Reliability generalization: an examination of the Positive Affect and Negative Affect Schedule

Area of Science:

• Psychology
• Social Sciences
• Psychometrics

Background:

• The Positive Affect and Negative Affect Schedule (PANAS) is widely used in social sciences.
• Understanding the reliability of PANAS across diverse studies is crucial for its application.
• Previous research often relies on single-study reliability, potentially limiting generalizability.

Purpose of the Study:

• To investigate population reliability scores for the Positive Affect (PA) and Negative Affect (NA) scales of the PANAS.
• To utilize reliability generalization (RG) meta-analytic methods to assess PANAS reliability.
• To examine factors influencing PANAS reliability, such as time frame, language, and sample characteristics.

Main Methods:

• A meta-analytic approach using reliability generalization (RG) was employed.
• Data from 147 studies were analyzed using a random effects model.
• Internal consistency and test-retest reliability coefficients were examined, correcting for measurement error.

Main Results:

• PANAS scales demonstrated moderate to high internal consistency coefficients.
• Reliability scores varied based on time frame instructions, item language, and sample characteristics.
• True score variance for PA and NA differed by up to 11% across subpopulations, indicating both state-like fluctuations and trait-like stability.

Conclusions:

• The reliability generalization analysis confirms the robustness of PANAS reliability.
• Population reliability coefficients derived from RG are more relevant for research and assessment than single-study coefficients.
• The findings support the continued use of PANAS in social science research, with an awareness of factors influencing its reliability.