Introduction

Emotion regulation is an essential process in the daily life of human beings that allows us to use different strategies to modify the course, intensity, duration and expression of emotional experiences depending on the situation or our goals [1]. Research on cognitive strategies in emotion regulation, regardless of behavioural factors and the influence of these on the presence of psychopathology, has become of great importance. These strategies can be defined as conscious cognitive processes aimed at controlling and regulating emotions in threatening situations [2].

As children grow older, their repertoire of emotion regulation strategies gradually increases, from the use of primarily behavioural strategies to cognitive strategies [3]. There is evidence that childhood is an important stage in which children's emotional control increases, reducing negative emotions over the years [4]. However, the type of cognitive emotional regulation (CER) strategies developed in infancy will have repercussions for the style of cognitive coping used for stressful situations in adolescent and adult life, where CER strategies have been widely related with the presence of psychopathology [5]. As in adults, the use of certain CER strategies are strongly related to the presence of emotional problems in children (i.e., anxiety and depression), having these strategies a potential value as an indicator of child psychopathology and as a target for the prevention and treatment of these problems [6, 7]. In this regard, it has been stressed that the development of adaptive CER strategies during childhood can have an important protective effect and reduce the risk of future disorders [8]. Therefore, the assessment of CER strategies can be important as it can be useful for the diagnosis process of childhood emotional problems and for establishing effective prevention or intervention strategies [8].

There are several instruments that evaluate emotion regulation processes, such as the Difficulties in Emotion Regulation Scale (DERS) [9], Emotion Regulation Questionnaire (ERQ) [10], Trait Meta-Mood Scale (TMMS) [11] and the Negative Mood Regulation Scale (NMR) [12]. However, the Cognitive Emotion Regulation Questionnaire (CERQ) [13] is the only questionnaire that focuses on evaluating purely cognitive strategies of emotion regulation, without encompassing the broad repertoire of intrinsic and extrinsic strategies for control, evaluation and modification of emotions.

The CERQ was initially developed by Garnefski et al. [13] as a self-report tool to evaluate the cognitive strategies of emotion regulation that the person uses in response to threatening or stressful situations. This multidimensional questionnaire consists of 36 items that are answered on a 5-point Likert scale, in which 1 corresponds to "almost never" and 5 to "almost always." The nine scales that comprise this instrument have 4 items each and correspond to the following cognitive strategies: a) Self-blame: thoughts that attribute the cause of the negative event and the emotion to oneself; b) Acceptance: to resign and accept the irreversibility of the negative experience; c) Rumination: state of excessive worry by negative thoughts and feelings; d) Positive focusing: having pleasant and joyous thoughts different from the negative event; e) Planning: thinking about how to solve the problem; f) Positive reappraisal: thoughts that highlight some positive aspect of the unpleasant event; g) Putting into perspective: decrease and relativize the severity of the event; h) Catastrophizing: to think about the horrible thing of what happened and conclude that it is the worse experience lived, even compared with what other people have experienced; and i) Other-blame: make others responsible for the negative event that happened.

In the Spanish adaptation of the CERQ, Domínguez-Sánchez, Lasa-Aristu, Amor, and Holgado-Tello [14] proposed an alternative model in which these nine cognitive strategies constitute the first order factors and can be grouped into two second order factors: adaptive strategies (putting into perspective, acceptance, positive reappraisal, positive refocusing, and planning) and maladaptive strategies (rumination, catastrophizing, self-blame, and other-blame). From this classification, also raised by the authors of the original CERQ [13], numerous studies have found that the use of maladaptive strategies are strongly negatively related to the symptoms of anxiety and depression, both in adult and adolescent populations [15–17] and in children [6, 7, 18]. In this regard, a recent review found that adaptive emotional regulation strategies were negatively associated with symptoms of anxiety and depression in youth, while maladaptive strategies were positively associated, showing that the presence of more symptoms was associated with the use of maladaptive strategies and fewer symptoms with the use of adaptive strategies [5].

The original version of the CERQ was applied to a sample of adolescents and adults from the Netherlands [13, 19]. Subsequently, the CERQ has been translated and adapted in different countries such as France [20], China [21], Romania [22], Iran [23], Turkey [24], Spain [14], Argentina [25] and Peru [26], among others. However, these adaptations focused on the adult population. Both the original version of the CERQ and the reported adaptations have demonstrated good psychometric properties, as the authors identified acceptable levels of internal consistency reliability for each dimension, and evidence of criterion validity consistent with measurements of depression, anxiety, and positive and negative emotions; and all are consistent with the nine-factor structure.

Several studies with samples of adolescents have shown good psychometric properties: maintaining the original factor structure, internal consistency reliability coefficients ranging from .68 to .83, and a test-retest stability with Pearson r values between .40 and .60 [2, 13, 19]. Although the CERQ was first developed with adult and adolescent populations, a version adapted for children under 12 was created, the CERQ-k [6]. This version and its adaptations to other languages maintain the original factor structure and have shown adequate internal consistency, with Cronbach's alphas between .62 and .79 [6, 8]. Despite being an instrument of great clinical and research utility to detect cognitive patterns of emotion regulation that predict the presence of emotional symptomatology in children, the measure has not been validated for Spanish-speaking children. As such, the present research aimed to provide a new tool for the evaluation of cognitive strategies of emotion regulation for use with Spanish-speaking children.

The main objective of the present study was to adapt and validate the CERQ-k in a community sample of children between 7‒12 years of age in Spain. To achieve this aim, the factor structure was analysed using confirmatory factor analysis (CFA), internal consistency and test-retest reliability of each dimension of the CERQ-k were examined, and convergent validity was evaluated. Based on previous studies [6–8], it was hypothesized that the psychometric properties of the Spanish version of the CERQ-k would be adequate, and the nine factors corresponding to the nine cognitive strategies of coping would be confirmed.

Methods

Results

Confirmatory factorial analysis

CFA for the Spanish sample was carried out (N = 582) and yielded a nine-factor model coinciding with the original version of the CERQ-k [6]. The adjustment for the model was adequate: CFI = .94, TLI = .93, RMSEA = .05 (.052, .058). According to the criteria of Hu and Bentler [32], the CFA showed suitable fit for the 36-item measure because the indexes were greater than .90. The factors of the Spanish version of the CERQ are the same as the nine proposed by the original authors. Each factor is composed of 4 items, namely Self-Blame, Acceptance, Rumination, Positive Refocusing, Planning, Positive Reappraisal, Putting into Perspective, Catastrophizing and Other-Blame. All the items showed factor loadings greater or equal to .50, with the exception of items 18, 19 and 31(factor loadings were .49, .47 and .46, respectively) (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Confirmatory factor analysis: factor loadings (N = 582).

https://doi.org/10.1371/journal.pone.0201656.t001

Analysis of items

Descriptive statistics of the items of the Spanish version of the CERQ-k were calculated. The corrected item-total correlations ranged from .22 to .68, which indicates a general adequate performance of the items. All item-correlations were above .30, except for item 20 (“I think that I can’t change it”). Since the removal of any item did not increase the ordinal alpha significantly, all items were kept. Table 2 shows the ordinal alphas, means, standard deviations and ranges for each questionnaire in the current study.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Scale properties of the CERQ, CDI and Trait Anxiety Inventory (STAI-C).

https://doi.org/10.1371/journal.pone.0201656.t002

Internal consistency and test-retest reliability

The CERQ-k demonstrated acceptable internal consistency with coefficients ranging from .56 (Acceptance) to .75 (Positive Refocusing) for the nine scales (Table 2). Spearman correlations among the nine factors were calculated, and low values ranging from -.01 (Other-Blame and Positive Reappraisal) to .54 (Planning and Positive Reappraisal) were found. Most of these correlations were significant at p < .01, as shown in Table 3.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Spearman correlations among CERQ scales and depression and anxiety problem scales.

https://doi.org/10.1371/journal.pone.0201656.t003

A subsample of 211 children (36.25% of the sample) was asked to complete the CERQ-k 8 weeks later in order to examine test-retest reliability, according to Domínguez-Sánchez et al. [14] in the Spanish validation of the CERQ for adults. ICC test-retest coefficients were low to moderate with the following values: .65 for Self-Blame, .54 for Acceptance, .70 for Rumination, .67 for Positive Refocusing, .60 for Planning, .65 for Positive Reappraisal, .63 for Putting into Perspective, .63 for Catastrophizing, and .56 for Other-Blame. These ICC coefficients indicated that the test-retest reliability was acceptable, except for the Acceptance subscale, whose value did not reach the recommended cut off [44]. The test-retest coefficient for the total scale score was adequate (ICC = .74).

Discussion

The current study tested the psychometric properties and factor structure of the Spanish version of the CERQ-k in a community sample of children aged from 7 to 12 years old. The CFA demonstrated an excellent fit and confirmed the original nine-factor structure of the CERQ-k [6] and the adapted Chinese version for children [8]. The CER strategies are self-blame, acceptance, rumination, positive refocusing, planning, positive reappraisal, putting into perspective, catastrophizing, and other-blame. Thus, the Spanish version of the CERQ-k maintains the same nine scales as the original 36-item version [2, 6] that has also been validated with adult populations across several countries: France [20], China [21], Romania [22], Iran [23], Turkey [24], Spain [14], Argentina [25] and Peru [26], among others. According to Field [45], when item-correlation coefficient is above .30 is expected a positive correlation between that specific item and the factor. However, a low item-correlation indicates that the item does not measure the same construct than other items included in the same factor. Given the item-correlation coefficients for the CERQ-k were above .30 (except for item 20) and factor loadings had medium-to-high magnitude, all items were retained.

The internal consistency reliability of the subscales was generally good, with alphas ranging from .56 to .75. These results are similar to other validation studies that have used the CERQ-k in a sample of 9-11-year-old 717 primary school youngsters in the Netherlands (from .62 to .79) [6, 7]. When the scale presents multidimensional nature, such as CERQ-k that includes four items per factor, it is reasonable that internal consistency is moderate-adequate [45]. Test-retest reliability was also examined over an interval of 8 weeks, and data established a satisfactory total scale test-retest reliability index (ICC = .74) and moderate to acceptable correlations for the subscales (ICC = .54 to .70), which suggests that the CER strategies can be considered relatively stable styles, although less than others such as personality traits [6]. Consistently, Liu et al. [8] found test-retest coefficients ranging from .53 to .70 over one month. This finding also suggests that CER strategies may vary depending on the situations children face and the context in which the unpleasant event unfolds [13].

In addition, the construct validity of the Spanish version of the CERQ-k was verified. Consistent with empirical studies using CERQ-k in children [14, 15, 25] and adults [17, 26] and a meta-analytic review on emotion-regulation strategies across psychopathology [15], the subscales corresponding to less adaptive CER strategies (Rumination, Self-Blame, Catastrophizing, and Other-blame) were more related to depression and anxiety symptoms, supporting the convergent validity of the CERQ-k. The results suggest that children who use maladaptive CER strategies (including Rumination, Self-Blame, Catastrophizing, and Other-blame) are more likely to develop depressive and anxiety symptomatology. Convergent validity was also demonstrated by the relationships between the adaptive CER strategies (positive refocusing and planning) and presence of fewer symptoms of depression. Therefore, positive refocusing and planning would act as strategies that have a positive effect on the prevention of depression in children. The reason why no further stronger negative correlations were found between adaptive strategies, anxiety and depression remains unclear. However, our findings are consistent with existing studies conducted with Chinese [8], Dutch [6, 13], and Spanish-Speaking children and youth [18], which found the strongest negative associations with anxiety and/or depression only with positive refocusing and/or positive reappraisal. Only the study with a Chinese sample [8] has found significant negative associations between most adaptive CER strategies and depression, including positive refocusing and planning as found in our study. In this regard, it is necessary further research with children from different settings to determine possible cultural differences in the concept of cognitive emotional regulator, the use of CER strategies, and their impact on emotional problems since cognitive coping styles may be more or less adaptive depending on the circumstances in which they are assessed [13]. In adults, the research is more extensive and the cross-cultural stability between the presence of these emotional problems and the use of CER strategies has been highlighted, and only minor differences have been found [14]. For instance, as Domínguez-Sánchez et al. [14] noted, compared to individualistic societies (e.g., Spain), in collectivistic societies (e.g., China) there has been found a greater relationship between anxiety and blaming others or a greater use of strategies such as positive refocusing instead of positive reappraisal. Lastly, the results found in this study suggest that more research is warranted with Spanish children from different areas of Spain and using different measures of childhood anxiety and mood problems to determine the relationships between the CER strategies and these emotional problems, as well as to confirm the psychometric properties of the adapted scale.