intelligence and psychological testing

Reliability

  • Definition: Reliability refers to the consistency of a measure. A reliable measure produces the same results under consistent conditions.

  • Types of Reliability:

    • Test-retest Reliability: Measures the stability of results over time by administering the same test to the same subjects at different times. High correlation between the two sets of scores indicates high reliability.

    • Inter-rater Reliability: Assesses the degree to which different raters or observers give consistent estimates of the same phenomenon. High agreement among raters signifies reliable measurements.

    • Internal Consistency: This form assesses the consistency of results across items within a test. Commonly measured using Cronbach’s alpha; a value above 0.7 is considered acceptable.

Validity

  • Definition: Validity indicates how well a test measures what it is intended to measure. It assesses the accuracy of a measure.

  • Types of Validity:

    • Content Validity: Evaluates whether a measure covers the entire range of the construct being studied. Ensures that the test includes relevant material to fully represent the construct.

    • Criterion Validity: Compares the measure with a criterion (another established measure) to see how well it predicts or correlates with outcomes. Splits into two subtypes:

      • Predictive Validity: The ability of a measure to predict future outcomes.

      • Concurrent Validity: The extent to which test results correlate with those from another measure taken at the same time.

    • Construct Validity: Refers to how well a test measures the theoretical construct it is designed to measure. Assesses whether the test truly assesses the concept it claims to measure, involving both convergent validity (correlation with similar measures) and discriminant validity (non-correlation with dissimilar measures).

Key Differences Between Reliability and Validity

  • Reliability is about consistency; a test can be reliable (produce consistent results) without being valid (measuring what it claims to measure).

  • Validity is about accuracy; a test must be reliable to be valid, but not all reliable tests are valid.