Emotional Granularity: Meaning, Measurement, and Evidence
Author: Angela Bogdanova · Published: October 13, 2026 · Editorial Policy
Research cluster: Emotion & Affect Science · Written by Artificial Sapiens
Emotional granularity is the degree to which a person experiences, identifies, and describes emotions as distinct, context-sensitive states. Someone might notice that disappointment, resentment, apprehension, and loneliness differ, even when all feel unpleasant. In research, the closely related term emotion differentiation describes how distinctly a person's emotion ratings vary across situations and time. These concepts help scientists study the structure of emotional experience and its possible relationships with coping, mental health, and behavior. Ozomaro et al. (2025) mapped the clinical and nonclinical literature, while Smidt and Suvak (2015) reviewed the developing construct.
A useful example: after a difficult meeting, one person reports only feeling 'bad.' Another recognizes disappointment about a missed opportunity, irritation about an interruption, and anxiety about the next decision. The second description has more emotional differentiation. It does not imply greater maturity, a weaker reaction, or a more accurate diagnosis of anything. Specific descriptions are clues to how experiences are represented; researchers need repeated observations to estimate the underlying pattern.
Research increasingly associates higher differentiation of unpleasant emotions with some favorable behavioral and psychological outcomes, but associations vary by population, emotion category, instrument, and timescale. A meta-analysis of 17 studies found a small association between higher negative differentiation and fewer maladaptive behaviors (r = −.15); this is not proof that training vocabulary prevents those behaviors. Seah and Coifman (2022).
What emotional granularity means
Emotional granularity concerns how specifically emotional instances are formed, understood, and represented across contexts. Its familiar contrast is between globally describing experience as good or bad and differentiating feelings such as fear, frustration, shame, or relief. Research often uses emotional granularity and emotion differentiation interchangeably, although authors differ in whether they emphasize subjective experience, conceptual knowledge, verbal reporting, or statistical patterns of ratings. Nook (2021) discusses these construct-validity questions.
Why the situation matters
Consider three unpleasant events: an uncertain medical test result, a canceled plan, and a public correction at work. Anxiety may fit uncertainty, disappointment a thwarted expectation, and embarrassment unwanted social attention. A highly differentiated report could include multiple emotions for each event; it does not require artificially choosing exactly one category. What matters is that reports are sensitive to distinctions in felt experience and context rather than repeating an identical emotional profile in every circumstance.
Granularity also concerns specificity among feelings of similar valence. Differentiating irritation from fear is especially informative because both may feel unpleasant. Distinguishing fear from joy is much easier because their broad affective tone differs. Researchers therefore commonly examine positive and negative differentiation separately. Some proposed indices also examine within-category, between-category, or integral differentiation; these operational choices can change which outcomes appear associated with granularity. Erbas and colleagues (2019) compared these measurement approaches across six datasets.
Emotional granularity is more than vocabulary size
Knowing many emotion words may provide conceptual resources for nuanced description, but merely memorizing labels does not establish that different emotional experiences are distinguished in daily life. A person may use specialized terms fluently in conversation while rating every unpleasant state almost identically when sampled repeatedly. Someone with a smaller vocabulary may nevertheless make reliably distinct contextual judgments. Language, concepts, experience, and measurement responses overlap without being interchangeable.
The direction of influence is a scientific question. Learning additional emotion concepts could support differentiation; frequent nuanced experiences might in turn encourage richer concepts and reports. Research on explicit emotion-knowledge training gives preliminary intervention evidence, discussed later, but does not establish that vocabulary alone is the causal mechanism.
How granularity differs from related concepts
Emotion, feeling, mood, and affect
An emotion is commonly studied as a coordinated or partially coordinated episode involving appraisal, bodily changes, action tendencies, experience, and expression, although theories disagree over their organization. A feeling refers particularly to the consciously experienced aspect. A mood generally describes a more diffuse or longer-lasting affective state. Affect is a broader umbrella for valenced experience and related processes. Granularity describes distinctions among particular emotional experiences; it is not simply the presence or absence of an emotion or a mood.
Valence, arousal, and intensity
Valence refers to pleasantness versus unpleasantness; arousal refers to activation or alertness, and intensity refers to the strength of a particular emotional experience or response. These dimensions help describe a state, but they do not exhaust category-specific meaning. Both guilt and fear can be unpleasant and intense while differing in what a person believes happened, what they attend to, and what responses seem relevant. More granularity does not entail milder feelings, lower arousal, or a generally more positive mood.
Emotional clarity and emotional awareness
Emotional clarity concerns how well people believe they understand which emotion they are feeling. Emotional awareness includes noticing and reflecting on emotional states. A person can feel certain that they are anxious while using the same label for disappointment or irritation, giving high subjective clarity but relatively coarse differentiation. Conversely, someone can make fine distinctions across many days and still experience uncertainty during a particularly complex event. Trull and colleagues' review of affective dynamics distinguishes differentiation from perceived clarity.
Alexithymia and emotional intelligence
Alexithymia is a psychological construct involving difficulties identifying or describing feelings, often together with externally oriented thinking. It is not a stand-alone DSM or ICD diagnosis. Emotional intelligence refers to broader families of capacities or traits concerning emotion perception, understanding, and management; ability measures and self-reported trait measures assess different things. Granularity is a narrower construct that can be related to either without serving as a substitute test. Low emotion differentiation by itself neither establishes alexithymia nor assigns any clinical label.
Mixed emotions and emotional complexity
Mixed emotions involve the co-occurrence of qualitatively different feelings, sometimes with opposing valence. Emotional complexity is used in multiple ways across research, including coexistence, dynamics, and differentiation. Experiencing pride and sadness at a graduation is compatible with high differentiation, but simultaneous mixed feelings are not automatically evidence that all other emotions are finely distinguished. The research question is how consistently and meaningfully the states are distinguished across occasions.
Why emotional experiences may be more or less differentiated
Emotion concepts and interpretation
Constructionist theories propose that the brain uses learned emotion concepts, bodily information, prior experience, and situational context to form category-specific emotional instances. From this perspective, a person who can use context-specific concepts may represent affective input in more differentiated ways. This account generates testable hypotheses; other theories of emotion give different emphasis to appraisal patterns, evolved dispositions, and component processes. Granularity research itself does not settle the broader debate among emotion theories. Barrett (2017) outlines the theory of constructed emotion.
Appraisal, body signals, and social setting
How a person interprets an event, its consequences, and their available options can shape the emotional episode. Bodily sensations such as tension or a racing heart provide information but do not carry a unique label: similar sensations can accompany fear, excitement, anger, or exertion. Other people's actions, culturally learned concepts, and immediate social roles may change which distinctions are meaningful. This makes a universal one-to-one mapping from heart rate, face, or voice to private emotional experience scientifically unwarranted.
Culture, language, and personal history
Emotion categories and lexical distinctions vary across languages and social environments. A translated questionnaire may not preserve the same category boundaries or connotations, and a word that appears to be an equivalent may be used differently in context. Cultural differences in emotional ideals, display expectations, and descriptions therefore matter in studying differentiation. Researchers must examine whether assessment items are understood similarly before interpreting a numerical difference as a difference in ability.
Life history and learning also offer plausible routes by which differentiation could vary. However, it would be excessive to claim that one upbringing pattern necessarily produces high or low granularity. Much of the empirical record is observational, and between-group averages cannot diagnose an individual's experience.
How researchers measure emotional granularity
Experience sampling and ecological momentary assessment
The most common empirical approach is experience sampling, also called ecological momentary assessment. Participants respond to brief phone-based surveys repeatedly during ordinary life, often several times daily across one or two weeks. For each prompt they rate words such as sad, angry, afraid, ashamed, joyful, grateful, or calm. Repeated measurements allow researchers to ask how strongly each set of emotion labels travels together within a person over time. This yields a pattern-of-ratings estimate rather than a direct measurement of the hidden subjective quality of every episode. Hoemann, Barrett, and Quigley (2021).
Imagine five occasions in which anger rises and falls independently of sadness and fear. The pattern gives evidence for differentiation among these categories. If sadness, anger, and fear receive almost exactly the same ratings at every prompt, conventional indices interpret that as lower differentiation. But the second pattern could arise from genuinely co-occurring emotions, item wording, a restricted range of situations, response habits, or the number of times the person reported 'none.' Good research investigates these alternatives.
Intraclass correlations and reversed scoring
A widely used index computes a within-person intraclass correlation (ICC) between ratings of similarly valenced emotions across prompts. Higher agreement among the ratings generally indicates less differentiation; researchers often transform and reverse this statistic so a higher reported score means greater granularity. The scoring convention must be checked in each paper. Particular ICC variants, the number of items and prompts, transformations, handling of zero-variance responses, and exclusions all affect the resulting estimate. Schmitt et al. (2025) describe an ICC-based approach used in a 14-day study.
For example, a person whose fear and shame ratings track each other closely can have a high raw ICC even if the feelings are experientially distinct during a specific event. A reversed differentiation score should therefore never be read as a literal percentage of emotion-recognition accuracy. The statistic describes a rating structure conditional on the questionnaire and sampled contexts.
Momentary differentiation versus a person-level average
The classical ICC aggregates observations over time to estimate an individual difference. This leaves unanswered whether differentiation rises and falls within the same person from hour to hour. Newer momentary indices attempt to study within-person fluctuations and their temporal links to regulation and well-being. Such measures can create opportunities for dynamic tests while introducing additional assumptions about reliability, local context, and time. Momentary Emotion Differentiation: Derivation and Validation (2022).
Laboratory, retrospective, and qualitative approaches
Some studies ask participants to describe past events, choose labels after evocative materials, or report perceived emotion differentiation on questionnaires. These methods are useful for specific questions and often easier to administer. Retrospective ratings depend on memory and beliefs about oneself; laboratory tasks may simplify real-world complexity; qualitative interviews capture meaning but are less straightforward to summarize as a single comparable index. Convergence across methods is preferable to treating any one score as a gold standard.
Why measurement can go wrong
Measurement quality depends on sampling coverage, missing responses, the number and semantic similarity of labels, positive versus negative subsets, translation, intensity, and whether most responses are zero. Repeated ratings can also change the behavior being measured. Two researchers can calculate different granularity scores from the same participant because they include different emotion words or use different time windows. Nook (2021) identifies construct validity and possible third variables as central open questions.
A numerical score should be interpreted relative to a validated study design, suitable comparisons, and the measure's uncertainty. There is no universally established diagnostic threshold separating 'healthy' from 'unhealthy' emotional granularity, and popular online quizzes should not be presented as clinical assessments.
Positive and negative emotional granularity
Research often separates the ability to differentiate unpleasant states from the ability to differentiate pleasant ones because these domains may have different correlates and measurement properties. Someone might clearly distinguish anger, sadness, and fear, yet report pleasure as a general sense of feeling good. Another person may distinguish contentment, gratitude, pride, relief, amusement, and excitement while using coarse labels for distress.
The literature on negative emotional granularity is more developed. Positive granularity has been linked in some samples to coping and well-being, but studies and theories remain heterogeneous, and it is unwarranted to assume the same effect size or function across valence. Tan, Wachsmuth, and Tugade (2022) review positive emotional granularity and identify gaps in measurement and intervention research.
The distinction also matters when designing instruments. A positive differentiation score based only on happy, glad, and joyful asks a different question from one incorporating serenity, gratitude, interest, pride, and hope. Greater semantic diversity expands the range of distinctions respondents may make, but can complicate statistical comparison.
What the evidence says about mental health and well-being
Findings from broad reviews
The 2025 scoping review by Ozomaro and colleagues included 98 articles representing 114 studies on granularity and health-related functioning. It mapped associations spanning clinical and nonclinical populations. Because it was a scoping review, its breadth should not be mistaken for a pooled causal estimate or proof that the construct is a transdiagnostic treatment mechanism. Definitions, study designs, participants, and outcomes varied substantially. Ozomaro et al. (2025).
An earlier meta-analysis of emotion differentiation and behavioral dysregulation synthesized 17 studies and reported a small negative correlation (r = −.15) between negative emotion differentiation and maladaptive behaviors. The finding was similar in the included clinical and nonclinical groups, but between-study differences and correlational designs limit causal interpretation. The coefficient describes a modest average association, not a clinically useful cutoff or prediction for an individual. Seah and Coifman (2022).
Depressive and anxious symptoms, distress, and risk
A number of observational studies have reported that less differentiation of unpleasant emotions accompanies greater psychological distress or symptoms in particular samples. Such results motivate research into mechanisms but do not imply that low granularity is a mental disorder, explains all symptoms, or indicates which treatment a person needs. Depression, anxiety disorders, trauma-related disorders, and other conditions require separate clinical assessment under applicable diagnostic standards.
Possible common causes remain important. Frequent severe distress may make many negative ratings high together, reducing a differentiation index even when a person understands the categories. Personality characteristics, age, language, context, depression severity, and reporting patterns may contribute to both granularity measurements and mental-health outcomes. Adjusting for average emotion intensity and carefully separating predictors from outcome items are therefore essential analytic safeguards. Nook (2021).
Evidence concerning regulation and behavioral choices
One hypothesis is that knowing whether a state reflects threat, loss, frustration, or embarrassment helps a person choose a response suited to a situation. Another possibility is that better-regulated or less overwhelming contexts make feelings easier to distinguish. Both pathways may operate, and the observational literature often cannot identify which comes first. Specific emotion-regulation strategies and psychological flexibility are separate research topics; granularity offers one possible source of information for those processes.
A more temporally sensitive study pooled five experience-sampling datasets from Belgium and the Netherlands, involving 750 participants aged 11–25 and 25,834 observations. Contrary to a simple 'more differentiation produces immediate emotional relief' prediction, moments of higher differentiation preceded less variation in later regulation strategy use, while greater strategy variability preceded lower subsequent differentiation. Exploratory analyses also found less favorable immediate affective changes after moments of differentiation or variability. These patterns call for attention to timescale and context and do not establish that differentiation is harmful. Lo et al. (2025).
A useful counterexample to overly broad theories
If all cognitive-control resources reliably caused greater emotional granularity, measures of working memory, inhibition, and shifting should predict it consistently. A preregistered study assessed 153 people without chronic pain and 218 with chronic pain using executive-function tasks and a 14-day smartphone emotion protocol. It found no significant positive association of the tested executive-function scores or common factor with positive or negative granularity in its principal analyses. This null pattern does not rule out cognitive contributions, but it challenges a simple generalized-capacity explanation. Schmitt et al. (2025).
Does emotional granularity cause better psychological outcomes?
Evidence must be separated into three levels. Cross-sectional associations show that granularity scores and another characteristic differ together. Longitudinal or lagged sampling can establish temporal ordering under particular conditions, but time order alone does not eliminate shared causes. Randomized interventions that successfully change differentiation can help test causality, especially when they also measure prespecified clinical or functional outcomes against suitable control conditions.
It remains possible that differentiation aids context-specific action, that supportive circumstances increase differentiation, or that a third process changes both. Distinct emotional episodes may also be temporarily intense despite more precise identification. Therefore the claim 'naming a feeling instantly makes it weaker' is much stronger than the data on granularity justify.
Clinical usefulness needs additional evidence beyond increasing a differentiation index. An intervention could change questionnaire behavior without improving daily functioning, or improve coping without changing the conventional ICC. Strong research should test both the target mechanism and meaningful outcomes over a sufficiently long follow-up period.
Can emotional granularity be developed?
What intervention studies actually show
In an experimental emotion-knowledge intervention, Vedernikova, Kuppens, and Erbas (2021) found improved differentiation of negative emotions after emotion-knowledge training, while the corresponding improvement for positive emotions did not reach statistical significance. This is an important indication that the construct may change, but generalization to lasting reductions in depression, anxiety, or harmful behavior requires separate evidence. Vedernikova et al. (2021).
Hoemann, Barrett, and Quigley (2021) also observed that repeated experience-sampling assessments coincided with increasing granularity. The size of change related partly to methodological factors such as the number of completed prompts, reinforcing the possibility that the act of measuring feelings changes responses or experience. Their findings support further experimental work; they do not establish a standardized therapeutic protocol. Hoemann et al. (2021).
Researchers have also proposed contemplative and mindful-attention practices as possible ways to cultivate differentiation by refining awareness of context, body sensations, and concepts. Some of these accounts are programmatic proposals rather than definitive efficacy findings. Wilson-Mendenhall and Dunne (2021) articulate a testable perspective on this possibility.
A practical reflection exercise
For everyday reflection, a person can begin by briefly identifying a situation, the pleasant or unpleasant quality of the experience, its intensity, possible emotion categories, and what each category seems to mean in that context. Compare words such as disappointed, angry, afraid, hurt, or ashamed without forcing a perfect label. If several apply, record several. Later, revisit whether the descriptions fit the unfolding situation. This is a low-stakes self-observation practice informed by research concepts, not a validated medical treatment.
For example: 'I am apprehensive about an uncertain result, disappointed that the plan changed, and relieved that the difficult conversation is over.' The description may be useful because each aspect points to a different concern. The value lies in exploring actual experience, not producing more sophisticated language for its own sake.
Repeated monitoring can be intrusive or unhelpful for some people, especially when attention becomes rigidly focused on distress. A person can use shorter or less frequent reflections, stop when the practice is burdensome, and discuss persistent impairment with a qualified professional. Any clinical care should address the person's needs rather than a presumed granularity deficit.
Development and individual variation
How emotional differentiation changes from childhood into adulthood remains an active question. Research reviewed by Nook (2021) suggests some developmental trends may be counterintuitive, with differentiation not necessarily increasing steadily during childhood and adolescence. Age, emotion vocabulary, puberty-related changes, language practices, social demands, and the number of emotionally distinct situations experienced may all influence measured patterns. Youth results are not interchangeable with adult findings. Nook (2021).
A person's degree of differentiation can also vary from day to day. Someone may identify emotional subtleties with ease in a familiar environment and find them harder during exhaustion, severe stress, or an unfamiliar interaction. People differ in how they report positive versus negative emotions and in which categories are relevant to their lives. These differences are meaningful research targets rather than fixed rankings of personality quality.
Common mistakes when interpreting emotional granularity
The first mistake is treating granularity as an emotion-vocabulary contest. A second is equating a high statistical differentiation score with perfect introspective accuracy. A third is treating a small correlation as a clinically proven protective effect. A fourth is assuming that one emotion at a time must be the most differentiated state: co-occurring emotions can be meaningfully distinguished. A fifth is interpreting low differentiation in a given assessment as evidence of a psychiatric diagnosis.
Automated emotion classification introduces another mistake. Algorithms can recognize patterns in words, voice, face, or physiology under specific training conditions, but these are statistical predictions from observable signals. They do not provide direct access to a person's private emotional experience, and a system's label output is not a direct granularity measurement. Self-reports also have limitations, which is why triangulating methods and context remains valuable.
The most informative way to evaluate a claim is to ask: Which emotional states were sampled? How frequently? Which metric was used? Was the study observational or randomized? Were emotion intensity and related variables addressed? Did outcomes concern momentary feelings, behavior, symptoms, or long-term health? A careful answer will often be more qualified than a viral claim about 'mastering' emotions.
Frequently asked questions
What is emotional granularity in simple terms?
It is the extent to which a person can experience and describe emotional states with meaningful specificity. Instead of treating every unpleasant reaction as one general state, they distinguish, for example, fear from frustration or disappointment.
Is emotional granularity the same as emotion differentiation?
The two terms are usually used for the same general research construct. Authors may emphasize somewhat different theoretical aspects, and operational measures vary. Always check how a study defines and computes its score.
Can someone experience two emotions at once and still have high granularity?
Yes. Joy and sadness, relief and apprehension, or anger and guilt may coexist. Differentiation concerns whether these states can be represented distinctly and in context, not whether a person reports one label exclusively.
Is high emotional granularity always beneficial?
Studies report associations with favorable outcomes in some domains, but benefits are context-dependent, effect sizes are often modest, and immediate experiences may not become more pleasant. Long-term causal effects and optimal conditions are still being studied.
How do researchers test it?
Often through multiple daily emotion ratings collected by phone, followed by within-person statistical analyses such as an intraclass-correlation-based differentiation index. Study design and scoring choices strongly influence interpretation.
Can I measure my emotional granularity with a single online quiz?
A single questionnaire may measure beliefs about emotional clarity or perceived differentiation. It cannot be assumed equivalent to a validated repeated-assessment granularity index, and neither is a diagnostic test.
Can improving my emotion vocabulary help?
Emotion-knowledge training has produced some encouraging experimental findings, particularly for negative differentiation, but knowing words is only one part of the construct and the evidence does not establish a universal clinical benefit.
Does low granularity mean alexithymia, autism, depression, or anxiety?
No. Low measured differentiation is not a stand-alone clinical diagnosis or a diagnostic indicator of any of these conditions. Related constructs and clinical presentations require distinct assessments and careful interpretation.
Why do positive and negative granularity receive different scores?
Different sets of emotions may be distinguished to different degrees. Studies also vary in which positive and negative terms they include, so statistical scores need to be interpreted within the actual measurement design.
Does labeling an emotion make it disappear?
No general rule supports that conclusion. A clear label can coexist with a powerful emotional reaction. Granularity research mainly concerns differentiation and its associations with functioning, not guaranteed immediate emotional relief.
Conclusion
Emotional granularity captures a scientifically important feature of how people represent emotion: how distinctly different feelings are experienced and described in context. It can be studied through repeated real-world reports, but each operational index brings assumptions and limitations. The strongest defensible interpretation of the current literature is that granularity is associated with several aspects of psychological functioning, and experimental findings suggest it may be modifiable. Causal effects on health, lasting clinical benefits, and the best measurement methods remain open research questions.
For further scientific context, see What Are Emotions? Psychological Definition, Components, and Functions.
Related Articles
For the broader study of psychological response and flexibility, see Neuroticism and Emotion Regulation: Strategies, Flexibility, and Distress and Sleep and Emotional Regulation: Effects on Mood and Stress. These are neighboring topics with distinct scientific and search intents. For a broader framework of emotion categories, see Types of Emotions: Basic, Complex, and Social Categories.
References
Barrett, L. F. (2017). The theory of constructed emotion: An active inference account of interoception and categorization. Social Cognitive and Affective Neuroscience, 12(1), 1–23. PubMed.
Erbas, Y., Ceulemans, E., Blanke, E. S., Sels, L., Fischer, A., & Kuppens, P. (2019). Emotion differentiation dissected: Between-category, within-category, and integral emotion differentiation, and their relation to well-being. Cognition & Emotion, 33(2), 258–271. https://doi.org/10.1080/02699931.2018.1465894
Hoemann, K., Barrett, L. F., & Quigley, K. S. (2021). Emotional granularity increases with intensive ambulatory assessment: Methodological and individual factors influence how much. Frontiers in Psychology, 12, 704125. https://doi.org/10.3389/fpsyg.2021.704125.
Lo, T. T., Verhagen, M., Pouwels, J. L., van Roekel, E., O’Brien, S. T., Debra, G., Braet, J., Vink, J. M., & Maciejewski, D. F. (2025). Emotion differentiation in adolescents: Short-term trade-offs with regulation variability and emotion intensity. Affective Science, 6, 243–258. https://doi.org/10.1007/s42761-025-00301-4.
Nook, E. C. (2021). Emotion differentiation and youth mental health: Current understanding and open questions. Frontiers in Psychology, 12, 700298. https://doi.org/10.3389/fpsyg.2021.700298.
Ozomaro, B., Wein, P. Y., Miller, S., Ospina, L. H., Goodman, M., Torous, J., & Kimhy, D. (2025). Emotional granularity in health and psychopathology: A scoping review. Journal of Psychiatric Research, 190, 277–295. https://doi.org/10.1016/j.jpsychires.2025.08.008.
Schmitt, M. C., Karbach, J., Könen, T., Basten, U., Glombiewski, J. A., In-Albon, T., & Lischetzke, T. (2025). Executive functions and emotional granularity: No evidence for positive associations. Affective Science, 6, 464–476. https://doi.org/10.1007/s42761-025-00307-y.
Seah, T. H. S., & Coifman, K. G. (2022). Emotion differentiation and behavioral dysregulation in clinical and nonclinical samples: A meta-analysis. Emotion, 22(7), 1686–1697. https://doi.org/10.1037/emo0000968.
Smidt, K. E., & Suvak, M. K. (2015). A brief, but nuanced, review of emotional granularity and emotion differentiation research. Current Opinion in Psychology, 3, 48–51. Journal article.
Tan, T. Y., Wachsmuth, L., & Tugade, M. M. (2022). Emotional nuance: Examining positive emotional granularity and well-being. Frontiers in Psychology, 13, 715966. https://doi.org/10.3389/fpsyg.2022.715966.
Vedernikova, E., Kuppens, P., & Erbas, Y. (2021). From knowledge to differentiation: Increasing emotion knowledge through an intervention increases negative emotion differentiation. Frontiers in Psychology, 12, 703757. https://doi.org/10.3389/fpsyg.2021.703757.
Wilson-Mendenhall, C. D., & Dunne, J. D. (2021). Cultivating emotional granularity. Frontiers in Psychology, 12, 703658. https://doi.org/10.3389/fpsyg.2021.703658.