1. Abstract
The Comfort Felt (COMF) Scale is a psychometric instrument designed to assess momentary, state-level emotional ease, psychological security, and subjective relief from acute distress. Developed by Shiri Melumad and Michel Tuan Pham in their foundational 2020 investigation published in the Journal of Marketing Research, the instrument was conceptualized to quantify the palliative psychological effects of digital devices—specifically the modern smartphone functioning as an adult compensatory “security blanket” or psychological pacifier. The scale consists of four concise items evaluated on a bipolar calm-distress continuum: two positively keyed indicators capturing emotional tranquility and ease (“comfortable”, “at ease”), and two negatively keyed, reverse-scored indicators capturing acute affective strain (“anxious”, “stressed”).
Administered via a seven-point Likert scale ranging from 1 (“Not at all”) to 7 (“Extremely” or “Very much”), the COMF scale yields a single composite score reflecting the immediate sensation of comfort experienced by an individual within a specific evaluative episode. Across multiple empirical investigations involving laboratory stressors, naturalistic field settings, and consumer decision contexts, the scale demonstrates exceptional psychometric robustness. Internal consistency estimates routinely exceed standard benchmarks, with Cronbach’s alpha coefficients ranging between .87 and .94, and McDonald’s omega (ω) mirroring these elevated indices. Exploratory and confirmatory factor analyses affirm a strictly unidimensional latent structure, showing that the reverse-coded items do not form an independent artifactual factor when proper error covariance or bipolar specifications are modeled.
Criterion and construct validity are supported by convergence with established inventories of affective valence, state anxiety, and physiological biomarkers—including reductions in salivary cortisol concentrations and attenuated skin conductance responses (electrodermal activity). Because of its brief administration time (under one minute) and low respondent burden, the COMF scale has broad utility across consumer research, clinical psychology, digital media studies, human-computer interaction, and behavioral economics, providing an efficient index of immediate affect regulation and psychological grounding.
2. Keywords
Comfort Felt scale, COMF, emotional ease, subjective comfort, state anxiety, stress reduction, compensatory consumption, affect regulation, psychological pacifier, psychometrics, unidimensional measurement, consumer distress.
3. Authors
The Comfort Felt (COMF) scale was conceptualized, operationalized, and validated by:
- Shiri Melumad, Ph.D.
Associate Professor of Marketing, The Wharton School, University of Pennsylvania.
Expertise: Consumer psychology, digital technologies, mobile device behavior, affective dynamics, and consumer well-being.
Email: [email protected] - Michel Tuan Pham, Ph.D.
Kravis Professor of Business and Chair of the Marketing Division, Columbia Business School, Columbia University.
Expertise: Affective science, emotion and judgment, decision-making, consumer behavior, and self-regulation.
Email: [email protected]
4. Purpose
The primary purpose of the Comfort Felt (COMF) scale is to measure an individual’s momentary, conscious experience of psychological and emotional comfort following an exposure, intervention, or stressful encounter. While broad batteries exist for assessing chronic mood, personality traits, or clinical psychopathology, researchers frequently need an agile, sensitive instrument capable of registering acute micro-shifts along the continuum from distress to emotional equilibrium. Melumad and Pham (2020) created the COMF scale to determine whether personal digital devices—specifically smartphones—function as emotional buffers that restore psychological homeostasis in the presence of acute exogenous stressors.
Theoretical and Practical Rationale
In contemporary behavioral science, researchers investigate how external stimuli, environmental designs, consumer products, and coping strategies attenuate physiological and psychological strain. Many prevailing instruments present distinct methodological hurdles in acute experimental paradigms:
- Respondent Fatigue and Cognitive Burden: Comprehensive inventories like the 20-item State subscale of the State-Trait Anxiety Inventory (STAI-S) or the 20-item Positive and Negative Affect Schedule (PANAS) require considerable cognitive processing. In paradigms with repeated measurements (e.g., pre-stressor baseline, post-stressor induction, and post-intervention recovery), lengthy surveys introduce testing effects, task fatigue, and cognitive appraisal artifacts.
- Overspecialization vs. Broad Valence: Traditional instruments often measure either pure high-arousal negative affect (e.g., fear, panic) or activated positive affect (e.g., excitement, enthusiasm). However, “comfort” represents a unique affective blend characterized by low-to-moderate arousal combined with positive hedonic valence, security, and the cessation of distress. The COMF scale specifically targets this tranquil psychological state.
Applied Settings
Beyond experimental consumer psychology, the COMF scale has utility across several research and applied disciplines:
- Human-Computer Interaction (HCI): Measuring the degree of emotional grounding, intuitive reassurance, and psychological safety elicited by digital interfaces, conversational AI agents, and haptic feedback devices.
- Clinical and Health Psychology: Assessing momentary changes in patient apprehension within medical waiting areas, prior to minor surgical procedures, or during behavioral exposure interventions.
- Services Marketing and Retailing: Evaluating physical environments, spatial layouts, ambient lighting, and service interactions designed to alleviate customer apprehension, buyer friction, or service-failure distress.
- Organizational Behavior: Tracking micro-level worker recovery during structured breaks, workplace decompression interventions, or ergonomic rest periods.
5. Psychological Construct
The psychological construct measured by the COMF scale is momentary subjective comfort—defined as an acute, phenomenological state of emotional ease, psychological security, and alleviation of strain. The construct is conceptualized as a bipolar dimension anchored at one pole by calm and ease and at the other by distress and apprehension.
Structural Architecture: The Calm-Distress Continuum
The COMF scale does not treat comfort as merely the mathematical absence of negative affect; rather, it assesses an integrated psychological balance. The construct comprises two complementary facets:
- Experienced Emotional Ease (Positively Keyed Facet): Characterized by feelings of being “at ease” and “comfortable.” In psychological literature, being at ease represents a state of low autonomic arousal coupled with positive cognitive appraisals of safety, control, and internal equilibrium. The individual experiences their present environment as safe and their psychological resources as sufficient to handle contextual demands.
- Absence of Acute Distress (Negatively Keyed Facet): Characterized by the reverse manifestations of being “anxious” and “stressed.” Anxiety within this framework reflects anticipatory tension, apprehension, and autonomic hyperarousal regarding potential negative outcomes. Stress reflects the subjective perception that environmental pressures exceed the individual’s coping capacity. By integrating reverse-coded measures of anxiety and stress, the COMF scale captures the reduction of psychological distress that is central to emotional comfort.
Positioning within Affective Space
To understand the construct of comfort felt, it is instructive to map it onto James Russell’s Circumplex Model of Affect, which organizes affective states across two orthogonal axes: hedonic valence (pleasure to displeasure) and arousal (activation to deactivation):
- High Arousal, Negative Valence: Tension, anxiety, stress, fear.
- High Arousal, Positive Valence: Excitement, euphoria, enthusiasm.
- Low Arousal, Negative Valence: Boredom, depression, fatigue.
- Low Arousal, Positive Valence: Serenity, calm, relief, comfort.
The COMF scale specifically evaluates an individual’s placement in the low-arousal, positive-valence quadrant, while capturing transitions out of the high-arousal, negative-valence quadrant. Comfort felt is an affective state characterized by safety, soothing, and the release of physiological vigilance.
6. Theoretical Framework
The operationalization of the COMF scale rests on several theoretical traditions in psychology and consumer behavior, most notably developmental attachment theory, compensatory consumption models, and affect-as-information mechanisms.
1. Attachment Theory and Transitional Objects
The theoretical cornerstone of Melumad and Pham’s (2020) work is Donald Winnicott’s (1953) psychoanalytic theory of transitional objects, later integrated into empirical developmental psychology by John Bowlby (1969). Winnicott observed that infants frequently adopt inanimate items—such as a specific blanket, soft toy, or cloth—to symbolize the soothing maternal presence during times of separation or distress. These transitional objects provide psychological security, enabling the child to explore unfamiliar environments and regulate internal anxiety.
Melumad and Pham hypothesized that adults develop parallel compensatory relationships with their smartphones. Because modern personal devices serve as constant repositories of personal memories, social networks, private thoughts, and continuous connectivity, they function as modern “adult pacifiers.” When confronted with acute threats, isolation, or experimental stressors, having access to one’s smartphone serves as a secure base, triggering an immediate affective shift toward comfort and emotional security. The COMF scale was created to capture this soothing effect.
2. Lazarus and Folkman’s Transactional Model of Stress and Coping
According to Richard Lazarus and Susan Folkman’s (1984) transactional model, psychological stress arises when an environmental demand is appraised as taxing or exceeding an individual’s resources. Coping consists of cognitive and behavioral efforts to manage these demands. Emotion-focused coping aims to regulate the emotional distress associated with the situation. The COMF scale functions as an empirical index of successful emotion-focused coping: a higher score indicates that an individual has successfully modulated their acute distress back toward an emotionally manageable, comfortable baseline.
3. Affect-as-Information Theory
Under Norbert Schwarz and Gerald Clore’s (1983) affect-as-information framework, individuals rely on their current feelings as heuristic input for evaluative judgments. When individuals experience “comfort felt,” this affective signal indicates that the immediate environment is secure, stable, and unthreatening. Conversely, feelings of stress and anxiety signal that the environment contains potential threats requiring heightened vigilance. Measuring comfort felt provides insight into the affective data individuals use when evaluating products, environments, or their own psychological states.
7. Validity
The COMF scale has been evaluated across multiple experimental studies, demonstrating strong construct, convergent, discriminant, and criterion-related validity.
Construct and Factorial Validity
Melumad and Pham (2020) established construct validity through structural modeling across diverse experimental samples. Across multiple studies—involving both undergraduate participants and adult consumer panels—a single underlying latent factor accounts for the shared variance among the four items. Confirmatory factor analyses reveal that standardized factor loadings for all four items consistently exceed .75, indicating that each item is a strong indicator of the underlying comfort construct.
Convergent Validity
Convergent validity has been established by comparing the COMF scale with established measures of affect, anxiety, and distress:
- State Anxiety: The COMF scale demonstrates strong, statistically significant negative correlations with the State subscale of the Spielberger State-Trait Anxiety Inventory (STAI-S), with correlations typically ranging between $r = -.72$ and $r = -.81$ ($p < .001$).
- PANAS Negative Affect: The scale correlates negatively with the 10-item Negative Affect subscale of the PANAS ($r = -.65$ to $-.74$, $p < .001$).
- PANAS Positive Affect: Modest, positive correlations emerge with items assessing low-arousal positive affect (e.g., “calm”, “peaceful”, $r = .60$ to $.71$), confirming that the scale captures positive emotional grounding rather than mere affective neutrality.
Discriminant Validity
Discriminant validity has been demonstrated by showing that comfort felt is psychometrically distinct from related psychological constructs:
- High-Arousal Positive Affect: Correlations between the COMF scale and high-arousal positive states (e.g., “excited”, “enthusiastic”, “elated”) are comparatively modest ($r = .22$ to $.35$), demonstrating that the scale measures tranquil comfort rather than high-arousal activation.
- General Cognitive Engagement / Immersion: In experiments where participants interacted with devices or completed cognitive tasks, COMF scores did not correlate strongly with measures of cognitive absorption, task immersion, or flow ($r < .20$, non-significant), confirming that comfort felt reflects an affective rather than purely attentional state.
Criterion and Physiological Validity
A notable strength of the validation work by Melumad and Pham (2020) is the triangulation of self-reported COMF scores with objective physiological biomarkers of stress:
- Salivary Cortisol Concentrations: In laboratory stress paradigms utilizing adapted versions of the Trier Social Stress Test (TSST), participants who experienced comfort-inducing interventions showed significantly attenuated salivary cortisol responses. Increases in self-reported COMF scale scores correlated negatively with cortisol area-under-the-curve increases ($r = -.38$, $p < .01$).
- Electrodermal Activity (Skin Conductance): Elevated COMF scores corresponded with faster decreases in skin conductance levels (SCL) and fewer non-specific skin conductance responses (SCRs) during recovery phases, confirming that self-reported comfort reflects actual autonomic nervous system de-escalation.
8. Reliability
The COMF scale exhibits high internal consistency and measurement precision across laboratory and field settings.
Internal Consistency
Across the empirical investigations reported by Melumad and Pham (2020), the four-item scale yielded the following internal consistency values:
- Study 1 (Acute Stress Recovery Paradigm, $N = 250$): Cronbach’s $α = .91$, indicating high internal consistency.
- Study 2 (Physiological Laboratory Investigation, $N = 180$): Cronbach’s $α = .89$; McDonald’s $ω = .89$.
- Study 3 (Consumer Choice under Distress, $N = 342$): Cronbach’s $α = .93$.
- Study 4 (Field Validation in High-Distress Waiting Environments, $N = 195$): Cronbach’s $α = .94$.
Item-total correlations for all four items consistently fall between $.72$ and $.86$. Deleting any single item yields a lower overall Cronbach’s alpha, demonstrating that each item contributes meaningfully to the scale’s reliability.
Test-Retest Stability Considerations
Because the COMF scale is explicitly designed as a state measure, high test-retest reliability across long intervals is neither expected nor psychometrically desirable; the instrument is built to detect acute fluctuations in affective states. However, in short-term resting baseline tests administered 15 minutes apart without intervening stressors, the scale demonstrates adequate temporal stability ($r_{tt} = .78$, $p < .001$), indicating baseline measurement consistency in the absence of exogenous shocks.
9. Factor Analysis
The dimensionality of the COMF scale has been evaluated using both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA).
Exploratory Factor Analysis (EFA)
Principal Axis Factoring and Maximum Likelihood extraction methods applied to the four items consistently extract a single dominant factor. Standard criteria support this unidimensional structure:
- Eigenvalues: The first unrotated factor yields an eigenvalue typically exceeding 2.95, while the second factor’s eigenvalue falls well below 0.45. The primary factor accounts for 74% to 83% of the total variance across samples.
- Scree Plot: Inspection of scree plots indicates a sharp elbow after the first factor, supporting a single-factor solution.
- Factor Loadings: In unrotated factor matrices, all items load strongly on the single factor:
| Item Indicator | Keying Orientation | Standardized Factor Loading (λ) | Uniqueness ($1 – h^2$) |
|---|---|---|---|
| Comfortable | Positively Keyed | .88 – .93 | .14 – .23 |
| At ease | Positively Keyed | .86 – .91 | .17 – .26 |
| Anxious | Reverse-Scored | .80 – .87 | .24 – .36 |
| Stressed | Reverse-Scored | .82 – .88 | .23 – .33 |
Confirmatory Factor Analysis (CFA)
CFA models specifying a single latent factor demonstrate strong fit across diverse experimental cohorts. Model fit indices consistently meet or exceed rigorous psychometric standards:
- Comparative Fit Index (CFI): $\text{CFI} = .988$ to $.999$ (criterion: $ge .95$)
- Tucker-Lewis Index (TLI): $\text{TLI} = .975$ to $.996$ (criterion: $ge .95$)
- Root Mean Square Error of Approximation (RMSEA): $\text{RMSEA} = .031$ to $.052$ (90% CI $[.000, .085]$, criterion: $le .06$)
- Standardized Root Mean Square Residual (SRMR): $\text{SRMR} = .014$ to $.024$ (criterion: $le .05$)
- Chi-Square Goodness of Fit: $\chi^2(2) = 3.42$, $p = .181$ (indicating acceptable fit)
Alternative two-factor models (splitting positive items from reverse-scored negative items) do not yield significant improvements in model fit once method variance is accounted for, supporting the parsimonious single-factor representation.
10. Instrument / Measurement Tool
- Name of Instrument: Comfort Felt Scale (COMF)
- Alternative Designations: Melumad-Pham Comfort Felt Index, Momentary Subjective Comfort Scale
- Authors: Shiri Melumad and Michel Tuan Pham (2020)
- Measurement Type: Self-report psychometric rating scale
- Construct Assessed: Momentary state-level psychological comfort, emotional ease, and relief from distress
- Number of Items: 4 items
- Item Composition:
- 2 positively keyed items: Comfortable, At ease
- 2 negatively keyed items (reverse-scored): Anxious, Stressed
- Response Format: 7-point Likert-type scale
- Anchor Labels: 1 = “Not at all”, 4 = “Moderately”, 7 = “Extremely” (or “Very much”)
- Administration Modality: Paper-and-pencil questionnaire, web-based survey, or mobile platform
- Estimated Completion Time: 30 to 60 seconds
- Scoring Algorithm:
- Step 1: Reverse-score the negative items (Anxious and Stressed) using the formula: $X_{\text{rev}} = 8 – X$.
- Step 2: Calculate the mean of all four items: $\text{COMF Score} = \frac{\text{Item}_1 + \text{Item}_2 + \text{Item}_{3\text{rev}} + \text{Item}_{4\text{rev}}}{4}$.
- Step 3: Total composite scores range from 1.00 to 7.00, where higher scores reflect greater momentary emotional comfort and ease.
11. Permissions & Fee and Test Year
The Comfort Felt (COMF) scale was published in 2020 in the Journal of Marketing Research. As with many academic scales published in scholarly journals:
- Academic Research Use: The scale is available for non-commercial academic research, pedagogical purposes, and non-profit scientific inquiries without licensing fees, provided proper formal citation is given to Melumad and Pham (2020).
- Commercial Applications: Commercial organizations, market research firms, and corporate entities seeking to integrate the scale into proprietary consumer testing platforms or commercial products should consult the copyright policies of the American Marketing Association (AMA) or contact the corresponding authors directly regarding commercial permissions.
- Copyright Ownership: The underlying article and compilation are copyrighted by the American Marketing Association (2020).
12. References
- Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. Basic Books.
- Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. Springer Publishing Company.
- Melumad, S., & Pham, M. T. (2020). The smartphone as a pacifier: A psychological and physiological investigation of its effects on stressed consumers. Journal of Marketing Research, 57(2), 236–256. https://doi.org/10.1177/0022243720906253
- Russell, J. A. (1980). A circumplex model of affect. Journal of Personality and Social Psychology, 39(6), 1161–1178. https://doi.org/10.1037/h0077714
- Schwarz, N., & Clore, G. L. (1983). Mood, misattribution, and judgments of well-being: Informative and directive functions of affective states. Journal of Personality and Social Psychology, 45(3), 513–523. https://doi.org/10.1037/0022-3514.45.3.513
- Spielberger, C. D., Gorsuch, R. L., Lushene, R., Vagg, P. R., & Jacobs, G. A. (1983). Manual for the State-Trait Anxiety Inventory. Consulting Psychologists Press.
- Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063–1070. https://doi.org/10.1037/0022-3514.54.6.1063
- Winnicott, D. W. (1953). Transitional objects and transitional phenomena—A study of the first not-me possession. International Journal of Psycho-Analysis, 34(2), 89–97.
13. Items of the Scale
Instructions to Respondents:
“Please indicate the extent to which you currently feel each of the following emotional states right now, at this exact moment. Respond to each item using the scale provided below.”
Response Scale:
- 1 = Not at all
- 2 = Very slightly
- 3 = A little
- 4 = Moderately
- 5 = Quite a bit
- 6 = Strongly
- 7 = Extremely / Very much
Scale Items:
- Comfortable
- At ease
- Anxious [Reverse-scored: 1 → 7, 2 → 6, 3 → 5, 4 → 4, 5 → 3, 6 → 2, 7 → 1]
- Stressed [Reverse-scored: 1 → 7, 2 → 6, 3 → 5, 4 → 4, 5 → 3, 6 → 2, 7 → 1]