The stability of individual response styles.

Area of Science:

• Psychometrics
• Survey Methodology
• Statistical Modeling

Background:

• Longitudinal survey data analysis is advancing, but method bias, particularly from response styles, can impact validity.
• Individual response styles, such as acquiescence and extreme responding, are potential sources of bias in survey data.
• The stability of these response styles over time is not fully understood, posing challenges for longitudinal research.

Purpose of the Study:

• To investigate the stability of individual response styles in longitudinal survey data.
• To model multiple directional response biases simultaneously over a 1-year period.
• To determine the extent to which demographics influence stable response style components.

Main Methods:

• Longitudinal data collection with two online questionnaires administered 1 year apart.
• Modeling of four key response styles: acquiescence, disacquiescence, midpoint, and extreme response style.
• Application of a second-order factor model derived from latent state-trait theory, including time-invariant and time-specific factors.

Main Results:

• Response styles exhibit a substantial stable component over the 1-year interval.
• Demographic variables accounted for a minor part of the stable response style variance.
• The proposed model effectively captured the complex interplay of stable and transient response style factors.

Conclusions:

• Individual response styles possess a significant stable trait, influencing longitudinal survey responses.
• While demographics play a minor role, the inherent stability of response styles is a key consideration for data interpretation.
• Findings underscore the importance of accounting for stable response styles to enhance the validity of longitudinal survey research.