Interitem Reliability: Measuring the Consistency of Test Results
Interitem reliability (IIR) is a measure of consistency in test results and a key indicator of a test’s reliability. It is also sometimes referred to as item-level reliability or item homogeneity. IIR is used to assess the internal consistency of a test or survey, to examine the degree to which different items on a test measure the same construct, and to identify items that are inconsistent with the others. By measuring the degree of consistency between items, IIR helps to identify and eliminate items that may be reducing the reliability of a test or survey.
To calculate IIR, the correlations between all pairs of items on a test are calculated. The average correlation is then used to estimate the reliability of the test. The higher the average correlation, the higher the IIR. A high IIR indicates that the items on a test are measuring the same construct.
IIR is one of several indicators of a test’s reliability. Other measures, such as Cronbach’s alpha, are also used to assess reliability. Cronbach’s alpha is the most commonly used measure of reliability, but IIR is sometimes used as well. IIR has the advantage of providing more detailed information about the consistency of items on a test.
There are several ways to improve IIR. One way is to reduce the number of items on a test, as this will increase the average correlation between items. Another way is to ensure that items are measuring the same construct and that they are worded in a consistent manner. This can be done by pre-testing items and conducting factor analysis. Finally, items that are too difficult or too easy should be eliminated, as these items may be reducing the IIR.
In summary, Interitem Reliability is an important measure of a test’s reliability. It is used to assess the internal consistency of a test and to identify items that may be reducing the reliability of the test. There are several ways to improve IIR, such as reducing the number of items on the test, ensuring that items are measuring the same construct, and eliminating items that are too difficult or too easy.
References
DeVellis, R. F. (2012). Scale development: Theory and applications (3rd ed.). Thousand Oaks, CA: Sage.
Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York, NY: McGraw-Hill.
Streiner, D. L. (2003). Starting at the beginning: An introduction to coefficient alpha and internal consistency. Journal of Personality Assessment, 80(1), 99-103. doi: 10.1207/S15327752JPA8001_18