Systematic comparison between ecological momentary assessment and day reconstruction method for fatigue and mood states in healthy adults

Abstract

Objectives: While both ecological momentary assessment (EMA) and the day reconstruction method (DRM) have been used to overcome recall bias, a full systematic comparison of these methods has not been conducted. This study was aimed to investigate the differences and correlations between momentary fatigue and mood states recorded by EMA and reconstructed ones recorded by simultaneous DRM in healthy adults.

Design: Each of two different designs (time-based and episode-based) of EMA and DRM were simultaneously conducted.

Methods: Twenty-five healthy adults recorded momentary fatigue and mood states with EMA, and then, reconstructed them with DRM. Differences between the mean and the variability of momentary and reconstructed recordings, and the correlations between them, are analysed for different EMA designs.

Results: No significant differences are found between the mean or the variability of EMA and DRM estimated over the monitoring period. However, correlations between EMA and DRM are low, albeit statistically significant.

Conclusions: Although the overall mean and variability of EMA recordings may be accessible with DRM, detailed changes over time of momentary fatigue and mood states are not retrieved by DRM. Statement of contribution What is already known on this subject? Day reconstruction method (DRM) may be a reliable substitute strategy for the measurement of subjective symptoms instead of ecological momentary assessment (EMA). Remembering the context of daily activities with DRM is assumed to be helpful in reconstructing subjective symptoms without recall bias. What does this study add? We are not able to reconstruct our diurnal time course (i.e. detailed changes over time) of subjective symptoms (e.g. fatigue and mood states in this study) with DRM, while their approximate mean and overall variability during the study period may be accessible with DRM. Reconstructed depression by DRM could be biased when the subjects remembered whether their behaviour was active or inactive.