Validating the Negative Emotion Regulation Inventory Measure of Emotion Regulation for Children in Preadolescence

Method

Results

Emotions Experienced Across Emotion Situation

Sadness

The one-way repeated measures ANOVA showed that sadness was significantly affected by the type of emotion situation, F (2, 182.49) = 123.83, p < .001. Post hoc tests using the Bonferroni correction revealed that, as expected, children reported more sadness in sad situations than in fear situations or in anger situations (see Table 1). They also reported more sadness in anger situations than in fear situations.

OPEN IN VIEWER

Table 1.

Means (and Standard Deviations) for Emotion and ER Strategies.

	Sad Situation	Fear Situation	Anger Situation	Across All Situations
Reported sadness	5.36 (1.24)	3.13 (1.55)	4.05 (1.37)	4.18 (1.12)
Reported fear	2.73 (1.39)	5.96 (1.20)	3.05 (1.45)	3.91 (1.03)
Reported anger	3.49 (1.43)	3.25 (1.52)	4.70 (1.41)	3.82 (1.13)
Adaptive ER	4.51 (1.12)	4.71 (1.35)	4.62 (1.31)	4.61 (1.17)
Social support seeking	4.68 (1.30)	4.97 (1.34)	4.22 (1.40)	4.62 (1.22)
Expressive suppression	3.86 (1.65)	3.78 (1.59)	3.53 (1.53)	3.72 (1.47)
Dysfunctional rumination	2.92 (1.03)	3.43 (1.25)	2.94 (1.27)	3.09 (1.11)

Fear

As expected, fear was significantly affected by the type of emotion situation, F (2, 198) = 278.66, p < .001. Post hoc tests revealed that participants reported more fear in fear situations than in sad situations or in anger situations (see Table 1). Fear was not significantly different in sad situations compared to anger situations.

Anger

As expected, anger was significantly affected by the type of emotion situation, F (2, 168.99) = 47.63, p < .001. Post hoc tests revealed that participants reported significantly more anger in anger situations than in sad situations or in fear situations (see Table 1). Anger was not significantly different in sad situations compared to fear situations.

Associations Between Experienced Emotions and ER Strategy Use

As shown in Table 2, when children experience greater sadness in sad situations, they engage in more social support seeking and dysfunctional rumination. We also found that when children have stronger feelings of fear in fear situations, they engage in more social support seeking. Further, results showed that when children experience greater anger in anger situations, they engage in less adaptive regulation in anger situations.

OPEN IN VIEWER

Table 2.

Bivariate Correlations Between Reported Emotion (Intensity) and ER Strategies.

	Adaptive Regulation	Social Support Seeking	Expressive Suppression	Dysfunctional Rumination
Sad situation reported sadness	−.049	.279**	.112	.213*
Fear situation reported fear	−.142	.321**	.030	.190
Anger situation reported anger	−.238*	-.024	.084	.110

Note. *p < .05. **p < .01. *** p < .001.

ER Strategies: Associations With Anxiety and Parent-Child Attachment

As shown in Table 3, children experiencing greater anxiety reported greater use of dysfunctional rumination and expressive suppression. Anxiety symptoms were not significantly associated with adaptive ER or social support seeking.

OPEN IN VIEWER

Table 3.

Pearson’s Correlations of Anxiety, Parent-Child Attachment, and ER Strategies.

	Adaptive Regulation Across All Emotion Situations	Social Support Seeking Across All Emotion Situations	Expressive Suppression Across All Emotion Situations	Dysfunctional Rumination Across All Emotion Situations
Anxiety symptoms	−.18	.02	.43***	.48***
Mother secure attachment	.14	.40***	−.47***	−.23*
Father secure attachment	.18	.29**	−.29**	−.28*

Note. *p < .05. **p < .01. *** p < .001.

Anxiety Symptoms N = 100, Attachment to Mother Figure N = 100, Attachment to Father Figure N = 98.

As predicted, children more securely attached to their mothers reported less use of dysfunctional rumination and expressive suppression and greater reliance on social support seeking (Table 3). Unexpectedly, there was not a significant association between mother-child attachment and adaptive ER.

The results were similar for father-child attachment, in that children more securely attached to their fathers reported less use of dysfunctional rumination and expressive suppression and greater use of social support seeking. Father-child attachment was not related to adaptive ER.

Because we obtained a similar pattern of results regarding relations between ER strategies and attachment to parental figures, we conducted partial correlations to examine unique effects of maternal and paternal attachment while controlling for security to the other parent. In two cases, we found that maternal attachment was uniquely related to ER strategies. Specifically, a significant negative correlation remained between mother-child attachment and expressive suppression when controlling for father-child attachment, r (95) = −.40, p < .001, and a positive correlation between mother-child attachment and social support seeking also remained significant when controlling for father-child attachment, r (95) = −.32, p = .001.

Finally, because the intensity of experienced emotions was related to ER strategy use, we tested whether the associations between ER and anxiety or attachment (findings in Table 3) changed after controlling for the intensity of experienced emotion. The results of these partial correlation analyses revealed the results did not change when controlling for emotion intensity.

Discussion

The purpose of our study was to test the reliability and validity of an adapted version of Zimmermann and Iwanski’s (2014) measure of ER strategy use in a sample of U.S. preadolescents. Overall, our results provided some evidence for the adapted measure’s reliability and validity. Specifically, scale reliabilities were, for the most part, acceptable, for both the total sample as well as the subsample of 9- to 11- year-olds. Also, the scenarios did elicit the intended emotions. These findings suggest that the situations may have relevance in different Western contexts, although we did modify some details (e.g., use of public transportation) to ensure the scenarios would reflect the common experiences of U.S. children. In addition, our hypothesis about ER strategy use varying depending on the emotion experienced was supported, and for the most part matched previous findings in the literature (Zimmermann & Iwanski, 2014). However, as discussed below, we did find that fear was a stronger elicitor of ER compared to Zimmermann and Iwanski’s findings. We also extended earlier work by showing that the specific ER strategies were associated in expected ways with anxiety and parent-child attachment security.

Our results showed that the use of specific ER strategies varied by emotion situation, supporting the notion that children have different goals based on the perceived emotions in a situation and achieve these goals through use of different ER methods (Thompson, 1994). Our study differed from Zimmermann and Iwanski’s (2014) study by finding differences in the use of adaptive regulation by emotion context; adaptive ER was used more often in fear situations than in sad situations. Additionally, while we replicated Zimmermann and Iwanski’s findings that social support seeking was used more often in sad situations than in anger situations, we also found that social support seeking was used even more often in fear situations than in sad situations. We replicated findings from Zimmermann and Iwanski’s study that children are less likely to engage in expressive suppression when angry, although we found no difference between suppression of emotion in fear and sad situations. We also replicated findings that dysfunctional rumination is used more often in fear situations than in sad situations, but we did not find that dysfunctional rumination was also used more often in anger situations than sad situations.

Fear was particularly distinguishing for ER strategy use in this younger sample; all four ER strategies were used most often in fear situations. Fear may be a more potent trigger for ER strategy use in our sample due to the inclusion of younger children in our study. In contrast, adolescents may want to present themselves as unafraid (e.g., not seek social support) and suppress feelings of fear (Zimmermann & Iwanski, 2018). Although preadolescents are more equipped to engage in ER in comparison to younger children (Saarni, 1999), they may nevertheless become overwhelmed in fearful contexts, and thus, feel a need to actively manage this emotion. Future studies could directly test our suggestion that fear may affect preadolescents’ use of specific ER strategies more than it does at older ages. More generally, the findings also point to the importance of considering context in future research when assessing ER strategies.

Contrary to our expectations, our results showed that ER strategy use did not vary significantly based on gender, although there was a trend for girls to report greater use of social support seeking than did boys (p < .06). It is important to note that in Zimmermann and Iwanski’s (2014) study, 1131 out of 1305 participants were over the age of 11. Gender differences may become stronger with age as individuals gain more exposure to societal norms and gender expectations. This could be explored in future studies that directly measure children’s adherence to gender norms and its relation to ER strategy use. Alternatively, effect sizes for gender may be small, and only reliably detected in larger samples.

We extended tests of the validity of the NERI by evaluating whether ER was associated with two developmental markers, anxiety symptoms and parent-child attachment, as both have been shown to be related to children’s use of ER strategies (see meta-analyses, Cooke et al., 2019; Mathews et al., 2016). Consistent with previous literature (Aldao, 2013; Mathews et al., 2016), we found that children with more anxiety symptoms made greater use of dysfunctional rumination and expressive suppression. These ER strategies could be considered “poor” regulation, as they either prolong the experience of the emotion or do not focus on resolving the issue at hand. Social support seeking and adaptive ER are less strongly related to anxiety symptoms (Mathews et al., 2016), and our sample size was modest, which may be why we did not find significant associations between these strategies and anxiety symptoms in our study. Successful interventions for children experiencing anxiety teach children to identify and recognize when they are feeling anxious, rather than suppress emotion, identify when they are having thoughts that prolong their anxiety (e.g., dysfunctional rumination), and address these thoughts with behavioral techniques encouraging the use of adaptive regulation or social support seeking (Suveg et al., 2009). The NERI could be used in intervention studies to test the extent to which children acquired adaptive ER skills.

As expected, we found that children who were more securely attached to their parental figures made greater use of social support seeking and less use of expressive suppression and dysfunctional rumination. Unexpectedly, more secure attachments to parental figures were not related to adaptive ER. Our findings are similar to those of Iwanski et al. (2021) who assessed attachment and ER in sad situations using the NERI with 9- to 17-year-olds. Although Iwanski et al. (2021) evaluated higher order dimensions of ER rather than the specific NERI ER scales, their findings showed that more secure attachment was associated with greater use of adaptive rather than maladaptive ER. Interestingly, when assessing unique associations of maternal and paternal attachment and ER, only maternal attachment was related to ER, suggesting that mothers might play an especially important role in socializing ER strategies. Overall, our findings are consistent with the notion that children with secure parent-child attachments are also able to use the parent as a secure base from which they can explore new emotional experiences, practice effective ER strategies, and openly express and discuss emotions (Cassidy, 1994; Cooke et al., 2019). Thus, they provide support for the idea that enhancing the quality of parent-child relationships in early adolescence promotes healthy adjustment, including adaptive ER skills such as social support seeking. Our study could be extended by examining how parent emotion socialization practices, particularly those of mothers, foster the development of ER strategies as assessed with the NERI.

Some limitations of our study should be noted. Our sample was primarily composed of White preadolescents, preventing generalization of findings to children of other ethnicities and races. The sample size was modest, and although we had power to detect medium and large effects, we did not have the power to detect small effects. Reliabilities were acceptable when looking at emotions across all situations but were lower when looking at emotions for specific emotion situations which could be due to a smaller number of responses for these items (six responses across all situations compared to two responses for specific situations). This could be addressed in a future study by adding another situation for each emotion. We focused on how children’s reports of their ER strategies were related to their self-perceptions of well-being and parent-child relationship quality. To ensure findings are not due solely to self-report bias, it will be important to examine individual differences in ER as assessed with the NERI in relation to outcomes reported by different observers using multiple methods (e.g., teacher report, behavior observation).

In conclusion, we showed that ER strategy use varied based on emotion, and we replicated previous findings regarding association of ER with anxiety and parent-child attachment. Fear contexts seemed especially likely to trigger ER efforts in this early adolescent sample. Further evaluation of the validity of the NERI in non-Western cultures is needed. Finally, preventive efforts aiming to enhance preadolescents’ healthy adjustment and skill building might need to target emotion regulation by emotion context, with particular attention to fear situations as they seem especially salient at this developmental stage.