Historically, the concepts of emotion and emotional regulation have been widely debated, as well as what characterizes the difficulties in emotional regulation, also known as emotional dysregulation. Emotions can be defined as transient changes in an individual’s subjective experiences, behaviors, and physiological responses, resulting from motivationally relevant internal or external stimuli (Gross, 2015; Lang, 1995). According to the functional perspective, emotions have an adaptive purpose, that is, they are promoters of survival (Kehoe & Havighurst, 2020).
The ability to manage emotions and express them appropriately requires skills in identifying, understanding, and regulating emotions in both intra- and interpersonal situations (Halberstadt et al., 2001). Emotion regulation involves monitoring, evaluating, and modifying emotional reactions (Thompson, 1994). It can be defined as a process that aims to intensify, decrease, or maintain the behavioral, cognitive, experiential, or physiological valences of emotion, depending on the subject’s objectives (Gross & Thompson, 2007). The effectiveness of emotion regulation implies the selection of appropriate strategies and flexibility in their use, which in itself means psychological adjustment (Campbell-Sills & Barlow, 2006). Deficits in emotion regulation are associated with psychopathology symptoms, other emotion-related constructs, and therapy progress (Victor & Klonsky, 2016).
Recognizing the importance of the functional role of emotions, Gratz and Roemer (2004) proposed a multifactorial conceptualization of emotion regulation, formed by the following dimensions: consistency and understanding of emotions; acceptance of emotions; ability, in the face of negative emotions, to control impulsive behaviors and act in line with the desired goals; and ability to use flexible and appropriate emotion regulation strategies, modulating emotional responses to achieve individual goals and deal with the demands of the situation. The relative absence of any or all these capacities may indicate the presence of difficulties in emotion regulation or emotion dysregulation. Usually, emotion regulation skills increase with age (Orgeta, 2009). However, difficulties in regulating emotions can occur throughout the life cycle (Kaufman et al., 2016).
Two different approaches to the concept of emotion regulation supported the development of self-report instruments: (1) emotion dysregulation responses in the form of emotional sensitivity, reactivity, and/or intensity and (2) maladaptive responses to emotions, regardless of the characteristics of the emotions (Gratz et al., 2020). The Difficulties in Emotion Regulation Scale (DERS) developed by Gratz and Roemer (2004) is based on the second perspective, which emphasizes the functional nature of emotions. There are other measures to evaluate this construct (e.g., Catanzaro and Mearns Generalized Expectancy for Negative Mood Regulation Scale; Trait Meta-Mood Scale; Emotional Regulation Questionnaire), but they all assess how individuals’ internal experiences impact their affective response and only reflect a single aspect or subset of emotions (Mekawi et al., 2021). Conversely, the DERS was designed to assess trait-level perceived emotion regulation ability in a multidimensional and comprehensive regulation perspective (Gratz & Roemer, 2004).
The DERS (Gratz & Roemer, 2004) is an extensively used instrument in different populations worldwide, covering adults and adolescents and presenting comprehensive empirical support (Charak et al., 2019). This self-response scale aims to assess clinically significant emotion regulation difficulties through 36 items that fall into 6 domains: non-acceptance of negative emotions (Non-Acceptance), inability to engage in goal-oriented behaviors when experiencing negative emotions (Goals), difficulties in controlling impulsive behaviors when experiencing negative emotions (Impulses), restricted access to emotion regulation strategies perceived as effective (Strategies), lack of emotional awareness (Awareness), and lack of emotional clarity (Clarity).
Numerous studies have confirmed this six-factor structure (Fowler et al., 2014; Gratz & Roemer, 2004; Neumann et al., 2010), although more recently, it has been suggested that items of the Awareness factor could be removed due to their lower validity and consistency (Hallion et al., 2018; Miguel et al., 2017; Osborne et al., 2017).
Contributing to the strength of the convergent validity, it was found that the emotional difficulties measured by this scale are significantly associated with a multiplicity of behaviors, such as self-mutilation, domestic violence, binge eating, substance abuse, and risky behaviors; psychopathological conditions, including borderline personality disorder, post-traumatic stress disorder, anorexia nervosa, social anxiety disorder, and behavioral disorder; and with countless other constructs in the scope of mental functioning, namely negative affect, the severity of depressive and anxious symptoms, tolerance to distress, experiential avoidance, and self-compassion (for a review see Gratz et al., 2020).
Given its relevance, DERS was translated and validated for several countries, originating multiple versions, including a Portuguese-European version (Coutinho et al., 2010) and a Portuguese-Brazilian version (Miguel et al., 2017), both with good psychometric qualities. Despite the wide use of DERS in both clinical and research contexts, its length represents an important limitation, given the expected burden effects on respondents and the difficulties associated with time constraints, particularly in assessments with close intervals and in large epidemiological studies (Kaufman et al., 2016; Shahabi et al., 2020). In addition, due to the similarity of some items, these tend to be perceived as repetitive, increasing fatigue and frustration, which advises the use of short instruments that can be equally effective in evaluating the same construct (Kaufman et al., 2016).
Considering these factors, three short versions of the DERS were developed: the Difficulties in Emotion Regulation Scale – Short Form (DERS-SF) (Kaufman et al., 2016), the DERS-18 (Victor & Klonsky, 2016), and the DERS-16 (Bjureberg et al., 2016). When comparing the DERS short forms (DERS-SF, DERS-18, DERS-16), all the three scales showed strong concordance with the original extended version, internal consistency fair-to-good, and reliability above 0.80 for all subscales except Awareness (Hallion et al., 2018). The authors of this comparative study reported no evidence that any of the short forms were psychometrically superior to the others. However, short versions that retained subscale scores (i.e., DERS-SF and DERS-18) have shown strong concurrent validity, given their capacity to predict current symptoms of anxiety and depression (Skutch et al., 2019). Considering that the short forms generally performed similarly to the original DERS, despite a slight loss of predictive utility (1–3% of the variance) to explain clinical severity, it was suggested that its use is acceptable in most clinical and research situations, with the long form of DERS being indicated for when a comprehensive assessment is required (Hallion et al., 2018; Shahabi et al., 2020). The DERS-SF has been the most used (Skutch et al., 2019) and the most recommended version intended to investigate emotion dysregulation throughout the life cycle, as it is the only scale that presents invariance among adolescents and adults (Charak et al., 2019). The DERS-SF is composed of half of the items in the original version and was developed using two adult and three adolescent samples, verifying the maintenance of the six-factor structure and the excellent psychometric properties of the original instrument, regardless of the age variation.
As with DERS, these short instruments also need to be subject to cross-cultural adaptations and the respective study of their psychometric properties. Recently, a study carried out in Portugal by Moreira et al. (2020), with an adolescent and adult community women sample and aiming to contribute to clarifying the adequacy of the DERS-SF, suggested a bifactorial model without the Awareness subscale.
Pointing out inconsistencies in the literature about the multidimensionality of DERS and its short forms (e.g., Hallion et al., 2018; Osborne et al., 2017), Moreira et al. (2020) tested several models for the scale structure. Between the first group of models tested, a correlated six-factor model and a bifactorial model exhibited a good fit to the data. Nevertheless, the former presented high correlations between all factors except for the correlations between Awareness and all the other factors, and the second had some problems with some items that had non-significant loading or weak factor loading. Considering the results of the first models, they tested a second set of models but removed the Awareness subscale. The correlated six-factor model without Awareness had a good fit to the data, but the bifactorial model without Awareness had a very good and better internal consistency of the subscales. Moreira et al. (2020) suggested that the items on the Awareness subscale should not be included in calculating the total value of the DERS-SF. The authors suggested that the Awareness subscale may be interpreted as a unique measure that evaluates difficulties in processing emotions, and not in the regulation of emotional responses.
Research is moderately extensive but far from conclusive concerning the differences between men and women in terms of emotion regulation and their difficulties. Nolen-Hoeksema and Aldao (2011) found that compared to men, women reported a wider range of emotion regulation strategies, including acceptance, social support, problem-solving reassessment, and rumination, being more likely to resort to more adaptive strategies at older ages. Furthermore, Gratz and Roemer (2004) during the validation of DERS found that men had greater difficulties in becoming emotionally aware. There were differences in the Clarity domain in the DERS validation studies for the Portuguese population, with men showing more difficulties recognizing their emotions (Veloso et al., 2011). However, this finding was not corroborated by a recent study, which found a greater lack of emotional clarity in female participants (Shahabi et al., 2020). In summary, gender differences have been documented in some DERS subscales, albeit sometimes inconsistently. It should be noted that these differences are often not found in the instrument’s overall score (Gratz & Roemer, 2004; Tull et al., 2012). However, this analysis has not yet been carried out using the DERS-SF. At the same time, considering the articulation between gender differences and factorial invariance, and considering that this has never been investigated, it is pertinent to explore gender invariance.
Based on a sample of adults from the general population, the present study aims to (a) analyze the factorial structure and psychometric properties of the Portuguese version of DERS-SF (Kaufman et al., 2016) and (b) analyze the DERS-SF factor structure invariance between men and women.