Educational Practices and Child Behaviors: Mothers’ and Teachers’ Evaluation

– The objective of this study was to describe and compare mothers’ and teachers’ reports concerning their educational social skills and negative practices and children’s behavioral problems and social skills. The mothers and teachers of 81 children participated in this study. The children were assigned to four groups depending on whether they presented problems exclusively at home, school, in both, or none of these contexts. Valid instruments measured educational practices and child behavior. The results show that: (a) mothers and teachers agreed that children with problems in both contexts presented the highest level of impairment; (b) the mothers more frequently reported skills such as affection, adopted negative practices, and identified problem behaviors; (c) the teachers reported more skills such as limit setting. The conclusion is that teachers’ and mothers’ practices differ, and such information can guide preventive programs and interventions.

In this study, positive child-rearing practices and positive educational practices are considered synonymous with educational social skills (HSE) as defined by Bolsoni-Silva et al. (2016), whose response classes are organized into communication, affection, and limit setting.Achenbach and Rescorla (2001) classify behavioral problems into externalizing (e.g., aggressiveness, disobedience) and internalizing problems (e.g., shyness, anxiety).The reports of parents and teachers regarding educational practices are relevant (Bolsoni-Silva & Mariano, 2018;Nunes et al., 2013;Mariano & Bolsoni-Silva, 2018;Vafaeenejad et al., 2019;Zalewski et al., 2017).Smith and Sheridan (2019) conducted a meta-analysis and verified that when teachers promote parental involvement and adopt positive communication practices, collaborative planning, and problem-solving strategies, students benefit from greater socio-emotional, behavioral, and academic development.
The reports of parents and teachers regarding children's behaviors are not always convergent.Some studies found divergent (Korsch & Petermann, 2014;Lavigne et al., 2015;Rescorla et al., 2014;Rudasill et al., 2014), while others found convergent reports (Bernedo et al., 2014;Ercan et al., 2015).In some of the studies in which parents and teachers disagreed, teachers reported behavioral problems more frequently than parents (Santiago et al., 2016), while in others, parents reported problems more frequently (Rescorla et al., 2014).Thus, no consensus has been achieved thus far.The same occurs with social skills; in some studies, parents reported more skillful repertories than teachers (Korsch & Petermann, 2014), while other studies report the opposite (Bolsoni-Silva & Loureiro, 2016).These differences and gaps in knowledge may result from multiple factors, one of which is possibly the use of self-report instruments, such as scales and questionnaires, or interviews and observation.These distinct methods produce different and complementary data (Bolsoni-Silva et al., 2011).Bernedo et al. (2014) assessed behavioral problems, impulsiveness, and attention in a sample of 104 foster children, concluding that parents and teachers agreed regarding externalizing problems, but disagreed regarding anxiety, depression, somatic symptoms, and impulsiveness/ attention.Rohenkohl and Castro (2012) addressed a sample of 59 preschoolers and verified that the parents more frequently reported total problems and somatic symptoms, anxiety, and depression.Regarding Attention-Deficit Hyperactivity Disorder (ADHD), Ercan et al. (2015) addressed 417 children aged between six and 14 years old and verified that parents and teachers agreed in general scales measuring internalization, externalization, and total problems.On its turn, Korsch and Petermann (2014) found divergences between parents and teachers reporting on 160 preschoolers.The parents reported social skills more frequently than teachers, though both parties agreed to behavioral problems.Rescorla et al. (2014) compared parents and teachers of 27,962 children from 21 different countries, reporting various divergences between the respondents; parents reported more problems than teachers.Llanes et al. (2020) examined the prevalence of ADHD and comorbidities among 180 pre-school and school-aged children using CBCL and TRF and verified that parents more frequently reported behavioral problems with little agreement between the reports.
De Los Reyes et al. ( 2015) conducted a meta-analysis using data from 341 studies published between 1989 and 2014.The meta-analysis was intended to address studies comparing different informants (parents, teachers, and children/adolescents) regarding children and adolescents' mental health.The main results reported were: (a) low to moderate correlation between the respondents concerning internalizing and externalizing problems; (b) the findings reported by the various studies investigating behavioral problems varied widely; (c) greater agreement was found between the reports of parents and teachers regarding externalizing problems; (d) the reports of fathers and mothers regarding the same child, that is, considering the same context, more frequently concurred than when the remaining pairs of respondents were compared.The authors considered that the divergent assessments resulted from two factors: the first is that the behaviors of children/adolescents were assessed in different contexts and the second was that different instruments with different psychometric properties were used.
Considering the previous discussion, the question "Why is there so much divergence?" may be answered by Lavigne et al. (2015) and De Los Reyes et al. (2015), that is, school and family are different contexts, each with its inherent demands.For this reason, children may present different behavioral patterns in these contexts, which in turn influence each other.Lavigne et al. (2015) note that family conflict predicted discrepant reports, and Santiago et al. (2016) verified that conflict with teachers was the strongest predictor of divergent assessments.Santiago et al. (2016) addressed 732 children attending kindergarten to the 4 th grade and verified that teachers reported behavioral problems more frequently than mothers, which influenced teacher-child conflict.
These findings suggest that conflicts may influence the assessment of children's behaviors in different contexts.Heatly and Votruba-Drzal (2017) addressed 1,364 1 st grade children and found that teacher-child conflict hindered the children's school engagement.The authors also noticed that the mothers' practices influenced the children's conflict with teachers and school engagement, noting that conflicting family interactions may lead to disruptive behaviors, which in turn may prompt conflicts with teachers.
There are some peculiarities regarding practices used to deal with behavioral problems.Silveira and Wagner (2009) addressed a small sample (4 parents and 4 teachers) and verified that parents and teachers adopted the same inductive and coercive practices to deal with problem behavior, though teachers used coercive practices more frequently than parents, but adopted the same amount of inductive practices as parents.In the same direction, Garcia et al., (2016), addressing 417 parents of children/adolescents aged from seven to 12 years old and 199 teachers verified that both parents and teachers adopted the same strategies.However, teachers less frequently adopted monitoring and power affirmation practices.
In summary, literature shows that, even though studies assessed children's behavioral problems among parents and teachers simultaneously, little agreement was obtained.At the same time, none of the studies differentiated the children exclusively presenting behavioral problems in one context (family or school) or in both contexts, which may be why parents and teachers report divergent reports.In this sense, there is a need for studies to investigate positive and negative educational practices, addressing both parents' and teachers' reports regarding children's behavioral problems and social skills considering these different interactional and essential environments for child development.
The objective of this study is to fill in this gap, describing and comparing the assessments of mothers and teachers regarding their educational social skills (HSE) and negative practices, along with the children's behavioral problems and social skills.

METHOD Ethical aspects
The Institutional Review Board at the hosting university approved this study, which is part of a larger project (Opinion Report No. 5826/46/01/10).

Sampling
A database collected between 2010 and 2012 was used in this study.It is composed of the reports of 112 teachers and 186 mothers/fathers/caregivers concerning 224 children.Of these, 118 children were eligible; however, those whose respondents were the fathers or caregivers (20 children) or whose mothers were foster mothers, widowed, or divorced (17 children) were excluded from the sample.That is, only the children of biological mothers from bi-parental families remained.Therefore, the final sample was composed of 81 biological mothers, married or in a civil partnership, and the children's teachers.

Participants
A total of 81 mother-child pairs participated in this study and were assigned to four subgroups: a non-clinical sample, composed of 32 children without behavioral problems; and three clinical samples: one subgroup with 23 children presenting behavioral problems at school and home, according to the reports of teachers and mothers; one subgroup with 18 children presenting problems exclusively at home, according to the reports of mothers; and the last subgroup was composed of eight children presenting behavioral problems exclusively at school, according to their teachers' reports.
The criteria used to assign the children to each clinical group was based on scores within the clinical or borderline ranges obtained in the Child Behavior Checklist (CBCL), completed by the mothers, or in the Teachers Report Form -TRF, completed by the teachers (Achenbach & Rescorla, 2001).Children with scores within the non-clinical range were assigned to the non-clinical subgroup.
Considering only the clinical subgroups, the number of children attending elementary education was equivalent (X 2 = 1.969, p = 0.374), as was the number of boys and girls (X 2 = 2.852, p = 0.240).No statistical differences were found between the clinical and non-clinical subgroups regarding their distribution into school grades (X 2 = 3.291, p = 0.349), though the distribution of boys and girls was not equivalent between the clinical and non-clinical subgroups (X 2 = 10.670,p = 0.014).The non-clinical subgroup was composed of 11 boys and 21 girls, while 33 boys (18 = both contexts; 10 = family; 5 = school) and 16 girls (5 = both contexts; 8 = family; 3 = schools) composed the clinical subgroups.
When analyzing each clinical group regarding school grade and sex variables, we found: (a) statistical differences (X 2 = 0.002; p = 0.964) between the distribution of boys (7 preschoolers and 11 school-aged) and girls (2 preschoolers and 3 school-aged) in kindergarten and elementary school; (b) the same occurred for the group with problems exclusively in the family context (X 2 = 0.000; p = 1.000), with 5 boys and 4 girls in kindergarten and 18 children (10 boys and 8 girls) in elementary school; (c) the number of children presenting problems exclusively at school was equivalent (X 2 = 0.036; p = 0.850), i.e., 3 preschool girls and 2 preschool boys, and two school-aged boys and one school-aged girl.The distribution of boys (5 preschool and 6 school-aged) and girls (15 preschool and 6 school-aged) was equivalent between kindergarten and elementary education (X 2 = 2.078; p = 0.149).The analysis showed that the school grade children were attending did not influence the comparisons between the non-clinical and sub-clinical subgroups.However, children's distribution between groups according to sex was not equivalent; boys presented problem behaviors more frequently.The distribution of boys and girls between kindergarten and elementary school was equivalent.Preschoolers were aged 4.3 years old on average (SD = 1.09), while school-aged children were 8.2 years old on average (SD = 1.84).
Mothers' ages ranged from 20 to 44 years old, 31 years old on average (SD = 5.69); all biological mothers, married or in a civil partnership.The number of children ranged from 1 to 5 (M = 2.33; SD = 1.10).Regarding maternal education, 32.1% reported middle education (complete or incomplete), 53% reported high school (complete or incomplete), and 13.6% reported higher education (complete or incomplete); 39.5% of the sample had a paid job, while 59.3% did not.Regarding family income, 8.6% reported up to one time the amount of the current minimum wage (MW); 29.6% reported 2 times the MW; 29.6%, 3 times the MW; 12.3% reported 4 times the MW; 11.1%, 5 times the MW; and 7.4% reported more than 6 times the MW.
Regarding the participant teachers, preschool teachers' average age was 43 (SD = 8.04), while elementary education teachers were 36 on average (SD = 9.48).The participants had ten years of experience on average (SD = 7.20), while 80% had a bachelor's degree; 6% of the sample was composed of male teachers.
Regarding the behavioral profile of the children in the clinical groups, we identified that: (a) children exhibiting behavior problems in both contexts presented a higher number of problems compared to the remaining two groups.According to the mothers' reports, (18 internalizing, 14 externalizing, and 18 total problems, comorbidities in 11 cases) and according to the teachers' reports (15 internalizing, 16 externalizing, and 16 total problems, comorbidities in 3 cases); (b) according to the mothers, children presenting problems exclusively in the family context totaled 15 internalizing problems, 6 externalizing, 9 total problems and 4 cases of comorbidities.The teachers of these children identified few problems (1 internalizing, 1 externalizing, and 1 total problem); (c) mothers identified few problems among children presenting problems exclusively at school (1 internalizing, 0 externalizing, and 1 total problem), while the teachers reported 4 internalizing problems, 8 externalizing, and 10 total problems, 3 with comorbidities.

Instruments
Child Behavior Checklist (CBCL) and Teachers Report Form -TRF (Achenbach & Rescorla, 2001) for preschool and school-aged children were used to assess behavior problems and assign children to clinical or non-clinical groups.Based on mothers' and teachers' reports, the instruments measured the frequency of answers indicating problem behaviors.The results are organized according to the scales of internalizing problems (e.g., anxiety and depression), externalizing problems (e.g., disobedience and opposition), and total problems.The instruments classify disorders according to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV); the instruments are undergoing updating for the DSM-V.All the problems are classified into clinical, non-clinical, or borderline.According to the recommended by Achenbach & Rescorla, (2001), in this study, borderline classifications were considered clinical.Bordin et al. (1995) found satisfactory test-positivity and morbidity criteria for the clinical and non-clinical profiles.
The Interview Script of Parental Social Educative Skills and Interview Script of Educational Practices for Teachers (RE-HSE-Parent, Bolsoni-Silva et al., 2016;RE-HSE-Teacher, Bolsoni-Silva et al., 2018) were used to assess parents-children and teachers-students interactions.Both the instruments are psychological tests approved by the Brazilian Federal Council of Psychology and assess, through an interview, how social educational skills (communication, affection, and limit-setting), negative practices, and contextual variables influence children/students' skillful or problem behaviors.The instruments present two factors: total positive and total negative of adult-child interactions and discriminate between children with and without behavioral problems (discriminant validity tests).
Socially Skillful Responses Questionnaires (QRSH-Parents, Bolsoni-Silva & Loureiro, 2020 and QRSH-Teachers, Bolsoni-Silva & Loureiro, 2020) were used.These instruments assess social skills based on the reports of parents/caregivers and teachers.The QRSH-Parent presents satisfactory evidence of construct and discriminant validity and alpha equal to 0.79.The QRSH-Teacher also presents satisfactory evidence of construct and discriminant validity, with alpha = 0.94.Both present cutoff points for social skills deficit, considering ROC curve analysis.

Data collection procedures
The Education Department from a city in the interior of São Paulo, Brazil, provided its consent before data collection was initiated.Early Childhood and Elementary Schools were contacted to verify their availability to participate in the study.Data were collected from 14 Municipal Early Childhood Schools and 12 Municipal Elementary Schools.Mothers were informed that this study addressed the interactions between parents and children and between teachers and children-those who consented signed free and informed consent forms.
The schools' principals and teaching coordinators, and later the teachers willing to collaborate, received clarification regarding the objectives of the study.Teachers who consented signed free and informed consent forms and nominated one child they believed had behavioral problems and one child they believed did not present behavioral problems.Mothers authorized teachers to complete the instruments concerning their children.
The data were collected among mothers in the places they considered to be more convenient (at their homes or school), while among the teachers, data were collected at school.Informative pamphlets addressing the topic were distributed, and upon the schools' request, lectures were given on behavioral problems, social skills, and educational practices to ensure ethical aspects.

Data process and analysis
Data were tabulated in the following order: (a) data were inserted in ASEBA and children's behaviors were coded according to TRF and CBCL in all the categories provided in the instruments; (b) participants were assigned to groups according to: having behavioral problems only at the family context (CBCL), only at school (TRF), in both contexts (CBCL and TRF), and no behavioral problems; (c) frequency of problems was described using descriptive statistical analysis, considering the groups with problems exclusively at school, home, and problems in both contexts; (d) data obtained from RE-HSE-Parent, RE-HSE-Teacher, QRSH-Parent, and QRSH-Teacher were tabulated according to each instrument's instructions; (e) mothers' and teachers' reports were compared (Wilcoxon Test) regarding the variables of interest in each of the clinical groups (both contexts, exclusively at home, and exclusively at school) and in the non-clinical group; (f) mothers' reports regarding the variables of interest in each clinical group were compared to the reports regarding the variables of interest in the non-clinical group (Mann Whitney); (g) teachers' reports regarding the variables of interest in each clinical group were compared to reports regarding the variables of interest in the non-clinical group (Mann Whitney).The level of significance was established at 5%.The results are presented in tables.

RESULTS
This section is organized into three tables.The first presents parents' and teachers' educational social skills, negative practices, and children's social skills and behaviors based on mothers' and teachers' reports.The remaining tables compare the clinical subgroups to the non-clinical group according to mothers' and teachers' reports.
Table 1 presents differences between parents and teachers for most of the behaviors assessed.Regarding the group of children with problems in both contexts: (a) mothers were more apt in total educational social skills; (b) mothers and teachers did not diverge regarding the varied subjects used in the communication with children, but diverged regarding affection (mothers were more affectionate than teachers) and limit-setting (teachers set limits more frequently); (c) the contextual variables indicate that mothers used HSE more frequently; (d) mothers adopted negative practices   more frequently, even though (e) teachers complained of behavioral problems more frequently than mothers; (f) social skills measured by RE-HSE-Parent/Teacher were more frequently reported by mothers, though responses to the QRSH-P/Parent showed no differences; (g) mothers scored higher in total positive, while no differences were found regarding total negative.Mothers of children exhibiting problems exclusively within the family context reported more skills (affection stood out) and also more negative practices.No differences were found between the mothers' and teachers' reports concerning behavioral problems, diverse communication subjects, HSE, and limit setting.This finding was more frequently found among mothers, which is in line with them being more skillful in the interactions with children.Mothers more frequently reported children's social skills measured by the non-structural instruments (RE-HSE-Parent/Teachers) and directive instruments (QRSH-Parent/ Teacher).Mothers also reported total positive skills more frequently, while total negative did not differ between teachers and mothers.

Comparisons between mothers' and teachers' reports regarding the clinical and non-clinical groups, considering the means of educational practices, and children's behavior problems and social skills.
Regarding children exhibiting problems exclusively at school, mothers reported more total educational social skills and contextual variables than teachers, though HSE concerning communication, affection, and limit setting did not differ between them.In spontaneous reports, mothers more frequently considered their children had social skills and total positive interactions.Negative practices, complaints of problems, and total negative did not differ between mothers and teachers.
Comparisons between mothers' and teachers' reports concerning the non-clinical group revealed that mothers were more skillful (HSE total and HSE-affection) than teachers in varied situations (context).Note that, differently from the clinical groups, teachers, more frequently than mothers, identified social skills in the non-clinical group when assessed by QRSH-Parent/Teacher.However, when interviewed (RE-HSE-Parent/Teacher), mothers reported more social skills, which resulted in higher scores in total positive.The mothers also more frequently complained of problems and negative practice and scored higher in total negative.
Table 2 presents a comparison between the reports of mothers of clinical children with the report of mothers of non-clinical children.
Table 2 shows that mothers of children with problems in both contexts more frequently reported behavioral problems, negative practices, and total negative interactions than mothers of non-clinical children.No differences were found regarding HSE and social skills.The comparison between mothers of children with problems exclusively within the family context and mothers of non-clinical children revealed that mothers in this clinical group (family) more frequently identified behavioral problems and adopted negative practices, resulting in more negative social interactions (total negative).Additionally, contrary to what was expected, these mothers reported to more frequently adopt hse-affection.Only complaints of behavioral problems differed between the group of children with problems at school and the non-clinical group; the clinical group reported behavioral problems more frequently.
Table 3 presents a comparison between the teachers' reports concerning clinical children and non-clinical children.Table 3 shows differences in teachers' reports regarding children with problems in both contexts and non-clinical children in almost all comparisons.Teachers reported more behavioral problems for this clinical group and more frequent use of negative practices.Regarding HSE, teachers mostly used limit setting with the clinical children, while they talked more frequently with non-clinical children regarding diverse subjects, considering children's interests, and in different contexts.Social skills were more frequently reported for non-clinical children (both in the interview with RE-HSE-Teacher and QRSH-Teacher).The comparison between teachers' reports regarding children with problems exclusively within the family context and non-clinical children revealed that complaints of problems, negative practices, and total negative were more frequently reported for the clinical group.At the same time, social skills were more frequently identified in the non-clinical group when QRSH-Teacher was used.Comparisons between teachers' reports regarding children exhibiting problems exclusively at school and the non-clinical group also revealed more problems, negative practices, and total negative.In contrast, children's social skills were more frequently reported for the non-clinical group using both the instruments.

DISCUSSION
The findings concerning the comparisons between the respondents based on CBCL and TRF show that mothers and teachers agreed that children with problems in both contexts present a greater impairment level, with a high frequency of internalizing, externalizing and total problems.The mothers also considered that these children more frequently presented comorbidities of internalizing and externalizing problems.However, the respondents' reports diverged concerning children exhibiting problems exclusively at home or school.These results are in line with the meta-analysis conducted by De Los Reyes et al. (2015) and the study by Llanes et al. (2020).Both report a low to moderate correlation between the parents' and teachers' reports.The authors explain that this divergence may be due to the use of different instruments and assessments conducted in different contexts; fathers and mothers in the same context more frequently agree than other pairs of informants.Family and school are distinct contexts, and each demands different repertories.However, it is worth noting that children with problems in both family and school contexts present, according to the respondents, similar behaviors in these contexts.Perhaps, the nature of the social interactions established in these two contexts explains this finding, a subject later discussed.
Thus far, literature comparing mothers' and teachers' reports concerning children has not achieved a consensus.Some studies found agreement between these two groups of respondents (Bernedo et al., 2014;Ercan et al., 2015), while others did not (Korsch & Petermann, 2014;Lavigne et al., 2015;Rescorla et al., 2014;Rudasill et al., 2014), which are consonant with the findings in this study.That is, part of the comparisons differentiated mothers from teachers, while others did not.However, none of the studies compared children exhibiting problems in both contexts with children exhibiting problems exclusively at home or school, as we did here.These comparisons are relevant to explain why the reports are consonant regarding some   aspects but disagree regarding others.Santiago et al. (2016) verified that the most important predictor for disagreement in parents' and teachers' reports was conflicts between teachers and children.Conflicts were strongly associated with the identification of behavioral problems.The findings in this study align with Santiago et al. (2016) because teachers more frequently report behavioral complaints and negative practices.Rohenkohl and Castro (2012) also found that mothers reported internalizing problems more frequently than teachers.

Categories
Mothers addressed in this study presented adequate resources concerning total HSE and HSE-affection, skills that admittedly prevent behavioral problems (Bolsoni-Silva & Loureiro, 2019;García-Linares et al., 2014;Nunes et al., 2013;Rohenkohl & Castro, 2012;Vafaeenejad et al., 2019;Zalewski et al., 2017).As previous studies show ( García-Linares et al., 2014;Price et al., 2013;Vafaeenejad et al., 2019;Zalewski et al., 2017), the widespread use of negative practices is likely to be strongly associated with children's behavioral problems.Heatly and Votruba-Drzal (2017) verified that the interactions between parents and children influenced school engagement among 1 st grade students and the occurrence of conflicts with teachers.In this sense, the fact that mothers reported the frequent use of negative practice to deal with their children is of concern, even if they also reported the use of social educational skills.
Internalizing problems are more frequent among children exhibiting problems exclusively at home, a result reported by studies that also find a higher frequency of negative practices (Tandon et al., 2009).Silveira and Wagner (2009) noted that parents and teachers tend to adopt both positive and negative practices when dealing with children's behavioral problems, though mothers more frequently used negative practices and social skills, such as affection.It is worth noting that teachers more frequently reported HSE-limit setting and diverse communication when dealing with the clinical group exhibiting problems in school and family contexts when compared to the non-clinical group, as already reported by Bolsoni-Silva et al. (2018).
Thus, the comparisons between the reports of mothers and teachers regarding educational practices and children's behaviors showed that: (a) mothers are more skillful than teachers, including contextual variables (diverse positive interactions), general HSE and HSE-affection when dealing with children with behavioral problems in both contexts, problems exclusively at home, and with the non-clinical group.Mothers of children with problems exclusively at school also scored higher in terms of general-HSE and contextual variables; (b) mothers and teachers did not differ regarding HSE-communication and HSE-limit setting; c) teachers of children with problems in both contexts more frequently reported HSE-limit setting; (d) mothers more frequently reported negative practices to deal with children exhibiting problems in both contexts, exclusively at home, and with children in the non-clinical group.However, the report of negative practices did not differ between teachers and mothers dealing with children exhibiting problems exclusively at school; (e) mothers of children in the non-clinical group more frequently complained of behavioral problems, while the teachers reported more behavioral problems among children with problems in both contexts.No differences were found regarding this aspect in the remaining groups; (f) through spontaneous (RE-HSE-P) reports, mothers mentioned more social skills than teachers in both clinical and non-clinical children.However, when responding to the structured instruments (QRSH-Parents/Teachers), mothers reported more social skills for children with problems exclusively at home, while teachers reported more social skills for children in the non-clinical group.
In summary, mothers and teachers reported positive and negative practices when interacting with the children, though teachers less frequently reported negative practices.This finding agrees with Silveira and Wagner (2009), which also verified that parents and teachers equally adopted inductive practices, which are related to HSE-limit setting addressed in this study.Note that the findings in this study corroborates part of the results reported by Silveira and Wagner (2009); that is, the teachers reported more frequently using these skills when exclusively dealing with children exhibiting problems in both contexts.Garcia et al. (2016) also verified that teachers used the same practices adopted by parents, though less frequently.This finding seems to be confirmed in this study, as both mothers and teachers used HSE and negative practices, though mothers more frequently than teachers.
In general, mothers identified more social skills, which seem to be in line with Korsch and Petermann (2014), who report that parents mentioned more pro-social behaviors than teachers.On the other hand, when responding to a structured instrument, teachers reported more social skills among children in the non-clinical group than mothers.Bolsoni-Silva and Loureiro (2016) verified that teachers identified a higher number of social skills than mothers when comparing clinical to non-clinical groups, which corroborate with the findings in this study.Social skill repertories are inversely proportional to behavioral problems, whether parents' (Fernandes et al., 2018;Kim et al., 2011;Reynolds et al., 2010), or teachers' reports are considered (Kettler et al., 2011;Pizato et al., 2014).Hence, a smaller social skills repertoire is expected when many behavioral problems are reported, in line with this study.The results of this study lead to the hypothesis that the use of HSE and negative practices depended on whether children exhibited behavioral problems or skillful behaviors (Assis-Fernandes & Bolsoni-Silva, 2020).
In this direction, mothers' HSE-affection, teachers' HSE-limit setting and HSE-diverse communication are good indicators of resources, considering that parentchildren affection (Bolsoni-Silva et al., 2016;Borden et al., 2014) and teacher-student closeness (Bolsoni-Silva & Mariano, 2018;O'Connor et al., 2011;Ribeiro, 2010), along with limit setting (Bolsoni-Silva & Loureiro, 2019), and positive practices (Bolsoni-Silva & Mariano, 2018), decrease the likelihood of children exhibiting externalizing behaviors.However, when dealing with children with behavioral problems, the quality of diverse communication, including subjects that interest children and academic subjects (Bolsoni-Silva et al., 2018;Mariano & Bolsoni-Silva, 2018), the less frequent use of affection on the part of teachers, and excessive use of negative practices (Santiago et al., 2016), may favor behavioral problems at home and school.
The use of two instruments to assess children's social skills was propitious.Both an interview/spontaneous report and a structured instrument rated on a Likert scale enabled identifying differences between teachers' and mothers' reports, as previously found by Bolsoni-Silva e Loureiro (2019).Spontaneous reports suggest greater discrimination and a greater likelihood of adults identifying and agreeing regarding children's behaviors, which occurred more frequently among mothers.The scale, in turn, favored the identification of a broader range of behaviors.Thus, the two instruments complemented each other.
The findings of this study show that children's skillful or problem behaviors are influenced by both the family and school contexts, and therefore, interventions need to be implemented in each of these contexts.

FINAL CONSIDERATIONS
This study presents similarities and differences between mothers' and teachers' reports regarding their practices and children's behaviors.Comparisons between the respondents showed that mothers more frequently reported general skills, especially affection, while teachers more frequently than mothers reported limit-setting skills in the group of children with problems in both contexts.In turn, mothers, more frequently than teachers, reported negative practices; mothers of children in the non-clinical more frequently complained of behavioral problems, while teachers complained of behavioral problems in the group of children exhibiting problems in both contexts; and mothers identified more social skills among children presenting problems exclusively at home, while teachers identified more social skills among the non-clinical children.This study's strengths include a comparison between groups considering the occurrence of children's behavioral problems at home and school, or only in one of these contexts.Additionally, only biological mothers from biparental families were included in the sample, which was essential for controlling confounding variables regarding educational practices.Finally, data regarding the different respondents' behaviors and children's behaviors improved understanding of the aspects investigated.The limitations refer to a small sample and different profiles of children exhibiting problems exclusively at school or home.These limitations can be addressed in future studies by including fathers as the informants and adopting observational methods in addition to reports.

Table 2
Comparisons between mothers' reports concerning clinical and non-clinical children, considering the means of educational practices, and children's behavior problems and social skills.