The Momentary Reliability of Single-item Emotion Measures in Ecological Momentary Assessment: Personal, Situational and Temporal Correlates

Single-item (emotion) ratings are commonly used in Ecological Momentary Assessment (EMA) research, yet their momentary reliability remains difficult to establish. Building on the test-retest procedure introduced by Dejonckheere and colleagues (2022), this study had two aims: (1) to evaluate the robustness of the test-retest approach by directly replicating prior estimates of measurement error variance in a similar, but independent sample, and (2) to examine whether personal, situational, and temporal factors systematically shape fluctuations in this error component. We conducted a one-week EMA study in which 122 participants completed 10 surveys per day, each containing 12 momentary emotion items and a randomly repeated retest item. Passive mobile sensing (GPS, pedometer, app logs, ambient noise levels) captured aspects of participants’ real-time context at each assessment. The average discrepancy between test-retest pairs corresponded to approximately 9 points of measurement error on a 0-100 slider scale, closely replicating earlier estimates and confirming the robustness of the test-retest reliability procedure. None of the examined person-level characteristics (social desirability, conscientiousness, neuroticism), contextual indicators (home location, walking, phone use, background noise), or diurnal patterns predicted momentary reliability. Overall, these findings tentatively suggest that, under typical EMA conditions, variability in single-item reliability reflects random rather than systematic influences. This would be an encouraging conclusion, as it implies that such influences may not require explicit consideration when designing EMA studies or modeling EMA data.