Messick (1995) suggested that the traditional view of construct is fragmented

Full Answer Section

• Overly reliant on single scores: Ignored the multifaceted nature of constructs and the potential for different aspects to be measured differently.
• Static and context-insensitive: Neglected the influence of testing context, task demands, and individual differences on responses.
• Limited focus on consequential validity: Underscored the importance of considering the actual use and consequences of test scores and how they impact decisions and actions.

Messick's Expanded View and Four Aspects:

1. Content Validity: Represents the degree to which the test content adequately reflects the intended construct.

• Example: A math test measures various aspects of mathematical ability, not just memorization of formulas.

2. Structural Validity: Examines the internal structure of the test and how items relate to each other and underlying sub-constructs.

• Example: A personality test shows separate clusters for "extraversion" and "neuroticism," reflecting their theoretical structure.

3. Criterion-Related Validity: Focuses on the test's ability to predict relevant performance or behaviors outside the testing context.

• Example: A cognitive ability test predicts academic success in college, showcasing its link to real-world performance.

4. Consequential Validity: Explores the actual impact and implications of test scores on individuals and society.

• Example: A language proficiency test used for immigration decisions should not unfairly disadvantage minority groups.

Unified Validity Framework (Progressive Matrix):

Messick proposed a four-category framework representing different ways evidence can accumulate for construct validity:

1. Content: Focuses on how the test reflects the construct's scope and characteristics.
2. Antecedent: Examines how individual characteristics and prior experiences influence test performance.
3. Concurrent: Considers how performance on the test relates to concurrent behaviors and measures.
4. Consequential: Assesses the actual effects and implications of using the test scores.

Each cell of the matrix (content x antecedent x concurrent x consequential) suggests specific types of evidence to evaluate construct validity.

Example: Measuring "critical thinking" through a problem-solving test (content) that considers individual reasoning skills (antecedent), correlates with real-world problem-solving tasks (concurrent), and leads to fair decision-making in job selection (consequential).

Conclusion:

Messick's view offered a more holistic and dynamic approach to construct validity, emphasizing the importance of different evidence types and considering the wider context of test use. This framework helps researchers and practitioners critically evaluate and improve the quality of their assessments.

Sample Answer

Messick's Unified Framework for Construct Validity

Traditional View of Construct Validity:

The traditional view of construct validity saw it as a single concept focused on matching a test score to a true score representing an underlying psychological construct. This approach relied heavily on convergent validity (correlating the test with other measures of the same construct) and discriminant validity (showing it doesn't correlate with measures of unrelated constructs). While useful, this fragmented view had limitations: