Beginning around age three, children start actively engaging in reminiscing with their parents (Fivush and Salmon, 2023). Their developing autobiographical skills enable them to recall and share past emotional experiences, while their parents primarily guide the conversations and support narrative coherence (Fivush and Salmon, 2023). Parents provide structure by (1) elaborating on past experiences and (2) evaluating children's contribution to these conversations (Fivush et al., 2006; Haden, 1998; Wu and Jobson, 2019).

Highly elaborative parents provide novel and detailed information to the conversation. For example, when reminiscing about a visit to the park, parents could provide novel information by saying “There were also ducks at the park. We fed the ducks some peas!”. Highly elaborative parents also make use of informative open-ended questions, often called W-questions (e.g., questions with “Why”, “When”, “What”) that elicit memory information from the child (Fivush et al., 2006). An elaborative style is contrasted with a repetitive style which is characterized by asking close-ended and/or repeating questions and reiterating previously presented information.

In terms of structure, parents also differ in their provision of evaluations regarding a child's contributions (Haden, 1998; Wu and Jobson, 2019). Parents can either provide positive evaluations by affirming their child's input, or negative evaluations by denying or criticizing their child's input. When children elaborate, parents could respond to their child, for example, by saying “Yes, indeed!” or “No, that is not what happened”.

Emerging research has shown that highly structured conversations that are high in elaborations and positive evaluations, and low in negative evaluations about past experiences relate positively to children's socioemotional development (Fivush and Salmon, 2023; Salmon and Reese, 2016). Previous studies show that mothers' highly elaborative reminiscing relates to children's increased emotion knowledge (Van Bergen and Salmon, 2010; Valentino et al., 2019) and appropriate emotional responding in social situations (Leyva et al., 2014). Likewise, positive evaluations were related to overall socioemotional competence (Song and Wang, 2013) and negative evaluations were related to disaffection in children (Van der Kaap-Deeder et al., 2020).

In addition to the provision of structure, the degree to which parents provide autonomy support during parent-child reminiscing is important for what children learn from these conversations. Studies have shown that autonomy support is distinct from structure (Cleveland and Morris, 2014; Cleveland and Reese, 2005). Specifically, while elaboration and evaluation involve organizing the conversation around novel aspects and topics and stimulating children's memory, autonomy support pertains to how that structure is implemented and conveyed (Leyva et al., 2009). According to the Self-Determination Theory (SDT; Ryan and Deci, 2017), parents who reminisce with high autonomy support encourage their child's volitional functioning by focusing the conversation on what their child wants to discuss and by empathizing with their child's perspective and feelings (Cleveland and Reese, 2005; Cleveland and Morris, 2014). In contrast, when parents reminisce with high psychological control, they minimize their child's experience and dictate their own perspective (Cleveland and Morris, 2014; Cleveland and Reese, 2005). Similar to the studies on general autonomy support, parents' autonomy-supportive reminiscing style has been found to relate to more adaptive socio-cognitive outcomes in children including better memory skills (Cleveland and Reese, 2005; Cleveland and Morris, 2014; Van der Kaap-Deeder et al., 2020), emotion knowledge (McClaine et al., 2023), and social competence (Kulkofsky et al., 2015; Wu et al., 2022). Additionally, children tend to be more interested and engaged in reminiscing when parents are more autonomy-supportive (Cleveland and Morris, 2014; Van der Kaap-Deeder et al., 2020), which may, in turn, facilitate greater learning from these conversations. Yet, how different components of parents' reminiscing style relate over time to emotion regulation in children has not been systematically examined.

While there is a growing body of evidence showing the importance of high-quality parent-child reminiscing for children's socioemotional development (Fivush and Salmon, 2023), less is known about its relation to emotion regulation specifically. Emotion regulation can be defined as the processes that people use to modulate their emotions, thereby increasing, maintaining, or decreasing their emotional response (Gross, 2002). As adaptive emotion regulation relates to positive mental health and academic outcomes in children (Bardack and Widen, 2019; Schäfer et al., 2017), it is important to study its promoting factors. In infancy, parents help children regulate their emotions with children increasingly learning to self-regulate as they get older (Guo et al., 2021). In accordance with that, mothers report that one of the main reasons that they reminisce with their child is to foster their child's emotion regulation (Kulkofsky and Koh, 2009). Cross-sectional studies showed that mothers' highly structured reminiscing was related to emotion regulation in children (Hernandez et al., 2019; Leyva et al., 2014, 2020). A recent study found that when different components of structure—elaboration and evaluation—were examined separately, negative evaluation (but not elaboration or positive evaluation) when discussing negative events was related to worse emotion regulation in children (Van der Kaap-Deeder et al., 2020). In terms of longitudinal relations, structure was found to be unrelated to children's emotion regulation skills (Leyva et al., 2020).

While parent- and child-reported autonomy support has been shown to be related to the development of adaptive emotion regulation in children (Brenning et al., 2015), the only study that examined the relation between observed autonomy support when reminiscing and child emotion regulation found these two to be unrelated (Van der Kaap-Deeder et al., 2020). Therefore, this study aims to examine how different components of mothers' reminiscing style—elaboration, positive evaluation, negative evaluation, and autonomy support—relate to the development of emotion regulation in children.

As most studies focused on how differences in parents' reminiscing style predict child outcomes, less is known about the predictors of parents' reminiscing style. The few studies thus far suggest that besides parent (e.g., gender) and contextual (e.g., the valence of the memory) factors, child characteristics (e.g., temperament) may play an important role in explaining differences in parental reminiscing style (Bird et al., 2006; Bost et al., 2010; Laible, 2004; Laible et al., 2013; Reese et al., 2019).

Although no research thus far examined the potential path from children's emotion regulation to parents' reminiscing style across time, research in the broader parenting literature suggests that how children express and regulate their emotions predicts parents' parenting practices such as emotion socialization behaviors (Kiel and Kalomiris, 2015) and perceived parental autonomy support and psychological control (Brenning et al., 2015; Gao et al., 2021). For example, parents have been found to affirm emotional experiences and promote adaptive emotion regulation in the child when children displayed more adaptive emotion regulation strategies (Morelen and Suveg, 2012). Given this research on child emotion regulation skills and general parenting behaviors, children's ability to adaptively regulate their emotions may also relate to parents talking about past emotional experiences with high elaboration, more positive evaluation, less negative evaluation, and high autonomy support. Yet, no study has examined cross-lagged relations between parents' reminiscing style and children's emotion regulation, which is therefore the aim of the present study.

Research has extensively examined the effects of parent-child reminiscing on children's cognitive and socioemotional capabilities. Yet, less is known about how parents' reminiscing style relates to children's emotion regulation with existing studies mainly focusing on concurrent relations and yielding mixed results (e.g., Hernandez et al., 2019; Van der Kaap-Deeder et al., 2020). Given that previous studies indicated the importance of examining the unique relations of different facets of parents' reminiscing style (e.g., Salmon and Reese, 2016), we assessed mothers' reminiscing style by coding for elaboration, positive evaluation, negative evaluation, and autonomy support separately. We exploratively examined how these facets of mothers' reminiscing style relate to concurrent and later emotion regulation in children. When examining emotion regulation, we examined both adaptive regulation as well as emotional lability (i.e., heightened emotional reactivity and inflexibility).

The present study also adds to the literature by examining reminiscing of both positive and negative memories. When reminiscing about negative experiences in comparison to positive experiences, mothers talk more about causes and focus on resolving negative emotions (Fivush and Wang, 2005). This may help children to learn how to regulate their emotions in a more adaptive manner (Leyva et al., 2014). For example, Van der Kaap-Deeder et al. (2020) found that negative evaluation related to more emotional lability only when discussing negative memories. Therefore, if mothers' reminiscing style would relate to children's emotion regulation, we expected this relation to be present when discussing negative (compared to positive) memories.

Finally, we extend current knowledge by examining the potential cross-lagged relation between maternal reminiscing and children's emotion regulation. Toward these aims, we measured both observed reminiscing style and child emotion regulation once at the beginning of the school year (at 4 years of age) and once around 7 months later. We predicted that high emotion regulation and low emotional lability would relate to mothers' provision of high elaboration, positive evaluation, and autonomy support and low negative evaluation both concurrently and across time.

This project employed data from a larger study on the antecedents and outcomes of parents' reminiscing style. The data were obtained from Ghent University, Belgium. Families (T1, n = 89; T2, n = 63) from the Flanders region of Belgium participated in the study, spanning two home visits. At the start of the study, children (47.2% girls) were between the ages of 2.5 and 6.5¹ (M_age = 4.02 years, SD = 0.85) and mothers were between the ages of 25 and 44 (M_age = 34.67 years, SD = 3.92). Most children had at least one sibling (94.3%). Most mothers were either married (73%) or living with a partner (25.8%) and had completed at least post-secondary education (90.8%).

Undergraduate students recruited families in exchange for course credits for a psychology course at Ghent University. Each student invited two mothers (could not be their own) and their eldest child between the ages of 2.5 and 6.5. Data were collected at two home visits: once at the beginning of the school year and once toward the end. On average, there were 7.37 (SD = 1.30) months between the two visits. The first home visits were carried out by the undergraduate students who initially recruited the families, primarily through their personal network. The second visits were conducted by one of the researchers or one of three master students. Therefore, the personal connection to participating families was lost from T1 to T2, which may explain the relatively high attrition rate (29.21%). To maintain consistency across data collection, all students received a one-hour training session covering the study procedures and guidelines for collecting data during home visits.

During the first visit, mothers were informed about the study's procedure, and they provided written consent for their own and their child's participation. Each visit consisted of mothers completing questionnaires (on sociodemographics and their child's language skills (SNEL; Luinge, 2005) and emotion regulation) followed by two reminiscing tasks, which are described below in the Measures section. At the start, mothers were informed that there are no wrong answers, that they could leave any question unanswered if they were unsure, and that they could ask questions at any time. At the second visit, mothers again filled out the questionnaire on child emotion regulation and repeated the reminiscing tasks with their child. Ethical approval was obtained from the ethical committee at the Faculty of Psychology and Educational Sciences, Ghent University. Families did not receive any reward and their participation was voluntary.

Using independent samples t-tests, we examined whether mother-child dyads that participated at both timepoints (70.8%) differed from those that only participated at T1. There were no significant differences between these two groups in terms of child (i.e., gender, emotion regulation, emotional lability; t values ranging between −0.21 and 0.59; p > 0.05) or parent (i.e., marital status, education level, number of children: t values ranging between −1.47 and 0.36; p > 0.05) characteristics, with one exception. That is, mothers (M = 33.19, SD = 4.31) and children (M = 3.70, SD = 0.82) who participated only at T1 were younger than those who participated at both timepoints [mothers: M = 35.32, SD = 3.59; children: M = 4.14, SD = 0.84; mothers: t(40.76) = −2.20, p = 0.03; children: t(39.90) = −2.22, p = 0.03].

Preliminary analyses were conducted in SPSS Version 28. The cross-lagged relations between mothers' reminiscing style and children's emotion regulation were tested with cross-lagged panel modeling using the lavaan package (Rosseel, 2012) in R Studio (R Core Team, 2022). Given the sample size, two separate models for positive and negative memories were tested. For both models, we employed bootstrapped confidence intervals with 1,000 resamples. As Little's Missing Completely at Random (MCAR) test (Little, 1988) showed that the missing data were missing completely at random [X²(141) = 129.47, p = 0.75], we employed full information maximum likelihood procedure to handle missing data. In the models, we examined all relations between the observed mothers' reminiscing style facets (elaboration, positive evaluation, negative evaluation, and autonomy support) and mother-reported child emotion regulation and emotional lability at T1 and T2. As all paths were allowed, the models were fully saturated (df = 0). Thus, model fits were not assessed.

First, to identify potential covariates to control for, we examined the relation between mother- and child-related background variables and all outcome variables by means of a MANCOVA. Results demonstrated no significant multivariate effects of children's or mothers' age, children's language skills, number of children a mother has, child gender, mothers' marital status or education level, or valence of the first memory discussed at T1 or T2. Therefore, we did not control for these background variables in the main analyses.

Descriptives of and correlations between the variables are displayed in Tables 2, 3, respectively. Mothers who reported that their children adaptively regulate their emotions and display low emotional lability at T2, also provided high elaboration when reminiscing about positive memories (T2). In addition, mothers' evaluations at T1 were related to children's emotion regulation. More specifically, when discussing negative memories, positive evaluation related to more adaptive emotion regulation in children over time. Moreover, a higher level of negative evaluation during negative memory conversations at T1 related concurrently to more emotional lability and less adaptive emotion regulation over time.

We examined the concurrent (at T2) and cross-lagged relations between mother's reminiscing style and child emotion regulation at T1 and T2. Significant standardized path coefficients are presented in Figure 1 and model summaries are presented in Table 4.

In line with previous studies, our results concerning positive memory conversations showed that mothers' high elaboration was concurrently associated with adaptive emotion regulation (Leyva et al., 2014, 2020), yet no prospective association was observed (Leyva et al., 2014). However, with respect to reminiscing about negative memories, we found no relation between elaboration and child emotion regulation. Additionally, we found elaboration to be unrelated to emotion lability, regardless of the memory valence. Although previous research suggests a positive link between high elaboration and socioemotional development in children (Fivush and Salmon, 2023; Salmon and Reese, 2016), for children to learn adaptive emotion regulation skills, reminiscing content may be just as important as the degree of elaboration. For example, a previous study has shown that mothers' provision of resolutions to negative emotions (but not the level of elaboration) promoted better social problem-solving skills in children (Leyva et al., 2014). Likewise, mothers' causal explanatory language was linked to fewer behavioral problems in children (Sales and Fivush, 2005). However, in the current study we did not code the content of mothers' elaborations. Thus, future studies may focus on the emotional content of mothers' speech as well as their level of elaboration in relation to child emotion regulation.

Moreover, the results contradicted our hypothesis stating that children are more likely to benefit (in terms of their emotion regulation) from discussing negative (instead of positive) memories with their mother. In the current study, elaborative discussion of negative memories may not be linked to children's emotion regulation because our instructions did not specify which type of negative memories mothers should discuss. Previous research indicates that mothers may elaborate on, for example, sad and angry memories differently. That is, mothers elaborate more on how to resolve negative emotions, which may include how to regulate negative emotions, when reminiscing about sad and fearful memories in comparison to angry ones (Fivush et al., 2003). Likewise, mothers use more emotionally negative words and explanations when discussing sad memories over painful memories (Pavlova et al., 2019). As previous studies showed, emotion explanations promote adaptive social behaviors (Leyva et al., 2014; Sales and Fivush, 2005). High elaboration that includes emotion explanations may also aid children's understanding of their own emotions and how to regulate them. Thus, future studies may differentiate between the type of negative memories parents and children reminisce about to shed more light on the relation between parents' level of elaboration and children's emotion regulation.

When reminiscing about positive memories, mothers' evaluations—positive or negative—were unrelated to child emotion regulation. On the other hand, when discussing negative memories, positive (but not negative) evaluations were related to better emotion regulation in children over time. One explanation could be that positive evaluation is especially important during negative memory conversations because children are less likely to engage in such conversations (Van der Kaap-Deeder et al., 2020). Thus, mothers' positive evaluation is essential during such situations, as to motivate the child to be an active participant in the conversation. Indeed, research has shown that when parents are highly affirming, children contribute more to reminiscing conversations (Cleveland and Reese, 2005; Larkina and Bauer, 2010). If children are more engaged in reminiscing, they may get more opportunities to learn about how to regulate their emotions. Given that many studies used a combined score of elaboration and evaluation, these findings highlight the importance of examining possible differentiating effects of components of structure in mothers' reminiscing style.

However, regardless of memory valence, negative evaluation was not significantly related to emotional lability. Notably, the relation between positive, and negative evaluations and emotional lability was marginally significant when discussing negative memories only. While a previous study has shown that negative evaluation was positively related to emotional lability in children (Van der Kaap-Deeder et al., 2020), our relatively small sample size may have prevented us finding such significant prediction. In the current study, negative evaluation when reminiscing about negative memories at T1 correlated with emotional lability in children at T1. Therefore, future studies with larger sample sizes may examine how negative evaluation may hinder children's socioemotional development.

For both positive and negative memories, mothers' autonomy support at T1 was unrelated to children's emotion regulation and emotional lability at T2. Only one former study (Van der Kaap-Deeder et al., 2020) has examined this, using a cross-sectional design and like the present research, no significant associations were found. While it may be important to support children's autonomy to engage children in reminiscing (e.g., Van der Kaap-Deeder et al., 2020), for children to learn how to regulate their emotions, parents may need to employ emotion coaching, which is coaching children to better express, understand, and regulate their emotions (Gottman et al., 1996). Parents also provide sensitive guidance when reminiscing by validating children's emotions and being open to discussing these emotions (Valentino et al., 2022). Indeed, previous studies have shown a positive link between emotion coaching and sensitive guidance, and the development of adaptive emotion regulation skills in children (Dunsmore et al., 2013; Ellis et al., 2014; Speidel et al., 2019, 2020). Although autonomy support shares similar components (e.g., acknowledging children's feelings) with emotion coaching and sensitive guidance, these measures additionally focus on the content of mothers' speech as well as their conversation style. While the current study focused on reminiscing about emotional memories, we have not coded for the emotional content of mothers' speech (e.g., emotion resolutions: Leyva et al., 2014) and have not included the item “acknowledging children's feelings”, which may explain the null findings. Additionally, autonomy support displayed relatively low interrater reliability, which might have contributed to the current non-significant relations. Building on the current results, future research should investigate the differential effects of emotional content and autonomy support (thereby including more items) when reminiscing on children's emotion regulation.

Finally, high levels of child emotional lability at T1 were significantly related to mothers providing less autonomy support at T2 when reminiscing about positive memories. The same relation was only marginally significant when reminiscing about negative memories. Although no former research examined this relation in a reminiscing context, in other studies (e.g., focused on parental requests toward children), high negative affect in children was negatively related to parents' provision of autonomy support via parents' stress (Andreadakis et al., 2020). Parents may have fewer mental and emotional resources to provide autonomy support when their children exhibit maladaptive emotion regulation. This points toward the importance of child characteristics in mother-child reminiscing. Likewise, when reminiscing, if a child is easily frustrated, parents may be more psychologically controlling, for example by imposing their own agenda. Yet, the research focus has been on how mother-child reminiscing may be benefiting children's socioemotional development and thus less is known about child factors associated with mothers' reminiscing style.

Although it was not the primary focus of the current study, we observed very few autoregressive effects for mothers' reminiscing style. Only positive evaluation was related over time. Although most research has focused on individual differences in how mothers reminisce (Fivush et al., 2006), recent studies show that context also plays a role in determining mothers' reminiscing style. For example, parents who reminisced to bond with their child were more autonomy-supportive than parents who reminisced focusing on their child's recall performance (Kulkofsky, 2011). Likewise, research on general parental characteristics show that contextual factors such as daily stress influence mothers' approach to parenting (e.g., Van Der Kaap-Deeder et al., 2019). Although in this study all researchers collecting the data received the same training, mothers personally knew the researcher at T1, but they did not at T2. Having someone they knew vs. not may have prompted differences in how mothers present themselves and reminisce with their child. As this study also demonstrates, child characteristics play a role in mothers' reminiscing style. Future studies could examine how contextual factors (e.g., daily stressors, conversation purpose) and child characteristics together influence mothers' reminiscing style.

In contrast, children's emotion regulation abilities remained relatively stable over time. Mothers who reported that their children were effective at regulating their emotions and exhibited low emotional lability at T1, tended to provide similar reports seven months later. It is important to note, however, that while relatively stable characteristics such as temperament influence children's later emotion regulation skills, these abilities are not fixed. With age and appropriate training, children gradually improve their capacity to regulate emotions (Morris et al., 2007).

This study had several important strengths such as adopting a cross-lagged design and differentiating between reminiscing of positive and negative memories. Also, we measured parents' reminiscing style with observational methods, and child emotion regulation with parent reports thereby preventing common method variance. To add nuance to the literature on the impact of reminiscing on child outcomes, the present inquiry examined whether different aspects of mothers' reminiscing style are uniquely associated with children's emotion regulation. More specifically, we examined possible differential predictive effects of mothers' provision of structure (distinguishing between elaboration, positive evaluation, and negative evaluation) and autonomy support.

This study also had several limitations. First, employing a parent-reported measure of children's emotion regulation may have introduced bias as parent characteristics—such as demographics and psychological well-being can influence parents' ratings of children's emotional characteristics (e.g., Durbin and Wilson, 2012). However, our initial aim was also to observe emotion regulation strategies used by children. By implementing tasks eliciting two different emotions, disappointment and frustration, we had aimed to capture different strategies that children use in different situations as prevalence and adaptiveness of an emotion regulation strategy may change from context to context (Aldao, 2013). However, the experimental tasks did not capture children's negative emotions and, therefore, also not their emotion regulation strategies (see Supplementary material). Thus, future studies may examine the relations between mothers' reminiscing style and children's use of emotion regulation strategies observed in different emotional contexts. Furthermore, we focused on mother-child reminiscing, and the current findings can therefore not be generalized to how children reminisce with other caregivers. Future studies could investigate the role of father-child reminiscing in children's development of emotion regulation skills, as fathers' emotion socialization has been found to relate to their child's emotional development (Kiel and Kalomiris, 2015). On a related note, generalizability of the results may be limited as we collected data through undergraduate students resulting in a homogenous sample (Bornstein et al., 2013). Finally, children are active participants when reminiscing with their parents (Salmon and Reese, 2016), and their level of engagement and interest influence how parents reminisce (e.g., Van der Kaap-Deeder et al., 2020). Consequently, it is important for future studies to consider these bidirectional relations when examining parent-child reminiscing.