Reliability -- consistency, trustworthiness, getting the same answer each time.

 

Major types:

 

1. Test-Retest -- The correlation of scores of a test that has been administered twice to a group of subjects.  Example see Shaw and Wright The custodial mental illness ideology scale. pages 108-111.

 

2. Alternate Form -- The correlation between scores of two distributions obtained on the administration of two tests that have been constructed as having the identical content.  Example see Shaw and Wright Attitudes toward feminism belief patterns scale page 278-287.

 

3.  Split-half Reliability -- From a single administration, a test is divided into two equal parts.  A correlation is calculated for the two halves.  Some common method for scales used in social work practice.  Example see Shaw and Wright  Opinionaire on attitudes toward education  pages 80-83.

 

4.  Interitem or Internal Consistency -- Assesses how well the different items measure the same issue of concept.  The mean of all possible split-half reliabilities.  Difficult to use without a computer.  Commonly used today.   Example see Corcoran and Fischer Hypercompetitive attitude scale. Pages 353 in volume II.  Two types are commonly found: Cronbach's alpha coefficient is used for continuous data and Kuder-Richardson (and variations of it) are used for discrete data.

 

5.   Scorer Reliability -- The correlation of two scores obtained by independent examiners.  Example see DSM I, II, III, III-R, IV, etc.

 

 

 

Validity -- What the test measures, how well it does so and that it does not measure something other than the measurement target.

 

1.    Content Validity -- The systematic examination of the test content to determine whether it covers a representative sample of the behavior domain to be measured.

 

Type: Face validity, item analysis

 

2.    Criterion-Related Validity -- the effectiveness of a test in predicting an individual’s behavior in specific situations.  Example does passing the state exam mean that one is going to be a good social worker.

 

Types: Concurrent, Predictive, synthetic validity

 

3.    Construct Validity -- The extent to which a test may be said to measure a theoretical construct or trait. 

 

Types: correlations with other tests factor analysis, internal consistency, and convergent and discriminate validation

 

 References