When “emotion regulation difficulty” is too vague to measure
A person can know exactly what they feel and still be unable to keep working while upset. Another can remain behaviorally controlled but have almost no language for the emotional state. A third can understand the emotion and accept its presence yet feel that there is nothing effective they can do with it.
Calling all three cases “poor emotion regulation” hides more than it explains.
The Difficulties in Emotion Regulation Scale, or DERS, was created to make that broad phrase more multidimensional. Rather than asking only how often someone uses one strategy, it assesses several kinds of self-reported difficulty that can appear when emotion is intense.
Its influence has been substantial in clinical and research settings. So has the temptation to read a total score as though it captured a single underlying deficit. The measure deserves more careful treatment.
What the DERS was designed to measure
Kim Gratz and Lizabeth Roemer introduced the 36-item DERS in 2004 alongside an integrative conceptualization of emotion regulation. Their framework emphasized more than reducing arousal. It included awareness and understanding of emotion, acceptance of emotional responses, the ability to pursue desired behavior while distressed, impulse control, and access to strategies perceived as effective.
The original factor structure yielded six commonly reported domains: nonacceptance of emotional responses; difficulties engaging in goal-directed behavior; impulse-control difficulties; lack of emotional awareness; limited access to emotion-regulation strategies; and lack of emotional clarity.
These labels are easy to over-literalize. A subscale is an operationalization, not an independent biological system. The factors arise from patterns in responses to questionnaire items. They should be interpreted as measurement dimensions within a particular model, not as six objectively separate modules inside a person.
Administration and scoring
The original DERS contains 36 self-report items. Respondents indicate how often statements apply to them. Researchers commonly calculate a total score and six subscale scores, with higher scores indicating greater reported difficulty under the scale's scoring convention.
The DERS-16, developed by Johan Bjureberg and colleagues, offers a shorter form. It retains five domains represented in the original model: clarity, goals, impulse, strategies, and nonacceptance. The awareness domain is not part of the DERS-16. That makes the short form useful when assessment burden matters, but it also means the shorter instrument is not simply the original scale with fewer redundant items. Its content coverage differs.
As with any measure, scoring procedures, reverse-coded items, missing-data rules, and current use permissions should be checked against authoritative documentation. This page does not reproduce item text.
The original validation and its boundary
Gratz and Roemer's original development work used two undergraduate samples. The paper reported high internal consistency, good test-retest reliability, and evidence of construct and predictive validity. Those findings were sufficient to establish an influential research tool, but they were an initial validation rather than the final word on factor structure across all populations.
The original paper later received an erratum noting item similarity or identity with material from the Negative Mood Regulation scale. That history belongs in a serious methods page because provenance is part of psychometric trust. It does not invalidate the DERS. It does remind users that measurement instruments have textual and intellectual histories that should be documented rather than treated as frictionless public objects.
The DERS-16 and Swedish research contribution
The DERS-16 was developed by a team that included researchers at Karolinska Institutet and Lund University, alongside international collaborators. Validation included a clinical sample of 96 participants and community samples of 102 and 482 participants. The authors reported excellent internal consistency, good test-retest reliability, and convergent and discriminant validity, with minimal differences from the full DERS on many relevant associations.
This is a substantive Nordic contribution to the measurement literature. It also illustrates a common psychometric trade-off: shorter measures reduce burden but necessarily change content density. A 16-item measure can be highly useful without replacing every purpose served by a 36-item measure.
Sörman and colleagues later examined the DERS-36, DERS-16, and ERQ in a Swedish community sample of 843 participants. In that study, the DERS-16 and ERQ showed adequate model fit while the hypothesized DERS-36 structure fit less well. Both DERS versions related more strongly to depression and anxiety measures than the ERQ strategy subscales did. That result supports the idea that DERS and ERQ should not be treated as interchangeable “emotion regulation scores.”
Reliability is not construct certainty
DERS studies frequently report high internal consistency. That is useful, but a high alpha or omega does not settle what a scale means. A group of items can cohere because they share substantive content, general distress, negative self-evaluation, response style, or overlapping wording.
The DERS is particularly vulnerable to a broader measurement problem in psychopathology: measures of difficulty can correlate with symptom severity partly because both ask people to report distressing experiences in negative language. Strong correlations can reflect genuine theoretical relationships and shared method variance at the same time.
For this reason, discriminant validity and factor structure matter. So do longitudinal and behavioral designs capable of examining whether DERS scores predict outcomes beyond general distress.
Construct validity and the meaning of “strategies”
One of the original subscales is commonly labeled limited access to emotion-regulation strategies. The wording can sound like an objective inventory of how many strategies a person possesses. It is better understood as perceived access or perceived efficacy as captured by the scale's items.
A person may know many strategies but believe none will work when distressed. Another may know few formal strategies yet regulate effectively through relationships, movement, task structure, or environmental change. The questionnaire cannot directly inventory the whole behavioral repertoire.
Similarly, “lack of emotional awareness” should not be equated with a clinical diagnosis or with absence of interoception. Awareness is a broad construct, and the DERS operationalizes one slice of it.
Populations and measurement invariance
DERS versions have been used across many clinical and nonclinical populations and translated into many languages. Widespread use is not itself proof of invariance. Researchers comparing groups should examine whether factor structure and item functioning support the comparison they intend to make.
Age is especially important. The original DERS was developed with adults/undergraduates, while adapted versions exist for adolescents and other populations. A score derived from one form should not be treated as directly interchangeable with another without evidence.
Clinical samples also require care. High DERS scores can be common across multiple diagnoses. That transdiagnostic pattern can be informative, but it does not make the DERS a diagnostic instrument.
Confounds and common interpretive errors
The first error is equating total score with a fixed regulatory capacity. Self-report difficulty can vary with current symptoms, stress, insight, expectations, and context.
The second is treating subscales as causal mechanisms. A correlation between nonacceptance and symptoms does not prove that nonacceptance caused the symptoms.
The third is reading high scores as emotional intensity. The DERS concerns difficulty around emotion, not simply how strongly a person feels.
The fourth is assuming a lower score is always preferable. Some questionnaire items reflect capabilities that are generally valuable, but a total score is not a moral ranking and does not capture contextual trade-offs.
The fifth is using DERS change as proof that a treatment's mechanism is emotion regulation. A mediator claim requires temporal and causal evidence beyond pre-post association.
Alternatives and complements
The ERQ is narrower and strategy-focused, measuring habitual reappraisal and expressive suppression. Experience sampling can examine regulation and emotion in daily contexts. Behavioral tasks can test persistence or inhibition under emotional load. Clinical interviews can clarify what “difficulty” means in a person's life. Physiological measures can characterize arousal but do not translate directly into regulatory success.
The best choice depends on the research question. If the question concerns broad perceived difficulties, DERS may be appropriate. If it concerns two specific strategy tendencies, ERQ is more direct. If it concerns moment-to-moment dynamics, repeated in-situ measurement may be more informative.
What the evidence does not show
The DERS literature does not show that one total score exhaustively measures emotion-regulation capacity. It does not show that high scores diagnose a disorder. It does not show that each subscale is a separate causal mechanism. It does not show that reducing a DERS score will necessarily improve functioning, or that observed treatment-related score changes reveal the mechanism by which treatment worked.
Psychometric validation in one language or population also does not prove measurement equivalence everywhere else.
Institute synthesis: difficulty is not the same as incapacity
The following is Institute synthesis.
For Inner Technology, the DERS offers a useful reminder that capacities can fail in different places. A person may have emotional awareness without acceptance, acceptance without impulse control, strategy knowledge without access under pressure, or clarity without the ability to continue goal-directed action.
That multidimensionality is more useful than a single label such as “regulated” or “dysregulated.” At the same time, the Institute would not equate self-reported difficulty with the capacity itself. Capacity should ideally be examined across self-report, context, action, persistence, recovery, and consequences.
This distinction becomes especially important in institutions. If an organization uses a scale to evaluate a programme, the outcome must match the intervention's actual target. A general DERS score should not become a proxy for employee worth, student readiness, or clinical diagnosis.
Used with that restraint, the measure can help locate where people report friction. It should not be asked to decide who they are.
Practical interpretation checklist
Use the correct DERS version. Report the population and language. Explain whether total or subscale scores are primary. State sample-specific reliability. Do not assume the six-factor structure fits every sample. Distinguish perceived difficulty from observed behavior. Treat symptom correlations as associations unless the design supports stronger inference. Verify licensing and source provenance before reproducing items.
Total score versus profile
The DERS total score is often attractive because it reduces complexity to one number. That can be useful for statistical modelling, but it also hides pattern. Two participants can have the same total while reporting very different constellations of difficulty. One may struggle mainly with impulse control and goal-directed behavior under distress. Another may report nonacceptance and poor clarity while remaining behaviorally controlled.
Whether subscale interpretation is justified depends on the version, factor structure, reliability, and research purpose. Profiles should not be treated as diagnostic phenotypes simply because six labels are available. The psychometric question is whether the subscales carry enough distinct, reliable information in the relevant sample to support the interpretation being made.
This is particularly important in small studies where subscale estimates can be noisy. A richly labeled radar chart can imply more precision than the data support.
The overlap with general distress
Measures of psychological difficulty often share variance. Someone who is currently depressed, anxious, sleep-deprived, or overwhelmed may endorse many negatively framed items across several questionnaires. That can produce substantial correlations even when the constructs are theoretically distinct.
Researchers can address this in several ways: include measures of general distress; test incremental validity; use longitudinal designs; compare self-report with behavior or ecological data; and model shared versus specific variance. None is perfect. Together they help determine whether DERS associations reflect emotion-regulation difficulty specifically or a broader negative-affectivity factor.
The distinction has clinical consequences. If a treatment reduces symptoms and DERS scores at the same time, both changes may be meaningful. It still requires stronger temporal evidence to conclude that the regulation change caused the symptom change.
Short forms solve burden, not every validity problem
The DERS-16 is valuable because participant burden matters. Long questionnaires can increase missingness, fatigue, patterned responding, and dropout, especially when several instruments are administered together. A short form can improve the quality of a larger protocol.
But shortening does not make the construct simpler. The DERS-16's omission of the awareness dimension means researchers should state clearly what is no longer represented. Comparisons between DERS-16 totals and DERS-36 totals should be made with evidence rather than convenience.
There are also other shortened or adapted forms in circulation. Instrument choice should be driven by validation in the intended population and by the conceptual domains required, not by whichever PDF is easiest to find online.
Clinical usefulness and the danger of thresholds
The DERS was not originally built around a universal diagnostic cut-off. A score can be useful continuously without defining a boundary between normal and disordered regulation. When studies propose thresholds for a specific sample or clinical purpose, those thresholds should not migrate into general use without validation.
This matters for digital screening. A platform can easily convert a continuous score into green, amber, and red categories. The interface may make the categories feel authoritative even when the underlying validation does not support them. Measurement design should resist that visual inflation of certainty.
Repeated measurement can change what the score means
When DERS is administered repeatedly during treatment or training, participants may learn the language of the instrument. They may become better at recognizing difficulty, more critical of patterns they previously normalized, or more confident in their ability to respond. That means a change score can reflect both underlying change and improved self-observation.
Researchers can partly address this by pairing self-report with behavioral or ecological measures and by interpreting trajectories rather than a single pre-post difference. In capacity research, increased awareness can sometimes make self-ratings look worse before behavior improves. A measurement system should be able to tolerate that possibility.
Translation is conceptual adaptation, not clerical work
DERS items refer to experiences such as clarity, control, acceptance, and perceived access to effective action. Those concepts carry ordinary-language meanings that can shift in translation. A technically literal translation can preserve grammar while altering emotional connotation.
Good adaptation therefore uses forward translation, back-translation or equivalent expert procedures, participant testing, and psychometric evaluation in the target population. If a Swedish translation performs well in one sample, that supports its use there; it does not remove the need to examine functioning in a different age, clinical, or cultural group.
The same discipline applies to digital implementations. Changing response labels, shortening instructions, or reformatting items can alter the instrument. A user interface is part of administration, not a neutral wrapper.
Research references
Gratz, K. L., & Roemer, L. (2004). Multidimensional assessment of emotion regulation and dysregulation: Development, factor structure, and initial validation of the Difficulties in Emotion Regulation Scale. Journal of Psychopathology and Behavioral Assessment, 26, 41–54. https://doi.org/10.1023/B:JOBA.0000007455.08539.94
Gratz, K. L., & Roemer, L. (2008). Erratum. Journal of Psychopathology and Behavioral Assessment. https://doi.org/10.1007/s10862-008-9102-4
Bjureberg, J., Ljótsson, B., Tull, M. T., et al. (2016). Development and validation of a brief version of the Difficulties in Emotion Regulation Scale: The DERS-16. Journal of Psychopathology and Behavioral Assessment, 38, 284–296. https://doi.org/10.1007/s10862-015-9514-x
Sörman, K., et al. (2022). Measures of emotion regulation: Convergence and psychometric properties of the Difficulties in Emotion Regulation Scale and Emotion Regulation Questionnaire. Journal of Clinical Psychology, 78, 201–217. https://doi.org/10.1002/jclp.23206