Types of validity in quantitative research

There are three types of validity in quantitative research:

1. face and content validity;
2. concurrent and predictive validity;
3. construct validity.

1. Face and content validity

The judgement that an instrument is measuring what it is supposed to is primarily based upon the logical link between the questions and the objectives of the study. Hence, one of the main advantages of this type of validity is that it is easy to apply. Each question or item on the research instrument must have a logical link with an objective. Establishment of this link is called face validity. It is equally important that the items and questions cover the full range of the issue or attitude being measured. Assessment of the items of an instrument in this respect is called content validity. In addition, the coverage of the issue or attitude should be balanced; that is, each aspect should have similar and adequate representation in the questions or items. Content validity is also judged on the basis of the extent to which statements or questions represent the issue they are supposed to measure, as judged by you as a researcher, your readership and experts in the field. Although it is easy to present logical arguments to establish validity, there are certain problems:

• The judgement is based upon subjective logic; hence, no definite conclusions can be drawn. Different people may have different opinions about the face and content validity of an instrument.
• The extent to which questions reflect the objectives of a study may differ. If the researcher substitutes one question for another, the magnitude of the link may be altered. hence, the validity or its extent may vary with the questions selected for an instrument.

2. Concurrent and predictive validity

‘In situations where a scale is developed as an indicator of some observable criterion, the scale’s validity can be investigated by seeing how good an indicator it is’ (Moser & Kalton 1989: 356). Suppose you develop an instrument to determine the suitability of applicants for a profession. The instrument’s validity might be determined by comparing it with another assessment, for example by a psychologist, or with a future observation of how well these applicants have done in the job. If both assessments are similar, the instrument used to make the assessment at the time of selection is assumed to have higher validity. These types of comparisons establish two types of validity: predictive validity and concurrent validity. Predictive validity is judged by the degree to which an instrument can forecast an outcome. Concurrent validity is judged by how well an instrument compares with a second assessment concurrently done: ‘It is usually possible to express predictive validity in terms of the cor­relation coefficient between the predicted status and the criterion. Such a coefficient is called a validity coefficient’ (Burns 1997: 220).

3. Construct validity

Construct validity is a more sophisticated technique for establishing the validity of an instrument. It is based upon statistical procedures. It is determined by ascertaining the contri­bution of each construct to the total variance observed in a phenomenon.

Suppose you are interested in carrying out a study to find the degree of job satisfac­tion among the employees of an organisation. You consider status, the nature of the job and remuneration as the three most important factors indicative of job satisfaction, and construct questions to ascertain the degree to which people consider each factor important for job satisfaction. After the pre-test or data analysis you use statistical procedures to establish the contribution of each construct (status, the nature of the job and remuneration) to the total variance (job satisfaction). The contribution of these factors to the total variance is an indica­tion of the degree of validity of the instrument. The greater the variance attributable to the constructs, the higher the validity of the instrument.

One of the main disadvantages of construct validity is that you need to know about the required statistical procedures.

Source: Kumar Ranjit (2012), Research methodology: a step-by-step guide for beginners, SAGE Publications Ltd; Third edition.