Methods of Assessing Reliability

1. Test-Retest - the same test is given 2 times, with a time interval between settings. The coefficient measured by this is stability. Memory effect, practice effect, and change over time are problems related with this procedure.

2. Alternate form - equivalent forms of the same test are given, with time between testing. This measures both equivalence between the two forms and stability. One of the problems associated with this form is that it may be hard to develop two equivalent forms of the test and changes in behavior over time (between the two tests) may affect the reliability.

3. Internal consistency or Split-half method - one test is given one time, by splitting the test into two parts. The correlation between the two halves is calculated. This measures internal consistency and equivalence. Due to splitting the test in half, it becomes necessary to see if this affects the reliability. This is achieved by applying the Spearman-Brown Formula which identifies the effect of shortening the length of a test on its reliability.

4. Inter-rater reliability - this depends on the rater's judgment...examples: an essay test, behavioral observational scale, or projective personality assessments. The correlation between two or more raters is found out. Sources of problems could include rater problems such as lack of motivation, rater bias, and characteristics of the measuring device.