Relationships Involved in Assessing Empirical Validity

[D]

This diagram depicts three types of assessments of the empirical validity of the Rosenberg Self-Esteem Scale. The boxes labeled "Item 1" to "Item 10" refer to the 10 questions on the scale. We assume we are measuring locus-of-control and marital satisfaction with other questions, but these are not shown on the diagram.

1. Construct validity is assessed by evaluating the statistical significance of each of the proposed causal relationships between self-esteem and the items used to measure self-esteem (i.e., the ten red-colored lines). Note the direction of causality; we assume that a person's self-esteem causes them to respond in a certain manner to each of the 10 questions on the Rosenberg Self-Esteem Scale.
   
   
2. Concurrent validity is assessed by evaluating the statistical significance of the proposed reciprocal relationship between self-esteem and locus-of-control (i.e., the blue-colored, double-headed arrow). The curved arrow indicates we are proposing that self-esteem and locus-of-control exist concurrently.
   
   
3. Predictive validity is assessed by evaluating the statistical significance of the proposed causal relationship between self-esteem and marital satisfaction (i.e., the two green-colored lines).

Return to: Validity and Reliability.