Introduction

Executive functions (EFs) can be described as a broad set of complex cognitive skills that enable children to monitor, regulate, and plan their actions, to perform a specific task (McClelland et al., 2010), and/or to solve new problems (Germano et al., 2017). According to a well-documented model (Miyake et al., 2000), it is possible to distinguish three components of EFs: updating working memory (the ability to hold, update, and manipulate information in mind), switching or cognitive flexibility (the ability to switch between different tasks and adapting the focus of attention to new goals and stimuli), and inhibitory control (the ability to suppress dominant responses in favor of more adaptive behaviors). There is however evidence suggesting that the distinction between these three components may be difficult to attain during the preschool period and that a mono-factorial model of EFs may be more appropriate in explaining children’s skills (Wiebe et al., 2008).

EFs are critical in children’s development, as they are associated with academic achievement, problem-solving ability, and social functioning (Diamond, 2013). Studies on preschoolers have indeed found positive links between EFs and a variety of developmental outcomes. Morgan and colleagues (2019), for example, showed that the EF skills of kindergarten children predicted their reading, math, and science achievement later in the second grade of primary school. EFs are also positively correlated with language and literacy skills (Lonigan et al., 2016), as well as with socio-emotional competence (Denham et al., 2015; Longobardi et al., 2022).

Even if direct measurements of EFs have been widely used in studies of preschool children (Silva et al., 2022), educators’ ratings are likewise considered to be valid indicators (McClelland & Cameron, 2019) and have the advantage of increasing the ecological validity of the assessment. In fact, parents and teachers may provide different information about children’s EFs, and their ratings can be useful to capture a wide range of skills that are observed in real-world contexts of school or home life (Toplak et al., 2013). The main aim of the present study was to compare parents’ and teachers’ ratings of preschoolers’ difficulties in the use of EFs and understand how they were related to cognitive and emotional skills.

Parents’ vs. teachers’ ratings

According to Toplak et al. (2013), direct (performance-based) EF measures assess the efficiency of the cognitive processes recruited for behavioral control, whereas indirect (rating-based) measures are more likely to tap issues of rational control, which are concerned with individual goals, beliefs relevant to those goals, and the choice of the most appropriate actions. When assessing the child’s EFs with the latter measures, it is relevant to consider the influence of the type of informants on the obtained measures. Parents and teachers are the most relevant informants and, as a consequence, the more frequently addressed in order to assess children’s skills and difficulties. Both parents and teachers may observe the child’s behavior in a range of meaningful daily life situations in which EFs are needed to display appropriate reactions. Yet, the distinctive interpersonal interactions, activities, and roles occurring in the family and school contexts, which Bronfenbrenner (1986) categorized as different microsystems, present the child with various challenges requiring different levels and types of executive control processes, resulting in a different assessment of their abilities.

Several questionnaires and checklists have been developed and used with parents and teachers (O’Meagher et al., 2019; Silva et al., 2022; Tamm & Peugh, 2019; Toplak et al., 2013). Specifically, the Preschool version of the Behavior Rating Inventory of Executive Function (BRIEF-P; Gioia et al., 20002003) is a standardized questionnaire measuring EFs outside of the clinic or laboratory settings. It consists of a single form that can be completed by different raters, including parents, caregivers, preschool teachers, and/or childcare workers. The BRIEF-P evaluates children’s difficulties in five different domains of EF functioning: Inhibit, Shift, Emotional Control, Working Memory, and Plan/Organize. In addition, three composite indexes (the Inhibitory Self-Control, Flexibility, and Emergent Metacognition Indexes) and an overall composite index (Global Executive Composite) can be also computed.

The aims of the present study were to investigate (a) whether parents’ and teachers’ ratings were correlated to each other (a stability issue, according to Bornstein et al., 2017) and (b) whether the mean levels of children’s EF difficulties were comparable across parents’ and teachers’ judgments (a consistency issue: Bornstein et al., 2017). With respect to the first point, a meta-analysis by Achenbach et al. (1987) reported an average correlation from 0.27 to 0.60 between different measures of parents’ and teachers’ reports. More recently, Dekker and colleagues’ (2017) study of EFs in school-aged children, which utilized both direct and indirect measures, revealed significant correlations between a direct measure of working memory and both parent and teacher reports, as well as moderate correlations between the two types of informant reports (ranging from 0.31 to 0.41). Other studies, however, reported little agreement between parents’ and teachers’ reports in typically developing children (Mares et al., 2007; Schneider et al., 2020).

Regarding the second point, there is evidence indicating that parents and teachers may differ in the mean levels of EF difficulties attributed to children (Bausela-Herreras, 2018; Germano et al., 2017). Several studies examining atypical populations reported that teachers’ scores had a higher probability of falling above the cut-off for clinically elevated symptoms, as compared to parents’ scores. Mares et al. (2007), for instance, performed a secondary data analysis on 240 children diagnosed with attention-deficit hyperactivity disorder and concluded that teachers reported more variety and severity of EF impairments than did parents. Also, the study by O’Meagher et al. (2019) found differences in parents’ and teachers’ reports of EFs, with the former reporting less difficulties than the latter on inhibition, working memory, planning, self-control, and metacognitive skills. The direction of the inter-rater discrepancies may however vary based on who is being rated (i.e., children with typical or atypical development), the specific subscales examined, and whether standardized or raw scores are analyzed. In particular, Schneider et al. (2020) found that teachers rated typically developing children as having more EF difficulties than parents when the analyses were based on standardized scores. Parents, on the other hand, rated children with ADHD as having more problems than teachers when the evaluation was based on raw scores. Likewise, differences in the way parents and teachers evaluated children’s problems with specific EF domains have been shown by Germano et al. (2017) in a sample of children with learning disorders. Both the informants, in fact, reported difficulties related to EFs, but where parents reported more problems than teachers in the emotional control, planning, and material organization subscales, teachers reported more difficulties than parents in the monitoring subscale.

The discrepancy in the type of difficulties reported by teachers and parents could be due to several factors. One is that EFs have different roles in the home and school contexts. Thus, both parents’ and teachers’ reports may be needed to have a comprehensive evaluation of the child’s skills because they may have different understandings of the same behavior. Another factor is that, given the higher demand for EF skills in the school setting than at home, and because teachers may have a better sense of which behaviors are normative, they may be better able to compare EF skills and difficulties across different children (Tamm & Peugh, 2019). Thus, parents and teachers may approach the same behaviors from different perspectives, resulting in discrepancies in their reports.

Gender differences in executive functions

Gender differences have been often observed in the development of EFs, although research on this question is inconsistent. Thus, this study aims also to ascertain whether gender differences in EFs were equally apparent across teachers’ and parents’ reports. While some research reported significant correlations between gender and EFs (e.g., Palomino & Brudvig, 2022; Yamamoto & Imai-Matsumura, 2019), other studies found no significant relations (Slot & von Suchodoletz, 2018; Wiebe et al., 2011). Cultural differences might in part explain these mixed results, since most of the studies reporting gender differences in EFs were conducted in the USA, while the study by Gestsdottir and colleagues (2014), conducted in France, Germany, and Iceland, and the study by Baptista and colleagues (2016), conducted in Portugal, did not find any differences.

When gender differences are found, however, the results are consistent in showing that girls outperform boys (Mulder et al., 2014; Palomino & Brudvig, 2022). A recent review by Schirmbeck and colleagues (2020) confirmed that most studies assessing gender differences in EFs reported better performance for girls than boys: girls outperform boys on direct tasks, and parents and teachers rate them higher than boys on indirect tests. Anyhow, even if most results go in the same direction, it is important to consider the potential bias of parents and teachers, which scholars have identified as a possible underlying factor contributing to divergences between indirect and direct assessments (Grabell et al., 2015; Schirmbeck et al., 2022; Thorell et al., 2013). Parents’ responses might be influenced by the fact that they have gender-specific cultural expectations toward the behaviors of their children (e.g., girls are typically judged as more attentive and well-behaved than boys, whereas boys are judged as more active and aggressive than girls; see Kollmayer et al., 2018, for a review). Likewise, teachers are subject to gender stereotypes when evaluating the behavior of their students (Glock, 2016).

Relations of executive functions with language, emotion comprehension, and nonverbal intelligence

A further goal of the present study was to determine whether parents’ and teachers’ indirect assessments of EFs were significantly associated with children’s performance on direct tests evaluating emotion comprehension, language ability, and nonverbal intelligence. Previous studies using the BRIEF-P have primarily looked at the potential correlations between direct and indirect EF assessments, reaching mixed results (e.g., Garon et al., 2016; Mahone & Hoffman, 2007; Miranda et al., 2015). However, to our knowledge, few studies investigated the utility of the BRIEF-P scores in predicting different cognitive functions. Obtaining positive results in this domain would lend credence to the use of indirect EF measures in applied and research settings (e.g., Ferrier et al., 2014).

Children’s EFs and emotional and linguistic competences are strictly related to each other. EFs and language develop rapidly in early childhood (Blair et al., 2005; Halliday, 2006), and the relationships between them have been substantiated by numerous studies (e.g., Palomino & Brudvig, 2022; Slot & von Suchodoletz, 2018). For example, White and colleagues (2017) investigated the relations between EFs and language skills in a sample of preschoolers and showed how EFs were strongly related to various components of language. More specifically, they found EF competence to be predictive of vocabulary, syntax, and language learning, when controlling for age, gender, and ethnicity. Similarly, Marini and colleagues (2020) highlighted the importance of specific EF skills in supporting the development of language competences: they found that preschoolers’ errors in an inhibition task were negatively correlated with phonological discrimination, grammatical comprehension, and sentence completion skills, as well as with the ability to select and produce words that were morphologically, semantically, and pragmatically appropriate to the task.

The relations of EFs with emotion understanding and nonverbal intelligence have been likewise demonstrated by several studies. With respect to the first point, Ferrier et al. (2014) examined 175 children aged 35–60 months by using a longitudinal design with 6-month lags between successive assessments. Consistent with the idea that both EF and emotion regulation are core aspects of self-control, the authors found that T1 emotionality was associated with T1 performance on direct EF tasks and the T2 BRIEF-P composite, whereas T1 scores in direct EF tasks were related both with emotionality and the BRIEF-P composite at T2. Regarding the second point, Rahbari and Vaillancourt (2015) reported that, in a sample of 126 preschool children, nonverbal intelligence, assessed using the verbal and performance subtests from Wechsler Preschool and Primary Scale of Intelligence–Third Edition (WPPSI-III), was concurrently and longitudinally associated with the BRIEF-P (especially the Working Memory subscale).

In summary, while the strong relations between EFs and both cognitive and emotional skills are evident, what is lacking to date is a comprehensive comparison of teachers’ and parents’ ratings and a thorough understanding of how the indirect assessments of different raters are correlated to children’s competences.

The present study

Considering the literature discussed above, the goals of the present study were to investigate differences in the way in which parents and teachers rated preschool children’s EFs and how their ratings were related to children’s emotional, linguistic, and cognitive skills. In particular, the objectives were (a) to compare parent and teacher ratings of EFs in preschoolers, to examine possible differences related to the context of assessment (home vs. school), (b) to analyze whether parent and teacher ratings of EFs differed as a function of children’s gender, and (c) to analyze the way in which parent and teacher ratings of EFs were related to children’s linguistic, emotional, and nonverbal intelligence skills.

Methods

Participants

A total of 130 children aged 34 to 71 months (M = 51.31, SD = 10.01; 68 boys and 62 girls), their mothers and teachers took part to the present study. They were recruited in several preschools and kindergartens in Rome (Italy). All the involved children and teachers had been attending school for at least 6 months at the time of data collection. No child had sensory, intellectual, language, or learning deficits, as reported by parents and teachers. In the present study, parents were always represented by mothers. The socio-economic level of families was medium or medium–high, as indicated by parents’ education: indeed, 33.1% of mothers and 40.8% of fathers had a high school diploma, and 49.3% of mothers and 32.3% of fathers had a university degree. An information sheet about the aims of the study was given to parents, who were also asked to sign an informed consent on behalf of their children and to allow teachers to fill in the BRIEF-P. Each parent provided also her own written consent to participate. The ethical committee of the local institution approved the present study, and all procedures were conducted in accordance with the ethical principles set out in the Declaration of Helsinki.

Measures

Executive functioning

The Behavior Rating Inventory of Executive Function–Preschool Version (BRIEF-P; Gioia et al., 2003; Italian adaptation by Marano et al., 2014) was completed by both parents and teachers to assess the executive functioning of children; specifically, this standardized test is appropriate for children from 2 to 5 years, and it is made up of 63 items that required parents and teachers to report how often each specific child’s behavior represented a problem in the last 6 months on a 3-point Likert scale ranging from 1 (never) to 3 (often). Higher scores indicate greater children’s difficulties and, hence, lower executive functioning. The BRIEF-P provides both a total composite score (the Global Executive Composite) and scores for five distinct scales (inhibition, shift, working memory, emotional control, and plan/organize). The Inhibition scale is composed of 16 items (e.g., “Is impulsive” or “Acts too wild or out of control”) and evaluates children’s ability to control impulses and to stop a given behavior at the appropriate time. The Shift scale, composed of 10 items (e.g., “Has trouble with activities that involve more than one step” or “Is upset by change in plans or routines”) measures children’s ability to shift from one activity to another one in a flexible way. The Working Memory scale, composed of 17 items (e.g., “Has trouble remembering something, even after a brief period of time” or “Cannot stay on the same topic when talking”) assesses children’s ability to keep information in mind. The Emotional Control scale, composed of 10 items (e.g., “Mood changes frequently and “Becomes upset too easily”) measures the ability of children to modulate their emotional reactions. Finally, the Plan/Organize scale, composed of 10 items (e.g., “When instructed to clean up, puts things away in a disorganized, random way” and “Gets caught up in the small details of a task or situation and misses the main idea”) evaluates if children are able to anticipate future events, to set goals, and to do appropriate steps to complete an ongoing task.

In the present study, the Global Executive Composite and the scores of the five BRIEF-P subscales were used. Moreover, on the basis of the results of an exploratory factor analysis, Gioia et al. (2003) proposed the use of three summary indices (see also Sherman & Brooks, 2010): the Inhibitory Self-Control index, which is the sum of the Inhibition and Emotional Control scales (range: 26–78) and evaluates the children’s ability to regulate and inhibit their actions, reactions, and behaviors; the Flexibility index, which is the sum of the Shift and Emotional Control scales (range: 20–60) and assesses the children’s ability to move flexibly among different actions, reactions, emotions, and behaviors; and lastly, the Emergent Metacognition index, which is the sum of the Working Memory and Plan/Organize scales (range: 70–81) and evaluates children’s ability to organize, plan, carry out, and maintain future-oriented problem-solving behaviors. The three-factor solution has been later replicated in the Italian adaptation of the BRIEF-P (Marano et al., 2014), and these indices have been already used in previous studies (e.g., Longobardi et al., 2022; Spataro et al., 2022).

Emotional comprehension

The Test of Emotion Comprehension (TEC; Albanese & Molina, 2008; Pons & Harris, 2000) was administered to evaluate children’s affective competence. Specifically, this test is used to assess emotional knowledge in children aged 3 to 11 years and measures nine different dimensions: (a) emotion recognition based on facial expression, (b) understanding of the external causes of emotion, (c) emotion understanding based on desires, (d) emotion understanding based on beliefs, (e) understanding of the influence of a reminder on the current emotional state, (f) understanding of the capacity to control a felt emotion, (g) understanding of the capacity to hide an emotion, (h) understanding of mixed emotions, and (i) understanding of moral emotions. Children were initially presented with a picture booklet with a series of cartoon scenarios at the top of each page. At the bottom of the same page, four possible emotional outcomes conveyed by different facial expressions were depicted. Children listened to a story that the experimenter told them as they looked at the cartoon scenario and were then asked to choose a facial expression representing the actual emotion felt by the protagonist of the story. If they answered correctly, one point was assigned for each component. Thus, the total score ranged from 0 to 9, with higher scores indicating better emotion comprehension skills. In previous studies (Pons & Harris, 2005; Pons et al., 2002), the TEC has shown good test–retest reliability when administered to children aged 3–10 years old after 3 months [r(18) = 0.84] and a good stability after 13 months [r (40) = 0.68]. Furthermore, in the validation study of the Italian version (Molina & Di Chiacchio, 2008), the instrument was found to have good internal consistency (KR-20 = 0.79).

Language competence

The Test of Language Evaluation (Test di Valutazione del Linguaggio: TVL; Cianchetti & Fancello, 1997), which is appropriate for children aged from 2 to 6 years, was administered to evaluate language abilities. The TVL assesses four different language abilities: comprehension of words and sentences, repetition of sentences, production of words from picture cards, and spontaneous speech production on a prescribed theme. In the present study, only the first three sections were employed to have a global assessment of receptive and productive abilities, including lexical and syntactic knowledge. The first section consists of two parts assessing comprehension of single words (referred to objects, actions, body parts) and simple and complex sentences. In the second section, children were asked to repeat different sentences of increasing length to evaluate the knowledge of morphological and syntactic rules of language and the ability to articulate words. Finally, in the third section, children were required to name 40 objects and actions depicted on picture cards, to evaluate productive vocabulary. For the purposes of statistical analyses, we used the total scores resulting from the sum of the scores assigned to these three sections (range: 0–171). The TVL has shown excellent test–retest reliability in the normative study (0.93 for comprehension, 0.87 for repetition, and 0.96 for naming; Cianchetti & Fancello, 1997).

Nonverbal intelligence

The Coloured Progressive Matrices (CPM: Raven, 1984) were used to assess nonverbal intelligence (i.e., the Spearman’s g factor), defined as the ability to form perceptual relations and to reason by analogy, independent of language and formal schooling. The instrument consists of 36 items arranged in three sets of 12 items, each in increasing order of difficulty. Each item contains a non-meaningful geometrical figure with a missing piece. Below the figure, four alternative pieces are presented: the children’s task is to choose the piece that completes the figure in the correct way (only one response is correct). The final score is the total number of correct answers (range: 0–36). In a previous study by Albanese and colleagues (2010), internal consistency ranged from 0.60 to 0.98, with a median of 0.90. The median test–retest value was approximately 0.82. Lastly, concurrent validity coefficients between the CPM and the Stanford-Binet and Weschler scales range from 0.54 to 0.88 (Albanese et al., 2010).

Procedure

Each child was administered in a randomized order the TEC, TVL, and CPM tests by a trained experimenter during a single session lasting about 30 min. Following the same procedure adopted in previous studies (Longobardi et al., 2022; Spataro et al., 2022), the administration took place in a quiet room of the kindergartens. At the same time, both parents and teachers were given a comprehensive explanation of the BRIEF-P and were asked to fill in and return the questionnaire within 1 week.

Statistical analyses

Differences between parents’ and teachers’ scores in the BRIEF-P and gender differences in children’s BRIEF-P scores were analyzed with a series of t-tests for paired samples (in the first case) or for independent samples (in the second case). The Welch correction was applied whenever the assumption of equal variances was violated. The stability of the BRIEF-P scores across parents’ and teachers’ reports and the associations between the BRIEF-P, TEC, TVL, and CPM scores were examined with Pearson’s correlations. Bayes factors (BF10) are always reported to quantify the evidence in support of the alternative hypothesis (i.e., the presence of significant differences for t-tests or the presence of significant associations for Pearson’s correlations). According to van Doorn et al. (2021), Bayes factors between 1 and 3 are considered to be weak evidence, Bayes factors between 3 and 10 are considered moderate, and Bayes factors greater than 10 are considered strong evidence. Finally, to assess which scales of the BRIEF-P were associated with children’s performance in the three direct tests used in the present study (TEC, TVL, and CPM) and to remove the confounding effects of age, we conducted a series of regression analyses, using the forward stepwise regression method. Age was always entered in the first step (since it was positively and significantly associated with children’s performance in all the direct tasks: r (128) = 0.34, p < 0.001 for the TEC, r (128) = 0.48, p < 0.001 for the TVL, and r (128) = 0.47, p < 0.001 for the CPM), while the three summary indices of the BRIEF-P were added as independent variables in the second step. We chose to use these indices because there were high intercorrelations between the BRIEF-P subscales (Pearson’s correlations ranged from 0.45 to 0.86 in teachers’ data and from 0.25 to 0.74 in parents’ data), leading to problems of multicollinearity.

Results

Table 1 reports descriptive statistics for the measures assessed in the present study. As can be noted, the values of skewness and kurtosis were generally comprised between − 2 and + 2 and were therefore acceptable, considering the recommendations set forth by George and Mallery (2010). We however conducted both parametric and non-parametric analyses. Since the overall pattern of results was highly comparable, only the former analyses will be reported in the following sections.

Table 1 Descriptive statistics (N = 130)

Differences between parents’ and teachers’ scores in the BRIEF-P and stability

Table 2 shows that significant differences in the paired-sample t-tests were observed for the Inhibition and Emotional control subscales, the Flexibility index, and the Global Executive Composite. The associated Bayes factors (BF10) were 5.49, 18.93, 7.78, and 1.12, respectively. Thus, except for the Global Executive Composite, the evidence in support of the existence of significant differences between parents’ and teachers’ scores ranged from substantial to strong. In all cases, parents’ scores were higher than teachers’ ones, suggesting that the former were more likely to judge the children’s behaviors as problematic.

Table 2 Mean scores reported by parents and teachers in the BRIEF-P, together with the results of t-tests and correlational analyses

Table 2 also reports Pearson’s correlations between parents’ and teachers’ scores in the BRIEF-P. Significant associations of moderate size were obtained in all cases, except for the Plan/Organize subscale. The Bayes factors (BF10) were all greater than 22.52, suggesting that there was strong evidence in favor of the significance of the above correlations. These results suggest that children who were assigned high (or low) scores by parents were likewise assigned high (or low) scores by teachers. In other words, there was considerable stability between parents’ and teachers’ scores.

Gender differences in children’s BRIEF-P scores

Table 3 shows that significant differences in children’s BRIEF-P scores were observed in teachers’ reports, but not in parents’ reports. Specifically, teachers’ scores differed between boys and girls in the Inhibition, Working Memory, and Plan/Organize subscales, as well as in the Inhibitory Self-Control and Emergent Metacognition indexes and the Global Executive Summary. In all cases, girls were reported to have less EF difficulties than boys. The Bayes factors for these differences were comprised between 3.23 and 79.19, suggesting that the evidence in favor of the existence of significant differences between boys and girls ranged from substantial to very strong.

Table 3 Differences between boys and girls in the BRIEF-P scores, as reported by parents and teachers, together with the results of the t-tests for independent samples

Correlations between BRIEF-P, TEC, TVL, and CPM

Table 4 illustrates Pearson’s correlations between parents’ and teachers’ scores in the BRIEF-P and the three direct tests used in the present study (TEC, TVL, and CPM). Starting with the TEC, significant negative associations were observed with teachers’ scores, but not with parents’ scores. The inspection of the Bayes factors suggests that, for teachers, there was substantial or strong evidence in favor of the hypotheses that the TEC was negatively correlated with the Shift, Working Memory, and Plan/Organize subscales, the Flexibility and Emergent Metacognition indexes, and the Global Executive Summary.

Table 4 Pearson’s correlations between parents’ and teachers’ BRIEF-P and the TEC, TVL, and CPM scores. Bayes factors (BF10) are reported in parenthesis

Similar results were observed for the TVL. For parents, we found that the TVL was negatively associated with the Shift and Working Memory subscales, the Flexibility index, and the Global Executive Summary. However, the inspection of the Bayes factors suggests that strong evidence in support of the alternative hypothesis was only available for the Shift subscale. In contrast, the negative associations with teachers’ scores were more robust in size and always significant; furthermore, the Bayes factors provided strong or very strong support for the alternative hypothesis in all cases—except for the Emotional control subscale and the Inhibitory Self-Control index (for which there was only anecdotal evidence).

Lastly, for the CPM, the analyses show that the negative associations with the BRIEF-P scores were significant for both parents and teachers, although being greater in size in the latter case. In agreement, the inspection of the Bayes factors suggests that the alternative hypotheses were more strongly supported when the BRIEF-P was filled in by teachers, than by parents.

Regression analyses

The final β coefficients resulting from the hierarchical regressions are reported in Tables 5 and 6.

Table 5 Stepwise regression analyses predicting TEC, TVL, and CPM total scores from teachers’ BRIEF-P
Table 6 Stepwise regression analyses predicting TEC, TVL, and CPM total scores from parents’ BRIEF-P

Beginning with teachers’ data (Table 5), we found that age and the Inhibitory Self-Control index of the BRIEF-P was associated with children’s performance in the TEC and TVL tasks, while age and the Emergent Metacognition index of the BRIEF-P was related to performance in the CPM task.

Turning to parents’ data (Table 6), the results were similar. Specifically, age and the Inhibitory Self-Control index of the BRIEF-P was associated with children’s performance in the TVL task, while age and the Emergent Metacognition index of the BRIEF-P to the performance in the CPM task. The main difference with teachers’ results was that children’s performance in the TEC task was associated with age, but not with the Inhibitory Self-Control index of the BRIEF-P.

Discussion

The present study aimed at investigating preschoolers’ EFs with a multi-informant approach, by comparing parent and teacher reports, and evaluating the associations of EF ratings with children’s emotional, linguistic, and cognitive skills. Previous research suggested that investigating EFs using both teacher or parent ratings represents a more ecological and non-intrusive strategy, especially with preschoolers, for which the administration of direct tasks may be difficult (Longobardi et al., 2022; Schroeder & Kelley, 2010; Spataro et al., 2022; Thorell and Catale, 2013). Moreover, an increasing number of studies assessed children’s EFs outside the laboratory using standardized questionnaires (Hertz et al., 2019; Hooper et al., 2020; Longobardi et al., 2022; Spataro et al., 2022), because they provide an index of how children’s difficulties in EFs, as reported by significant others (parents or teachers), have an impact on everyday life (Sherman & Brooks, 2010).

The analyses confirm the existence of significant differences between the BRIEF-P reports produced by parents and teachers, with parents rating their children’s EF difficulties as higher than teachers. Specifically, the findings showed that parents are more inclined to detect difficulties in the inhibition and emotional control subscales of the BRIEF-P, as well as in the flexibility index and the Global Executive Composite. The present results could be linked to the different contexts (home versus school) in which EFs are observed (Tamm & Peugh, 2019) and are in line with those obtained in a number of previous studies evaluating both typically developing and atypical populations. The normative study by Gioia et al. (2003), for example, reported higher BRIEF-P ratings for parents than for teachers. Similarly, Kenealy (2002) showed that teachers report less EF impairments than parents in a sample of children with ADHD. Yet, other studies dealing primarily with clinical samples found just the opposite pattern (Mares et al., 2007; O’Meagher et al., 2019).

Even if parents and teachers may approach children’s behaviors from different perspectives, as reflected in the differences that emerged between their reports, the data should not be taken as evidence that teachers’ judgments were more valid or reliable than parents’ ones. In fact, we observed significant, moderate associations between parents’ and teachers’ scores, suggesting that the two types of informants were quite consistent in evaluating children’s rank order. That is, parents and teachers were likely to detect similar discrepancies between children having different levels of EF difficulties. Again, it should be taken in mind that interrater agreement may vary considerably, depending on the nature of the sample. Indeed, while Gioia et al. (2003) reported moderate correlations between parents’ and teachers’ ratings in a normative sample of typically developing children, low and largely nonsignificant correlations have been instead observed in children with ADHD (e.g., Mares et al., 2007; but see Schneider et al., 2020).

Differences between the two types of respondents were also detected concerning gender differences in EFs. Indeed, while in the parents’ reports, no differences emerged according to children’s gender, teachers rated boys as having more EF difficulties than girls in the inhibition, working memory, and plan/organize subscales of the BRIEF-P, as well as in the inhibitory control and emergent metacognition indexes, and the Global Executive Composite. Such a discrepancy may reflect either the cross-situational differences in children’s behaviors or the differences in raters’ perceptions and expectations (Mares et al., 2007). As mentioned above, teachers may take advantage of the opportunity to compare several children with each other and this may allow them to have a greater sensitivity in understanding and identifying gender differences. Alternatively, the differences between parents’ and teachers’ reports that emerged in the present study might reflect the different behaviors that boys and girls show in different contexts and the different interactions they have with the different informants. The school context is typically more structured than the home environment, and children may be required to regulate and inhibit their behaviors more often in the former than in the latter context. Thus, the disruptive and potentially aggressive behaviors of boys are more likely to emerge at school than at home. Anyway, regardless of the correct explanation, the female advantage observed in teachers’ reports is well in line with a recent review by Schirmbeck et al. (2020), showing that both direct and indirect measures of EFs tend to detect a better performance of girls over boys.

An important aim of the present study was to analyze the relationship between parents’ and teachers’ EF reports and the children’s performance in direct tests assessing emotional understanding, language competence, and nonverbal intelligence. This is a critical point for demonstrating the utility of indirect EF measures in applied and research settings (Ferrier et al., 2014). The results of correlational analyses showed that the EFs rated by teachers—but not parents—were negatively associated with children’s emotional comprehension, suggesting that children having more EF difficulties achieved lower scores in the TEC. The relationship between EFs and emotion understanding is indeed widely documented in the literature (for a review, see Schmeichel & Tang, 2014). Children with good EF skills can easily shift from their own perspective to the perspective of other ones, moving beyond the immediate environmental context and reflecting on their own and others’ mental states (Devine & Hughes, 2014). The finding of stronger associations in teachers’ than in parents’ reports may likely reflect the fact that the multiple interactions occurring in the school context (with both peers and adults) offer teachers more opportunities to detect and observe the children’s ability to activate their EFs for understanding their own and others’ emotions.

Similar results were also obtained for the links between EFs and language abilities and between EFs and nonverbal intelligence. Specifically, findings from correlational analyses suggest that there were negative relationships between preschoolers’ EF difficulties and both language and nonverbal intelligence abilities (again, suggesting that children who had less EF difficulties tend to have higher performance in the TVL and CPM tasks) and that these relationships were somewhat stronger in the case of teachers’ than parents’ reports. These results are in line with previous studies (Longobardi et al., 2022; Martins et al., 2016; Wade et al., 2018) showing how EFs are an important prerequisite for the development of verbal and nonverbal cognitive skills (Carlson et al., 2004; Hughes & Ensor, 2007; Marcovitch et al., 2015; Wade et al., 2018). In particular, regarding the relationship between EFs and language abilities, it has been repeatedly demonstrated that children with difficulties in language development are more likely to experience significant deficits of EFs and self-regulation (Clark et al., 2021; Diamond, 2006; Singer & Bashir, 1999; Spataro et al., 2022).

Finally, the regression analyses indicated that the inhibitory self-control index of the BRIEF-P was associated with the children’s performance in direct tasks tapping language ability and emotional comprehension, whereas the emerging metacognition index was related to the performance in the task assessing nonverbal intelligence. As stated above, the inhibitory self-control index is made up of the combination of the Inhibition and Emotional Control subscales of the BRIEF-P. In agreement, there is now consistent evidence suggesting that the development of inhibitory functions plays a critical role in language processing (e.g., Gandolfi & Viterbori, 2020; Larson et al., 2020). Among the other things, it has been proposed that inhibition would be necessary to select the right word over competing alternatives, to integrate different elements into a coherent sentence representation, to monitor morphological inflections or gender and number agreements, to inhibit the tendency to produce inappropriate expressions, to facilitate sentence comprehension (especially when it is necessary to revise initial misinterpretations of ambiguous sentences), and to resolve lexical ambiguity (Gandolfi & Viterbori, 2020). Inhibitory skills are likewise necessary for emotional modulation, as they enable children to suppress unwanted emotional experiences and inappropriate emotional displays (Carlson & Wang, 2007). In fact, deficits in EFs and emotion regulation tend to co-occur in atypical and at-risk populations (e.g., children with ADHD: Barkley, 1997), and the development of the two domains share similar timetables (Bell & Wolfe, 2004). On the other hand, the emergent metacognition index is made up of the working memory and plan/organize subscales of the BRIEF-P. The present findings, demonstrating the utility of this index in relation to the children’s performance in nonverbal intelligence tasks, ally with a large body of research showing that working memory and fluid intelligence develop in concert and share bidirectional relations in the preschool period (e.g., de Abreu et al., 2010; Fry & Hale, 2000; Giofrè et al., 2013).

Conclusions

In summary, the present study suggests that EF skills are important for children’s successful adjustment in many important areas of life, and therefore, their development should be assessed and monitored during the preschool period. Direct measurements, which are useful for testing the efficiency of the EF processes involved in laboratory tasks, may not be sufficiently ecologically reliable. Findings from this study indicate that the indirect measurement of preschoolers’ EFs through the BRIEF-P represents a valid option, because it allows researchers to detect and measure EF difficulties in everyday life; more interestingly, we showed that the BRIEF-P scores can be fruitfully related to the children’s emotional, language, and cognitive abilities in direct tasks, above and beyond the effects of age. The results obtained confirm that parents and teachers can have different perspectives about children’s EFs, as they observe children in situations that may activate different behaviors. Although teachers’ reports seem to provide more useful and discriminant information, as they are better suited to differentiate between boys and girls and are more tightly associated with direct measures, it is important to reiterate that the two types of reports were significantly correlated to each other and could be therefore used together to assess children’s EFs in everyday contexts.

The present study has a number of limitations that must be taken into account. First, the data were cross-sectional and correlational, and this prevented us from making strong inferences about the causality and the direction of the observed associations. This is especially true for the results of the regression analysis, in which the summary indices of the BRIEF-P were used as concurrent predictors of children’s performance in the direct tests. The results of these analyses cannot be taken as evidence of the unidirectionality of the relations between EFs and cognitive-emotional skills, because, as stated above, previous longitudinal studies demonstrated that most of the assessed relations are likely to be bidirectional during the preschool period (Slot & von Suchodoletz, 2018; Xing et al., 2022). Second, a convenience sample limited to typically developing children was used: it would be opportune to verify the generalizability of these findings to children with atypical development. Lastly, even if investigating preschoolers’ EFs using a more ecological measurement was a specific aim of the study, only questionnaires were used. Further research could be realized by combining both direct and indirect measures of EFs. Despite these limitations, the findings suggest that indirect assessments of EFs are associated with preschoolers’ performance in direct tasks and that training programs aimed at increasing EFs can have a positive impact on the development of emotional, linguistic, and cognitive abilities (Diamond & Lee, 2011; Fisher & Happé, 2005; Longobardi et al., 2022; Traverso et al., 2015).