Investigating Measurement Invariance Across Situations in Intensive Longitudinal Data

Area of Science:

• Psychometrics
• Longitudinal Data Analysis
• Methodology

Background:

• Intensive longitudinal data (ILD) captures dynamic constructs but faces measurement noninvariance across clustering dimensions.
• Cross-classified factor analysis (CCFA) is used for person and time invariance, but other dimensions are under-explored.
• Situational characteristics, both perceived and objective, can influence responses in ILD.

Purpose of the Study:

• To test the measurement invariance of vigor across people and time using CCFA.
• To investigate situational characteristics as a novel clustering dimension for measurement invariance in ILD.
• To compare invariance across perceived situations versus objective situational properties.

Main Methods:

• Utilized ILD from 225 participants with up to 50 observations each.
• Applied CCFA to assess measurement invariance of vigor items.
• Derived perceived situation profiles via latent profile analysis and included objective aspects (location, company, activity).

Main Results:

• Substantial variation in vigor item measurement parameters across participants indicated noninvariance.
• Measurement invariance held across time and objective situational properties.
• Measurement parameters varied markedly across perceived situations, demonstrating significant noninvariance.

Conclusions:

• Measurement invariance in ILD is influenced by perceived situations, extending beyond time as a clustering dimension.
• Researchers should incorporate individuals' situation perception when evaluating measurement invariance in ILD.
• The CCFA approach can be adapted to include additional clustering dimensions in future ILD research.