Models Represent; Theories Explain; Assessments Measure
(Click here to find out about the history of assessments)

 

How does the Platinum Rule "Measure Up"?

Reliability of very acceptable levels was achieved for both the Openness scale (.7527; see table 1) and for the Directness scale (.7954; see table 2). These reliability coefficients are well above the.7 level of reliability generally considered as acceptable and just short of the .8 ideal standard for reliability as defined in the Standards for Educational and Psychological Testing published by the American Psychological Association.

 

The construct validity studies, showing a two factor structure, were not unexpected given the wide body of research specifically from Merrill to Stogdill.

The Science of Assessments

There are three things to consider to determine psychometric rigor: reliably, validity and social desirability.

Reliability (PRECISION) determines how well that the items on a scale accurately reflect the scale itself.   

Reliability:
In laymen’s terms: does it actually give the same results if taken by the same person on different occasions?

Test Retest Reliability – Stability

This refers to how well an instrument yields consistent results. Scores range between 000 and 1.00. A reliability index of 0.85 or higher is generally considered to be effective for all purposes. An index of between .65 and .85 indicates a good range for groups but is considered less confident in an individual sore.

Validity (ACCURACY) determines the extend to which the association among the scores represents the theory and model on which the instrument is based.

Validity:
In laymen’s terms: does it actually measure what it says it measures?

Validity establishes the confidence with which we can interpret any given result on a given test. Validity is obviously a much more complex question than reliability, and is ultimately a more important question. Beware, there are at least five different types of validity and often the one that is claimed is that it appears to have face validity.

  • Content Validity or Face Validity – a person’s perception, usually from the appearance of the instrument, that it is likely to measure something meaningful.
  • Predictive Validity – measures what it claims to measure
  • Concurrent Validity – relationship to some other concrete criteria
  • Convergent Validity - results on individual instruments are compared to results obtained on other instruments that measure abilities in the same ability domain.
  • Divergent Validity - results on individual instruments are compared to results obtained on other instruments that measure abilities in different ability domains.

Social Desirability (TRANSPARENCY) determines how easy it is to fake the results.

 		 
     

  ©  Ward-Green & Hill Associates Ltd. ~ 877 . 323 . 9137