Abstract
Ambulatory assessment (AA) captures within-person fluctuations in daily life, yet not all observed change exceeds measurement error, and not all reliable change is subjectively experienced. We propose a preregistered dual-threshold framework distinguishing statistical detectability (smallest detectable change, SDC) from experiential salience (minimal important change, MIC). Adults (N = 250) completed a 14-day AA protocol with twice-daily and event-contingent assessments (5,982 observations), rating momentary stress, positive affect, and negative affect (0–4) and perceived change since the previous assessment. SDC₉₅ estimates ranged from 1.56 to 1.92 scale points, so roughly three-quarters of consecutive changes fell below reliable detection. Anchor-based MICs, in contrast, were 1 point for all constructs. Correspondence between detectable and perceived change was asymmetric: sensitivity was high (.77–.87), but specificity (.46–.56) and positive predictive values (.24–.31) were low, as participants frequently reported change without reliable change; directional reversals were rare. Reliable change strongly predicted perceived change (ORs 4.6–12.5). Event-contingent assessments, higher prior states, and person-level characteristics (e.g., impulsivity, emotion regulation) increased the general propensity to report change but did not moderate this link. Felt change thus precedes measurable change: detectability and salience are distinct thresholds that intensive longitudinal research must evaluate separately.