On the Utility of Indirect Methods for Detecting Faking.

Area of Science:

• Psychometrics
• Survey Methodology
• Behavioral Science

Background:

• Questionnaire data validity is crucial in research.
• Direct faking detection scales (e.g., social desirability) can be detected by respondents.
• Indirect indices offer a covert method for identifying response biases.

Purpose of the Study:

• To compare the performance of 12 distinct indirect faking detection indices.
• To evaluate the effectiveness of indirect indices individually and in combination.
• To assess indirect indices against established direct faking measures.

Main Methods:

• Selection of 12 conceptually different indirect faking detection indices.
• Comparative analysis of individual and combined indirect index performance.
• Benchmarking against a direct faking measure (validity scale).

Main Results:

• The agreement factor of the Likert-type item response process tree model, desirable scale endpoint responses, and covariance index showed superior performance.
• Combined indirect indices achieved detection rates comparable or superior to direct measures.
• Effective indirect indices exhibited low correlation with substantive scales, preserving data integrity.

Conclusions:

• Indirect faking detection indices are a viable and effective alternative to direct measures.
• Combining indirect indices enhances their utility for detecting response biases.
• Researchers are encouraged to adopt indirect indices for improved data quality in questionnaires.