or should take shape in any given application.

Basic Concepts in Validity

One of the earliest proposed conceptions of validity (and one that remains popular among many scholars and practitioners) is that validity refers to the extent to which a test measures what it claims to measure. Validity is often introduced alongside the concept of reliability, which refers to the extent to which the results of an assessment are free from random sources of measurement error; in other words, the higher the reliability, the less “noise” there will be in the measurement process. Seen this way, validity refers to the accuracy of an assessment, whereas reliability refers to its consistency. A classic pictorial analogy (Figure 1) helps make this concept intuitive: A test that is both valid (accurate) and reliable (consistent) is analogous to a situation in which the bull’s-eye of a target is struck regularly. If a test is reliable but not valid, the results will be consistent but consistently off the mark; if it is valid but has lower reliability, the scatter around the bull’s-eye will be greater, but on average, the shots will be in the right location. This image also helps motivate the intuition that at least some degree of reliability is a precondition for validity: If all of the observed variation in the results of an assessment were due to random measurement error (i.e., zero reliability), it is difficult to imagine how one could claim that the assessment’s results have validity.

...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Sage Recommends

We found other relevant content for you on other Sage platforms.

Loading