Validation of the short version of the Cognitive Emotion Regulation Questionnaire for Spanish children

Abstract

This study aimed to validate a short version of the Cognitive Emotion Regulation Questionnaire for Spanish kids (CERQ-Sk) based on the 18-item version available for adults. A sample of 654 children aged 7–12 years completed the CERQ-Sk and tests for depression and anxiety measures. Confirmatory factor analysis supported the 18-item version and the original nine-factor structure, which includes self-blame, acceptance, rumination, positive refocusing, refocus on planning, positive reappraisal, putting into perspective, catastrophizing, and other-blame (comparative fit index = .99, Tucker–Lewis index = .98, root mean square error of approximation = .02). Internal consistency was adequate (ordinal α = .80), and the eight-week stability of this version was moderate (intraclass correlation = .69). Criterion validity was supported by correlations among self-blame, rumination, and catastrophizing (positive) and among positive reappraisal and depression and anxiety symptoms (negative). Results suggest that the short version of the CERQ-Sk is a valid and reliable instrument for assessing these cognitive emotion regulation strategies during the middle childhood developmental period. Clinicians and researchers will benefit from this briefer acceptable version when time is not available for the 36-item version. This study offers preliminary results for the first short version of the CERQ for children.

Keywords

Anxiety CERQ-k children cognitive emotion regulation

Introduction

Emotional regulation plays an important role in our adaptation to stressful situations and even in our well-being (Garnefski et al., 2001; Garnefski and Kraaij, 2006b). Emotional regulation involves the use of strategies to manage emotions by type, intensity, and duration to respond to environmental demands (Gross, 1999; Koole, 2009). The current study focuses on cognitive emotion strategies, since they have been found to be effective in regulating emotions in adverse situations (Garnefski and Kraaij, 2007). Theoretical models of psychopathology postulate that inadequate emotional management facilitates exposure to stress for long periods, and this exposure increases the risk of emotional problems (Mennin, 2005; Mennin et al., 2007).

Years of research have shown that certain cognitive strategies of emotional regulation (e.g. self-blame or rumination) are related to the development and/or maintenance of the practice of internalizing problems (e.g. depression and anxiety), while others (e.g. acceptance and positive reappraisal) are considered protective factors for emotional problems in general population, including children (see Aldao et al., 2010). The capacity for emotional regulation begins in childhood and progressively develops throughout life. Cognitive strategies used during childhood may have important implications for mental health during adulthood (Orgilés et al., 2018), so it is of special relevance to deepen their development from childhood. From the perspective of preventive and treatment approaches to psychopathology, it is essential to have valid and reliable tools to evaluate cognitive emotion regulation (CER) strategies.

Background

The Cognitive Emotion Regulation Questionnaire (CERQ; Garnefski et al., 2001) is one of the most widely used tools worldwide to assess cognitive emotional management strategies (e.g. Perte and Miclea, 2011; Tuna and Bozo, 2012; Zhu et al., 2008). The CERQ is especially valued because it focuses purely on cognitive strategies of emotional regulation, which is an advantage compared to other similar instruments, such as the Negative Mood Regulation (Catarizaro and Mearns, 1990) or the Emotion Regulation Questionnaire (Gross and John, 2003). The CERQ was developed according to the model postulated by Garnefski’s team that includes nine cognitive strategies of emotional regulation: self-blame, rumination, catastrophizing, other-blame, acceptance, positive focusing, refocus on planning, positive reappraisal, and putting into perspective (Garnefski et al., 2001; Garnefski and Kraaij, 2007). The first four strategies are considered maladaptive strategies because they have been shown to be directly related to depression and anxiety, while the rest are considered adaptive strategies because they may act to protect against these disorders (Domínguez-Sánchez et al., 2013; Garnefski et al., 2004, 2007; Liu et al., 2016; Orgilés et al., 2018). The original version of the CERQ consists of 36 items divided into the nine factors mentioned. Respondents must indicate how often they have used each of these strategies after having experienced stressful events.

The CERQ allows the evaluation of cognitive strategies for emotional regulation in different contexts (personal life and work and school environments) (Cakmak and Cevik, 2010). Therefore, this versatile instrument has been adapted to different languages and contexts for adults, adolescents, and children. A reliable and validated version of the CERQ for adults and adolescents is available in France (Jermann et al., 2006), China (Zhu et al., 2008), Romania (Perte and Miclea, 2011), Hungary (Miklósi et al., 2011), Iran (Abdi et al., 2012), Turkey (Tuna and Bozo, 2012), Spain (Domínguez-Sánchez et al., 2013), Argentina (Medrano et al., 2013), Germany (Görgen et al., 2015), Portugal (Martins et al., 2016), and Italy (Balzarotti et al., 2016). These cross-cultural adaptations of the CERQ generally maintained the structure of 36 items under nine factors, except for the Italian validation of the CERQ, which retained only 27 items (Balzarotti, et al., 2016), and showed adequate psychometric properties in terms of reliability and criterion validity according to the original version (Garnefski et al., 2001).

Based on the CERQ, Garnefski et al. (2007) developed a specific version for children (CERQ-k) that was initially tested on a sample of 717 primary school students aged 9–11 years in the Netherlands. Using principal component analysis, the authors confirmed the original nine-factor structure. Reliability across subscales was appropriate (α = .62, .79), and evidence of criterion validity was obtained with anxiety and depression measures. Liu et al. (2016) validated the Chinese version of the CERQ-k in a sample of 1403 elementary students aged 9–11 years. Confirmatory factor analysis (CFA) supported the original version. Internal consistency was also appropriate (α = .66, .73). Acceptable one-month test–retest coefficients were found in a subsample of 76 participants (r = .53, .70). More recently, Orgilés et al. (2017) validated the CERQ-k for Spanish-speaking children (CERQ-Sk) with a sample of 582 students aged 7–12 years in Spain. Scale reliability was excellent (α = .88), and moderate indexes were found for each subscale (α = .56–.75). The eight-week test–retest coefficient for a subsample of 221 children was adequate for the total scale (intraclass correlation, ICC = .74) and moderate for the subscales (ICC = .54, .70). Criterion validity was also provided. Children who self-reported the use of rumination, self-blame, and catastrophizing as strategies tended to present higher anxiety and depression symptomatology, while those who reported the use of positive refocusing and refocusing on planning showed a positive effect on the prevention of these internalizing disorders. These positive results obtained for the CERQ-Sk suggest the usefulness of the CERQ-k for children under nine years. In this regard, Garnefski et al. (2007) indicate, supported by prior research, that by the ages of eight and nine, children have learned to regulate their emotions by means of cognition or thoughts, but that knowledge of how cognitions regulate emotions in children remains scarce. Moreover, there are studies examining emotion regulation strategies at even earlier ages (e.g. 4–7 years; Silk et al., 2006). All of this would support the idea of continuing to investigate the use of emotion regulation strategies in younger children.

The original 36-item version of the CERQ has demonstrated good psychometric properties across countries and different populations. Nevertheless, this tool may be too long to be administered in certain community, clinical, and research settings, especially when time or resources are limited. The short versions of theory-based questionnaires are recommended for at least three reasons: (1) eliminating redundant items, (2) reducing the tedium of responding to similar items, and (3) generating less fatigue in respondents (Robins et al., 2001). Based on the 36-item CERQ, Garnefski and Kraaij (2006b) developed an 18-item short version in a study with 611 adults aged 18–65 years in the Netherlands. The original nine-factor model was kept, and satisfactory internal consistency (α = .67–.81) and criterion validity were proved. The 18-item version of the CERQ has been validated in Turkey in a sample of 317 undergraduate university students studying management, economics, and public finance (Cakmak and Cevik, 2010) and in Peru in a sample of 286 students of psychology from a private university located in Lima (Dominguez and Merino, 2015). So far, all 18-item versions of the CERQ have proven to be reliable and valid in general adult populations. Despite the advantages of using short self-reported measures, especially in child populations, there is no short validated version of the CERQ for children.

Objectives

The objective of the present study was to validate the first short version of the CERQ-k in a community sample of children aged 7–12 years in Spain. The factor structure of the short version of the CERQ-Sk was analyzed using CFA, based on short versions of the CERQ (Cakmak and Cevik, 2010; Dominguez and Merino, 2015; Garnefski and Kraaij, 2006b). We explored internal consistency, test–retest reliability, and criterion validity of the new short version of the CERQ-Sk to analyze the implications of CER strategies for children’s mental health. Based on previous CERQ studies, we hypothesized that the nine-factor structure of short versions of the CERQ will be confirmed with Spanish-speaking children (hypothesis 1). Evidence of reliability and temporal stability of the short version of the CERQ-Sk will also be provided (hypothesis 2). Criterion validity will be proved through the relationship between CER strategies and depression and anxiety measures. Adaptive CER strategies will be negatively related to depression and anxiety measures, while maladaptive CER strategies will be positively related to depression and anxiety measures (hypothesis 3).

Methods

Participants

Participants were 654 children aged 7–12 years. All were recruited from 11 public and private schools in Alicante Province, Spain.

Measures

Cognitive emotion regulation

The 36-item version of the CERQ-Sk (Orgilés et al., 2018) was administered. The CERQ-Sk consists of 36 items evaluating nine different subscales about what children think and what they actually do following the experience of threatening or stressful life events. These are as follows: (a) self-blame: putting the blame on oneself for what happened; (b) acceptance: resignation to what happened; (3) rumination: excessive worry after a negative event; (4) positive focusing: positive and joyous thoughts about the actual situation; (5) refocus on planning: thinking about how to solve a negative event; (6) positive reappraisal: thoughts focused on positive aspects in terms of personal growth; (7) putting into perspective: relativizing the importance of the event; (8) catastrophizing: thoughts about how horrible an event was; and (9) other-blame: putting the blame on someone else or the environment (Garnefski et al., 2001). Each subscale consists of four items, and response categories for items range from 1 (Almost never) to 5 (Almost always). The CERQ-Sk presents good psychometric properties (α = .88). The higher the score in subscales, the greater the use of that particular CER.

Depression

The Spanish version of the Child Depression Inventory (CDI; Del Barrio and Carrasco, 2004) was used. It consists of 27 items divided into two factors, dysphoria (17 items) and negative self-esteem (10 items), to assess depressive symptomatology over the preceding two weeks in children and adolescents aged 7–17 years. The response categories for items range from 0 (No symptoms) to 5 (Depressive symptoms). A total score is calculated; the higher the score, the more severe the depression symptomatology. Internal consistency of the CDI in the current sample was excellent (α = .82).

Anxiety

The trait anxiety subscale of the Spanish version of the State-Trait Anxiety Inventory for Children (STAIC; Pons-Salvador et al., 1994) was used, according to Orgilés et al. (2017). The response categories for items range from 0 (Not at all) to 3 (Very much). This version is used with children aged 9–15 years and even with younger children so long as they have above-average reading and comprehension ability. The higher scores indicate higher levels of anxiety. Internal consistency of the trait anxiety subscale in the current sample was excellent (α = .85).

Procedure

Internal review board approval for this study was provided by the Miguel Hernández University (DPS.MO.02.14). For the present study, the validated 36-item version of the CERQ-Sk was used (Orgilés et al., 2018). For this full version of the CERQ-Sk, a back-translation method was used based on the original version of the CERQ-k in English. Three bilingual clinical psychologists were involved in the process. Two of them translated the English version into Spanish; then, all 36 items composing the CERQ-Sk were translated back into English by the third. Possible discrepancies were discussed in-group to ensure the appropriateness and accuracy of the Spanish version of the CERQ. Recommendations of the International Test Commission for adapting psychological tests across cultures were followed (Muñiz et al., 2013). In order to test the short 18-item version of the CERQ-Sk, the items were selected based on the corresponding 18 items of the CERQ short version (Cakmak and Cevik, 2010; Dominguez and Merino, 2015; Garnefski and Kraaij, 2006b).

Principals of the participating schools provided their permission to evaluate the children during school hours. Only children whose parents provided signed written informed consent were evaluated. A member of the researcher group read the items aloud and responded to questions. Children answered the battery of questionnaires in groups of about ten. The data were treated confidentially, and participation was voluntary. No incentives were provided to the participants.

Statistical analysis

CFA using Lavaan, an R package for structural equation modeling, version 0.5–12 (BETA) (Rossel, 2012), was used to test the latent structure and internal consistency of the short version of the CERQ-Sk. The overall goodness of fit of the model was tested using a χ² test with comparative fit index (CFI), the Tucker–Lewis index (TLI), and the root mean square error of approximation (RMSEA). For good model fit, strict criteria were set as follows: CFI and TLI ≥ .95 and RMSEA ≤ .06 (Hu and Bentler, 1999). Diagonally weighted least squares method was used as the estimator method since it is recommended for use with ordinal data and in cases of absence of normality (Forero et al., 2009; Li, 2016). Ordinal α was calculated (Gadermann et al., 2012; Zumbo et al., 2007). Values closer to 1 indicate stronger reliability of the scale. The temporal stability of this new version was explored using ICC with data collected in the baseline and posttest (after eight weeks) measurements in a subsample. An ICC coefficient ≥ .60 was considered as showing acceptable reliability of the scale over time (Anastasi, 1998; Baumgartner and Chung, 2001).

Descriptive analyses were carried out using SPSS version 24 (with .05 level of significance). Sociodemographic variables of the sample as well as anxiety and depression measures were explored using descriptive statistics. Based on previous studies of the CERQ-k and CERQ short version (Garnefski and Kraaij, 2006b; Liu et al., 2016; Orgilés et al., 2018), criterion validity of the short version of the CERQ-Sk was examined across the total score and subscales, the trait anxiety subscale of the STAIC, and the total score on the CDI (Spearman–Brown correlations, ρ) in a subsample. Attrition analyses were conducted to test the equivalence between the subsamples used to test temporal stability of this tool and criterion validity and the rest of participants. The relationship across CER measures and anxiety and depression symptoms was explored using multiple regression analysis following the original study of the CERQ-k (Garnefski et al., 2007). The CERQ subscales were included as independent variables in the regression model and depression/anxiety as dependent variables. Predictor variables were tested for collinearity using variance inflation factor (VIF): values under three are considered ideal, although values up to ten are commonly accepted (Hair et al., 1995). In the current study, VIF values ranged from 1.14 to 1.35 (< 3), which indicate that there was no evidence of multicollinearity among predictor variables. Therefore, all were retained in the regression models with a p = .05 cutoff point.

Ethics approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the Institutional and/or National Research Committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individual participants included in the study.

Results

Study participants

A total of 654 children (313 girls; 47.9%), aged 7–12 years (M = 9.49 years; standard deviation, SD, = 1.2), participated in this study. The age distribution was as follows: 2.6% (n = 17) were 7 years old; 19.6% (n = 128) were 8 years; 31.8% (n = 208) were 9 years; 23.1% (n =151) were 10 years; 18.3% (n = 120) were 11 years; and 4.6% (n =30) were 12 years old. The mean number of participants’ siblings was 1.16 (SD = 0.83). All were Spanish-speaking children; most of them (98.6%) were born in Spain.

Confirmatory factor analysis

The nine-factor model of the short version of the CERQ-Sk fit the data with Spanish-speaking children very well: CFI = .99, TLI = .98, RMSEA = .02 (.019, .037). These coefficients showed a suitable fit for the 18-item model. Two items were included in each factor: self-blame, acceptance, rumination, positive refocusing, refocus on planning, positive reappraisal, putting into perspective, catastrophizing, and other-blame. All the items showed factor loadings greater than .40 except for item 4, ‘It just happened; there is nothing I can do about it’ (factor loading of .35; Table 1). All covariances between the CER strategies of the CERQ had a moderate or large effect size (Cohen, 1988).

Table 1.

Confirmatory factor analysis: factor loadings (N = 654).

F1. Self-blame
10. I think that I have been stupid	.64
28. I think that it’s all caused by me	.62
F2. Acceptance
2. I think that I have to accept it	.54
11. It just happened; there is nothing I can do about it	.35
F3. Rumination
3. Again and again, I think of how I feel about it	.63
12. I often think of what I am thinking and feeling about it	.49
F4. Positive refocusing
13. I think of nicer things that have nothing to do with it	.45
22. I think of something nice and not about what happened	.45
F5. Refocus on planning
23. I think of how I can change it	.50
32. I think of what I can do best	.57
F6. Positive reappraisal
6. I think that I can learn from it	.63
15. I think that it makes me feel ‘older and wiser’	.74
F7. Putting into perspective
25. I think that it’s not as bad as other things that could happen	.53
34. I think that there are worse things in the world	.58
F8. Catastrophizing
17. Again and again, I think about how terrible it all is	.44
35. I often think about how horrible the situation was	.56
F9. Other-blame
18. I think that others have been stupid	.42
36. I think that it’s all caused by others	.65

Note: The number of each item refers to the 36-item Spanish version of the CERQ-k (CERQ-Sk) administered for this study. CERQ-Sk: Cognitive Emotion Regulation Questionnaire for Spanish kids.

Table 2 presents the descriptive statistics of the items of the short version of the CERQ-Sk, including means, SDs, ordinal α values, and the corrected item-total correlations (r _it ^c). Values above .30 for all items—except for the item ‘I think that it’s all caused by others’—were obtained for the corrected item-total correlation. This indicates an adequate performance of all items. The item ‘I think that it’s all caused by others’, having a figure below 0.30, was retained, as the removal of any item did not significantly increase internal consistency; therefore, the short version included the original 18 items.

Table 2.

Scale properties of the CERQ short version.

	Item	M	SD	r _it ^c	α-i
F1	10. I think that I have been stupid	3.25	1.40	.47	.78
	28. I think that it’s all caused by me	2.83	1.45	.53	.78
F2	2. I think that I have to accept it	3.28	1.38	.49	.78
	11.It just happened; there is nothing I can do about it	2.26	1.29	.49	.78
F3	3. Again and again, I think of how I feel about it	3.31	1.38	.52	.78
	12. I often think of what I am thinking and feeling about it	2.95	1.42	.58	.78
F4	13. I think of nicer things that have nothing to do with it	2.89	1.50	.46	.79
	22. I think of something nice and not about what happened	2.82	1.51	.44	.80
F5	23. I think of how I can change it	2.65	1.43	.48	.78
	32. I think of what I can do best	2.15	1.28	.32	.80
F6	6. I think that I can learn from it	2.88	1.47	.46	.78
	15. I think that it makes me feel ‘older and wiser’	3.34	1.40	.58	.78
F7	25. I think that it’s not as bad as other things that could happen	3.01	1.41	.52	.78
	34. I think that there are worse things in the world	2.28	1.32	.49	.78
F8	17. Again and again, I think about how terrible it all is	3.36	1.42	.55	.78
	35. I often think about how horrible the situation was	3.20	1.50	.51	.78
F9	18. I think that others have been stupid	2.51	1.43	.42	.78
	36. I think that it’s all caused by others	1.76	1.12	.16	.78
	Total	2.91	1.45	—	—
		M	SD	α
F1		4.54	2.12	.54
F2		6.55	2.40	.70
F3		5.77	2.39	.60
F4		5.76	2.55	.70
F5		6.70	2.36	.62
F6		6.10	2.31	.47
F7		6.21	2.43	.61
F8		5.16	2.41	.66
F9		3.90	1.97	.61
Total		50.73	11.46	.80

Note: CERQ: Cognitive Emotion Regulation Questionnaire; M: mean; SD: standard deviation; r _it ^c: corrected item-total correlation; α-i: ordinal alpha if the item is removed; α: ordinal alpha; F1 = self-blame; F2 = acceptance; F3 = rumination; F4 = positive refocusing; F5 = refocus on planning; F6 = positive reappraisal; F7 = putting into perspective; F8 = catastrophizing; F9 = other-blame.

Internal consistency and test–retest reliability

The nine subscales showed adequate reliability, with ordinal α values ranging from .60 (rumination) to .70 (acceptance and positive refocusing), except for self-blame (ordinal α = .54) and positive reappraisal (ordinal α = .47). Internal consistency of the total questionnaire was excellent (ordinal α = .80). Test–retest reliability was tested in a subsample of 221 children (33.79% of retention) from the total sample eight weeks after baseline evaluation. Attrition analyses confirmed that this subsample was equivalent to the rest of participants in sociodemographic variables and the main outcomes, including scores on CERQ, CDI, and STAIC (p > .05). ICC test–retest coefficients for each scale ranged from .46 (other-blame) to .66 (catastrophizing), indicating acceptable test–retest reliability. The test–retest coefficient for total scale score was adequate (ICC = .69).

Correlations across subscales

Spearman correlation coefficients (ρ) across the short version of the CERQ-Sk subscales ranged from −.18 (acceptance and other-blame) to .38 (acceptance and refocus on planning; Table 3). Most of these correlations were significant at p < .01. Based on Cohen (1988), these correlations are considered small and moderate. When the correlation was positive, this suggested that the children who use a specific CER strategy were more likely to use other specific strategies (it does not imply causality). For example, children who self-reported using planning as a strategy were more likely to use acceptance (ρ = .38) and positive refocusing (ρ = .24). However, when the correlation was negative, this indicated that the children who use a specific CER strategy were less likely to use other specific strategies (again, it does not imply causality). For example, children who self-reported that they use of other-blame as a strategy were less likely to use acceptance (ρ = −.18) or positive refocusing (ρ = −.11).

Table 3.

Spearman correlations (ρ) between CERQ-short scales, depression (CDI) and anxiety (trait anxiety of STAIC) measures.

CERQ subscales	1	2	3	4	5	6	7	8	Depression	Trait anxiety
1. Self-blame	—								.14*	.18**
2. Acceptance	.25**	—							−.08	−.008
3. Rumination	.20**	.15**	—						.26**	.27**
4. Positive refocusing	.15**	.17**	.07	—					−.12*	−.004
5. Planning	.22**	.38**	.24**	.24**	—				−.11*	−.004
6. Positive reappraisal	.18**	.33**	.19**	.25**	.27**	—			−.13*	.05
7. Putting into perspective	.11**	.21**	.22**	.22**	.28**	.25**	—		.001	.08
8. Catastrophizing	.16**	−.007	.37**	−.10	.08	−.05	.07	—	.28**	.25**
9. Other-blame	−.01	−.18**	.04	−.11*	−.02	−.12*	.02	.28**	.07	.04

Note: CDI: Child Depression Inventory; CERQ: Cognitive Emotion Regulation Questionnaire; F1 = self-blame, F2 = acceptance; F3 = rumination; F4 = positive refocusing; F5 = refocus on planning; F6 = positive reappraisal; F7 = putting into perspective; F8 = catastrophizing; F9 = other-blame; STAIC: State-Trait Anxiety Inventory for Children.

* Correlation is significant at the .05 level (two-tailed).

** Correlation is significant at the .01 level (two-tailed).

Criterion validity

In order to study the implications of CER strategies for children’s mental health, criterion validity was tested in a subsample of 318 children (48.62% of retention). This subsample was equivalent to the rest of participants in sociodemographic variables and main outcomes, including scores on CERQ, CDI, and STAIC (p > .05). We calculated the Spearman bivariate correlation index (ρ) between each of the short version of the CERQ-Sk subscales and the scores on depression (total score on the CDI) and anxiety (trait anxiety subscale of STAIC). The results are presented in Table 3. Two multiple regression analyses were performed with measures of depression (CDI), anxiety (trait anxiety subscale of STAIC), and the nine subscales of the short version of the CERQ-Sk. Both regression models were significant (p < .001). Percentages of explained variance were 16 and 14 for the ‘prediction’ of depression and trait anxiety, respectively. In the ‘prediction’ of depressive symptoms, self-blame, rumination, positive reappraisal, and catastrophizing reached the level of significance, and refocus on planning was marginally significant (p = .07). Self-blame, rumination, and catastrophizing (maladaptive CER strategies) were positively related to reporting depressive symptomatology. However, positive reappraisal and refocus on planning (adaptive CER strategies) were negatively related to depression. The most important predictors of anxiety were self-blame, rumination, and catastrophizing. Refocus on planning reached a marginally significant level (p = .07) in predicting anxiety. Self-blame, rumination, and catastrophizing (maladaptive CER strategies) were positively related to reporting anxiety symptomatology. However, refocus on planning (adaptive CER strategy) was negatively related to anxiety.

Discussion

This study was aimed at developing and validating the first short version of the CERQ-Sk in a community sample of children aged 7–12 years in Spain. The results provided support for hypothesis 1, since the theoretical nine-factor structure of the short CERQ-k for Spanish-speaking children was confirmed. This is consistent with previous short versions of the CERQ for adults (Cakmak and Cevik, 2010; Domínguez-Sánchez et al., 2013; Garnefski and Kraaij, 2006b). The nine-factor structure has also been supported by previous validations of the 36-item CERQ-k (Garnefski et al., 2007; Liu et al., 2016; Orgilés et al., 2018). This indicates that the latent structure of this tool seems to be similar for its long and short versions, as well as across countries (the Netherlands, China, Spain, and many others) and samples (adults and children). In the current study, factor loadings were medium and high, ranging from .35 to .74. These values are considered adequate and similar to those of the Chinese version (Liu et al., 2016); however, they are slightly lower than that on the original short version of the CERQ (Garnefski et al., 2007). Because the removal of any item did not increase internal consistency significantly, the original 18 items of the short version of the CERQ-Sk were retained in the final version.

The short version of the CERQ-Sk emerges as a brief and reliable tool to evaluate cognitive strategies of emotional regulation in Spanish-speaking children. Internal consistency and temporal stability results supported hypothesis 2. Total reliability was excellent (α = .80), although slighter lower for the 36-item version of the CERQ-Sk (α = .88) (Orgilés et al., 2018), which is less than desirable as it indicates that the short version of the CERQ-Sk is slightly less reliable than the full-version. Reliability depends on the number of items per factor, and as the short version of the CERQ-k reduces the number of items on each dimension from four to two, this logically and statistically reduces the reliability indices when compared with the original 36-item version (Garnefski and Kraaij, 2006b). Reliability coefficients for the subscales were acceptable, although two of them—self-blame (α = .54) and positive reappraisal (α = .47)—proved less reliable than is desirable. However, earlier studies found similar results (Dominguez and Merino, 2015; Garnefski and Kraaij, 2006b). For example, reliability coefficients lower than .60 were found for self-blame and acceptance for the Peruvian short-CERQ (Dominguez and Merino, 2015). Self-blame consistently had the lowest reliability index (.67) in the original short version of the CERQ for the general adult population (Garnefski and Kraaij, 2006b). Such low values may call into question the fidelity of evaluation for these two factors when taken separately.

The eight-week stability of the short version of the CERQ-Sk was considered good (Anastasi, 1998), which suggests that similar results will be obtained when the tool is administrated to the same sample. This value is higher than the five-month stability indexes provided in an earlier study of the CERQ among adolescents aged 12–16 years (Garnefski et al., 2001) and is similar to the one-month stability indexes reported in a validation study with Chinese children (Liu et al., 2016), but slightly lower than that for the 36-item CERQ-Sk for a Spanish-speaking population over a period of one and a half months (Orgilés et al., 2018). Thus, these results suggest that the short version of the CERQ-Sk is as stable over time as the full versions of the CERQ and CERQ-k tested with children from other countries, despite the fact that the full version of the CERQ-Sk (Orgilés et al., 2018) demonstrated a slightly higher test–retest coefficient (ICC = .74) over a shorter assessment period. Future studies should analyze whether there are significant differences in test–retest reliability between the long and short versions of the CERQ-k, as currently there are still few studies available to determine this accurately. On the other hand, changes in responses over time may be due to other factors than the validity of the test. For example, the use of certain cognitive emotion strategies might be influenced by newly experienced stressors and consequently may vary over time. Similarly, it is possible that children’s responses could change over time because the respondents actually change how they feel about each item. All of this may explain differences in temporal stability across time and studies. The ICC coefficients for the subscales indicate that the CERQ-Sk’s stability over time was acceptable. Self-blame (ICC = .49) and other-blame (ICC = .46) subscales had the lowest temporal stability. These were strategies less used by the children of this sample, which suggests that if self-blame and other-blame are barely used, they may be less likely to be consistent over time. Similar low test–retest coefficients for certain subscales of the CERQ were reported earlier (Liu et al., 2016; Miklósi et al., 2011; Orgilés et al., 2018).

The most used cognitive strategies of emotional regulation for children were refocus on planning and acceptance (adaptive strategies). Of the maladaptive CER strategies, rumination and catastrophizing were self-reported as the most used, while blame and other-blame were clearly the least used. Children who reported the use of self-blame, rumination, and/or catastrophizing as emotion regulation strategies were more likely to present higher depression scores, whereas a higher use of positive reappraisal was related to lower depression scores. These relationships across CER strategies and depression and anxiety measures provide evidence of criterion validity of the short version of the CERQ-Sk. These results also fit in with the findings of other studies using the CERQ in children (Liu et al., 2016; Orgilés et al., 2018), adolescents (Garnefski et al., 2001; Garnefski et al., 2005), and adults (Garnefski et al., 2004; Garnefski and Kraaij, 2006b).

In the current study, the use of other-blame—referring to thoughts of putting the blame on someone else or the environment—was unrelated to anxiety and depression symptomatology, as has been found by other studies conducted in Western countries, such as the Netherlands (Garnefski and Kraaij, 2006a) and Spain (Domínguez-Sánchez et al., 2013; Orgilés et al., 2018), with individualist models. Nevertheless, the use of other-blame has been related to depression in countries, such as China (Liu et al., 2016), with a collectivist model. Although other-blame is a maladaptive subscale, it does not seem to have the same implications as the rest of the maladaptive strategies across countries. This cross-cultural difference highlights the importance of studying factors associated with child emotional problems while taking into account their cultural framework.

Limitations

This study has several limitations. Because this is a cross-sectional study, conclusions about any direction of influence cannot be drawn. The results are based on children’s self-reported information on cognitive emotion strategies, depression, and anxiety. Because children may be susceptible to social desirability, the influence of social desirability on the child’s willing to endorse anxiety symptoms should be taken into account (see Arabiat et al., 2013). Future studies should collect information not only from children but from their parents as well, so that the information can be contrasted. Although the sample is not small, the study sample cannot be considered representative of the general child population in Spain. This is a general limitation found most validation scale studies, as mentioned by Loh et al., (2015). Additionally, future studies should examine convergent validity by examining relationships between the short version of the CERQ-Sk and other related measures evaluating emotional or coping strategies. More evidence is needed of the psychometric properties of the short version of the CERQ-Sk in the clinical child population. Since this is the first short version of the CERQ-k, the findings of the current study were discussed and compared to earlier research on short versions of the CERQ for adults and the 36-item CERQ-k versions across countries.

Conclusion

The present study provides initial support for the reliability and criterion validity of the short version of the CERQ-Sk, a multidimensional tool that evaluates a wide range of cognitive strategies of emotion regulation at the same time. This version coincides with the nine-factor structure of previous studies. To our knowledge, this is the first paper that provides support for a short version of the CERQ for children. Despite the good psychometric properties of the short version of the CERQ-Sk, the original 36-item Spanish version for children (Orgilés et al., 2018) proved to be a slightly more reliable and valid instrument to assess CER strategies. Therefore, the Spanish 36-item version of the CERQ-k should definitely be used when it is feasible to do so. This has been consistently found across the different short versions of the CERQ when comparing them with their original 36-item versions (Cakmak and Cevik, 2010; Dominguez and Merino, 2015; Garnefski and Kraaij, 2006b).

Findings showed that the use of CER strategies is related to depression and anxiety symptomatology in children. A considerable percentage of the variance of self-reported depression and anxiety was explained by the use of strategies of self-blame, rumination, and catastrophizing and the lesser use of positive reappraisal. Information collected using the short version of the CERQ-k may be especially valuable for the development and assessment of treatments that address anxiety and depression symptomatology, as well as the study of risk factors for psychological issues in the child population. Therefore, health practitioners and researchers who work on children’s mental health may benefit from this tool to quickly evaluate a wide range of CER strategies related to emotional problems. The next steps should address the gender and age invariance of the short version of the CERQ-Sk, as well as the validation of the short version of the CERQ Spanish version for adolescents aged under 16 years, which is not yet available for the Spanish-speaking population. Finally, although the results suggest that the short version of the CERQ-Sk is a reliable and valid instrument to assess CER strategies in children, it should be noted that these results need further support and should be interpreted with caution, as each strategy was evaluated by means of only two items, few factor loadings were above .70, some subscales showed low reliability coefficients (i.e. self-blame and positive reappraisal) or low temporal stability (i.e. self-blame and other-blame), and correlations with depression and anxiety were significant, but relatively small.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the Ministry of Economy and Competitiveness of Spain (PSI2014-56446-P).

References

Abdi

Taban

Ghaemian

(2012) Cognitive emotion regulation questionnaire: validity and reliability of the Persian translation of the CERQ (36-item). Procedia – Social and Behavioral Sciences 32: 2–7. DOI: 10.1016/j.sbspro.2012.01.001.

Ainuddin

Loh

Chinna

. (2015) Psychometric properties of the self-report Malay version of the Pediatric Quality of Life (PedsQLTM) 4.0 Generic Core Scales among multiethnic Malaysian adolescents. Journal of Child Health Care 19(2): 229–238.

Aldao

Nolen-Hoeksema

Schweizer

(2010) Emotion-regulation strategies across psychopathology: a meta-analytic review. Clinical Psychology Review 30(2): 217–237. DOI: 10.1016/j.cpr.2009.11.004.

Anastasi

(1998) Psychological Testing, 6th ed. New York: Macmillan.

Arabiat

Jabery

Wardam

(2013) Screening for anxiety symptoms and social desirability in children and adolescents living with chronic illnesses in Jordan. Journal of Child Health Care 17(1): 62–71. DOI: 10.1177/1367493512450623.

Balzarotti

Biassoni

Villani

. (2016) Individual differences in cognitive emotion regulation: implications for subjective and psychological well-being. Journal of Happiness Studies 17: 125–143. DOI: 10.1007/s10902-014-9587-3.

Baumgartner

Chung

(2001) Confidence limits for intraclass reliability coefficients. Measurement in Physical Education and Exercise Science 5(3): 179–188. DOI: 10.1207/S15327841MPEE0503_4.

Cakmak

Cevik

(2010) Cognitive emotion regulation questionnaire: development of Turkish version of 18-item short form. African Journal of Business Management 4(10): 2097–2102.

Catarizaro

Mearns

(1990) Measuring generalized expectancies for negative mood regulation: initial scale development and implications. Journal of Personality Assessment 54(3–4): 546–563. DOI: 10.1080/00223891.1990.9674019.

10.

Cohen

(1988) Statistical Power Analysis for the Behavioral Sciences. New York: Routledge.

11.

Del Barrio

Carrasco

(2004) CDI Inventario de Depresión Infantil. Madrid: TEA Ediciones.

12.

Dominguez

Merino

(2015) Una versión breve del cognitive emotional regulation questionnarie: análisis estructural del CERQ-18 en estudiantes universitarios limeños. [A brief version of the cognitive emotional regulation questionnaire: structural analysis of the CERQ-18 in university students from Lima]. Revista Peruana de Psicología y Trabajo Social 4(1): 25–36.

13.

Domínguez-Sánchez

Lasa-Aristu

Amor

. (2013) Psychometric properties of the Spanish version of the cognitive emotion regulation questionnaire. Assessment 20(2): 253–261. DOI: 10.1177/1073191110397274.

14.

Forero

Maydeu-Olivares

Gallardo-Pujol

(2009) Factor analysis with ordinal indicators: a Monte Carlo study comparing DWLS and ULS estimation. Structural Equation Modeling 16(4): 625–641. DOI: 10.1080/10705510903203573.

15.

Gadermann

Guhn

Zumbo

(2012) Estimating ordinal reliability for Likert-type and ordinal item response data: a conceptual, empirical, and practical guide. Practical Assessment, Research & Evaluation 17(3): 1–13.

16.

Garnefski

Kraaij

(2006a) Relationships between cognitive emotion regulation strategies and depressive symptoms: a comparative study of five specific samples. Personality and Individual Differences 40(8): 1659–1669. DOI: 10.1016/j.paid.2005.12.009.

17.

Garnefski

Kraaij

(2006b) Cognitive emotion regulation questionnaire - development of a short 18-item version (CERQ-short). Personality and Individual Differences 41(6): 1045–1053. DOI: 10.1016/j.paid.2006.04.010.

18.

Garnefski

Kraaij

(2007) The cognitive emotion regulation questionnaire: psychometric features and prospective relationships with depression and anxiety in adults. European Journal of Psychological Assessment 23(3): 141–149. DOI: 10.1027/1015-5759.23.3.141.

19.

Garnefski

Kraaij

Spinhoven

(2001) Negative life events, cognitive emotion regulation and emotional problems. Personality and Individual Differences 30(8): 1311–1327. DOI: 10.1016/S0191-8869(00)00113-6.

20.

Garnefski

Kraaij

van Etten

(2005) Specificity of relations between adolescents’ cognitive emotion regulation strategies and internalizing and externalizing psychopathology. Journal of Adolescence 28(5): 619–631. DOI: 10.1016/j.adolescence.2004.12.009.

21.

Garnefski

Rieffe

Jellesma

. (2007). Cognitive emotion regulation strategies and emotional problems in 9-11-year-old children: the development of an instrument. European Child and Adolescent Psychiatry 16(1): 1–9. DOI: 10.1007/s00787-006-0562-3.

22.

Garnefski

Teerds

Kraaij

. (2004) Cognitive emotion regulation strategies and depressive symptoms: differences between males and females. Personality and Individual Differences 36(2): 267–276. DOI: 10.1016/S0191-8869(03)00083-7.

23.

Görgen

Loch

Hiller

. (2015) Kognitive emotionsregulation und psychopathologie: erste teststatistische untersuchung des deutschen cognitive emotion regulation questionnaire (CERQ) in einer klinischen stichprobe. [Cognitive emotion regulation and psychopathology: psychometric evaluation of the German cognitive emotion regulation questionnaire (CERQ) in a clinical sample]. Zeitschrift Fur Psychiatrie, Psychologie und Psychotherapie 63(4): 255–265. DOI: 10.1024/1661-4747/a000248.

24.

Gross

(1999) Emotion regulation: past, present, future. Cognition & Emotion 13(5): 551–573. DOI: 10.1080/026999399379186.

25.

Gross

John

(2003) Individual differences in two emotion regulation processes: implications for affect, relationships, and well-being. Journal of Personality and Social Psychology 85(2): 348–362. DOI: 10.1037/0022-3514.85.2.348.

26.

Hair

JFJ

Anderson

Tatham

. (1995) Multivariate Data Analysis, 3rd ed. New York: Macmillan.

27.

Bentler

(1999) Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal 6(1): 1–55. DOI: 10.1080/10705519909540118.

28.

Jermann

Van Der Linden

D’Acremont

. (2006) Cognitive emotion regulation questionnaire (CERQ): confirmatory factor analysis and psychometric properties of the French translation. European Journal of Psychological Assessment 22(2): 126–131. DOI: 10.1027/1015-5759.22.2.126.

29.

Koole

(2009) The psychology of emotion regulation: an integrative review. Cognition and Emotion 23(1): 4–41. DOI: 10.1080/02699930802619031.

30.

(2016) Confirmatory factor analysis with ordinal data: comparing robust maximum likelihood and diagonally weighted least squares. Behavior Research Methods 48(3): 936–949. DOI: 10.3758/s13428-015-0619-7.

31.

Liu

Chen

Blue

(2016) Chinese adaptation and psychometric properties of the child version of the cognitive emotion regulation questionnaire. PLoS One 11(2): e0150206. DOI: 10.1371/journal.pone.0150206.

32.

Martins

Freire

Ferreira-Santos

(2016). Examination of adaptive and maladaptive cognitive emotion regulation strategies as transdiagnostic processes: associations with diverse psychological symptoms in college students. Studia Psychologica 58(1): 59–73.

33.

Medrano

Moretti

Ortiz

. (2013). Validación del cuestionario de regulación emocional cognitiva en universitarios de Córdoba, Argentina. [Validation of the cognitive emotion regulation questionnaire in university students of Córdoba, Argentina]. Psykhe 22(1): 83–96.

34.

Mennin

(2005) Emotion and the acceptance-based approaches to the anxiety disorders. In: Orsillo

Roemer

(eds), Acceptance and Mindfulness-Based Approaches to Anxiety: Conceptualization and Treatment. New York: Springer, pp. 37–68.

35.

Mennin

Holaway

Fresco

. (2007). Delineating components of emotion and its dysregulation in anxiety and mood psychopathology. Behavior Therapy 38(3): 284–302. DOI: 10.1016/j.beth.2006.09.001.

36.

Miklósi

Martos

Kocsis-Bogár

. (2011). Psychometric properties of the Hungarian version of the cognitive emotion regulation questionnaire. Psychiatria Hungarica: A Magyar Pszichiátriai Társaság Tudományos Folyóirata 26(2): 102–111.

37.

Muñiz

Elosua

Hambleton

(2013) Directrices para la traducción y adaptación de los tests: Segunda edición. Psicothema 25(2): 151–157. DOI: 10.7334/psicothema2013.24.

38.

Orgilés

Morales

Fernández-Martínez

. (2018) Spanish adaptation and psychometric properties of the child version of the cognitive emotion regulation questionnaire (CERQ-Sk). Manuscript under revision.

39.

Perte

Miclea

(2011) The standardization of the cognitive emotional regulation questionnaire (CERQ) on Romanian population. Cognitie, Creier, Comportament 15(1): 111–130.

40.

Pons-Salvador

Frías

Del Barrio

(1994) Propiedades psicométricas del state-trait anxiety inventory for children (STAIC). [Psychometric properties of state-trait anxiety inventory for children (STAIC)]. Psicológica 15: 45–53.

41.

Robins

Hendin

Trzesniewski

(2001) Measuring global self-esteem: construct validation of a single-item measure and the Rosenberg self-esteem scale. Personality and Social Psychology Bulletin 27(2): 151–161.

42.

Rossel

(2012) Lavaan: an R package for structural equation modeling and more. Version 0.5–12 (BETA). Available at: http://users.ugent.be/∼yrosseel/lavaan/lavaanIntroduction.pdf (accessed 15 October 2017).

43.

Silk

Shaw

Skuban

. (2006) Emotion regulation strategies in offspring of childhood-onset depressed mothers. Journal of Child Psychology and Psychiatry 47: 69–78. DOI: 10.1111/j.1469-7610.2005.01440.x.

44.

Tuna

Bozo

(2012) The cognitive emotion regulation questionnaire: factor structure and psychometric properties of the Turkish version. Journal of Psychopathology and Behavioral Assessment 34(4): 564–570. DOI: 10.1007/s10862-012-9303-8.

45.

Zhu

Auerbach

Yao

. (2008) Psychometric properties of the cognitive emotion regulation questionnaire: Chinese version. Cognition and Emotion 22(2): 288–307. DOI: 10.1080/02699930701369035.

46.

Zumbo

Gadermann

Zeisser

(2007) Ordinal versions of coefficients alpha and theta for Likert rating scales. Journal of Modern Applied Statistical Methods 6(1): 21–29. DOI: 10.22237/jmasm/1177992180.