The concept of validity

It refers to quality which could be applied to any aspect pertaining to research process. In regard to procedures of measurement it implies to whether research instrument measures what it is suppose to measure. Basically there are two approaches through which validity of an instrument could be checked: a logical link established between the objectives of the study undertaken and the instrument answering established set of questions, using therefore statistical analysis for the purpose of demonstrating the link.

 Validity could be of three types:

  • Face and content
  • Concurrent and predictive
  • Constructive validity

The ability of producing consistent measurements every time refers to the reliability of an instrument. When an instrument is administered under  similar conditions obtaining same results, it is regarded as the ‘reliable’ instrument- the reliability increases with increase in the similarity of results. Reliability can be looked from two sides:

  • Reliability or the extent of accuracy
  • Unreliability or the extent of inaccuracy

If the wording of questions is ambiguous, there is a change in the physical setting for the purpose of data collection, mood of respondent’s while providing information, interviewer and interviewee’s interaction, and the instrument’s regressive effect are few factors that affect the research instrument’s reliability.

Procedures for determining reliability are external and internal consistency. Parallel forms and test/re-test of the same test form the two procedures which would determine the research instrument’s external reliability. The internal consistency involves the split-half technique.

%d bloggers like this: