Content Validity and Reliability of Single Items or Questionnaires

Abstract

Procedures for computing content validity ( V) and consistency re liability ( R) coefficients and determining the statistical significance of these coefficients are described. These procedures, which employ the multinomial probability distribution for small samples and nor mal curve probability estimates for large samples, can be used in a variety of situations where judgments of the content validity of items or questionnaires are made on ordinal rating scales. Computing for mulas for determining the statistical significance of V and R for large samples of raters and any number of rating categories are given. A computer program has been written to determine the right- tail probabilities associated with values of V and R obtained from ratings by N raters using c rating categories.