Let sleeping dogs lie? How to deal with the night gap problem in experience sampling method data.

Area of Science:

• Psychological Methods
• Quantitative Psychology
• Behavioral Science

Background:

• Experience Sampling Method (ESM) data collection inherently includes 'night gaps' between daily measurements.
• The impact of these night gaps on analyzing lagged relations (autoregression, cross-lagged regressions) is often overlooked.
• Existing methods for handling night gaps include ignoring them, removing them, or treating them as missing data.

Purpose of the Study:

• To explicitly detail the theoretical implications of three common methods for handling night gaps in first-order autoregressive models.
• To introduce an alternative modeling framework for more granular investigation of night gap effects.
• To empirically test which night gap handling method best fits different variables in ESM data.

Main Methods:

• Theoretical analysis of three established night gap handling techniques within autoregressive modeling.
• Development and proposal of a novel, more detailed modeling approach for night gaps.
• Empirical application using an N=1 design with multiple ESM variables to compare model fits.

Main Results:

• The study demonstrates that common night gap handling methods are special cases of the proposed alternative approach.
• Empirical findings indicate that the optimal method for modeling night gaps is variable-dependent.
• This variability suggests that psychological processes may exhibit distinct dynamics during nighttime compared to daytime.

Conclusions:

• The choice of method for handling night gaps in ESM data significantly influences the interpretation of lagged relationships.
• A one-size-fits-all approach is insufficient; variable-specific modeling is necessary for accurate analysis of ESM data.
• This research provides a foundation for more sophisticated understanding and modeling of night gaps in ESM studies.