The Role of Family in Bullying and Cyberbullying Involvement : Examining a New Typology of Parental Education Management Based on Adolescents ’ View of Their Parents

The influence of the family in children’s involvement in bullying and cyberbullying has been well documented. However, previous research into this relationship seems to have overlooked recent social changes, which have affected the family context. The aim of this study is to put forward a categorization of the current educational management of Spanish parents and examine how this is linked to their children’s involvement in bullying and cyberbullying. To achieve this, 2060 schoolchildren from the South of Spain (47.9% girls with mean age = 14.34) answered four questionnaires including the Scale for the Assessment of the Parenting Styles of Adolescents’ Mothers and Fathers, the Discipline Dimensions Inventory, the European Bullying Intervention Project Questionnaire, and the European Cyberbullying Intervention Project Questionnaire. The Cluster Analysis results revealed a typology containing six styles: permissive, authoritarian, strict, normative democratic, indulgent democratic, and punitive democratic. Lower levels of victimization and aggression in bullying and cyberbullying were found to be linked to the indulgent democratic or normative democratic styles and higher levels to the authoritarian and strict styles. The value of parents’ educational practices and how they are combined in general styles, since these are elements that can predispose or prevent adolescent’s involvement in bullying and cyberbullying, is discussed.


Introduction
Despite the fact that a considerable body of scientific evidence has accumulated on the nature and prevalence of bullying and cyberbullying, there are still many gaps in our knowledge.For this reason, further studies are needed into ways to prevent both of these phenomena.
Bullying is a group phenomenon, occurring within more or less complex networks of relationships, in which a single pupil or a group of schoolchildren pick on another intentionally and without provocation.The aggression often takes place regularly over time and the victim finds it extremely difficult to defend themselves against their aggressor or aggressors, given the imbalance of power that exists between them (Smith 2016;Olweus 1999).The aggression is generally made worse by two processes, which form a part of the group's micro-culture.First, the law of silence, which prevents those involved and specially bystanders from reporting the unfair aggression, makes it difficult that it is reported to adults and, hence, stopped.The submissive domain scheme also transforms the usual horizontality that characterizes relations between equals into a kind of perverse verticality.
The aggression involved in bullying can be verbal, physical, sexual, psychological, or relational (Ortega and Mora-Merchán 2008).
Currently, the advent of information and communication technology has opened up new opportunities for social aggression, which gives rise to a parallel phenomenon closely connected to bullying known as cyberbullying.Although there is a degree of controversy over the exact definition of cyberbullying, initial approaches have defined it as repeated, aggressive behavior conducted through electronic devices with the intention of hurting or intimidating the recipient (Hinduja and Patchin 2008;Tokunaga 2010).Cyberbullying is characterized by sending aggressive messages or humiliating images, making intimidating telephone calls, impersonating the victim's identity, or recording and then sharing videos in which the victim is ridiculed or attacked.Since it was demonstrated that there is a considerable overlap between both violent phenomena, a number of authors have agreed to label cyberbullying as a sub-category of bullying (Waasdorp and Bradshaw 2015).We will examine both of these problems of interpersonal violence in this work.
Interpersonal violence between peers exists as a phenomenon in all schools (Craig et al. 2009).In fact, Modecki et al. (2014), in their meta-analysis, which included 80 studies from all around the world, found that cyberbullying is less frequent than conventional bullying, with a prevalence of 15% when compared to 35% for bullying.
The need to prevent these phenomena by addressing all the relevant factors of both protection and risk is, therefore, becoming more and more urgent.Most research has centered on individual aspects in an attempt to delimit the factors associated with these forms of interpersonal violence, which include social competence, emotional intelligence, empathy, judgment, and moral sensitivity.The relevant research shows that schoolchildren involved in bullying have lower levels of socio-moral competence than those who have not been affected (Del Rey et al. 2016;Romera et al. 2018;Gómez-Ortiz et al. 2017;Elipe et al. 2015).
In addition to this, research has also focused on studying the connection between the family context and involvement in bullying or cyberbullying, and this idea has been linked to the parenting styles of the parents.The research seems to point out where the problem may be originating from, but it is still rather imprecise (Nocentini et al. 2018).
Traditionally, most research on parental styles followed the theoretical model originally proposed by Baumrind (1968) and concluded by Maccoby and Martin (1983).The empirical and practical validity of this model has been widely accepted, whereby four categories of parenting were established, based on the dimensions of involvement and control or coercion.The categories include: (a) democratic (high levels of both factors), (b) permissive (high involvement and low coercion), (c) authoritarian (low involvement and high coercion), and (d) negligent (low levels of both factors).Other authors have also accepted this categorical approach, and other typologies have been proposed to classify the behavior and attitudes of parents toward their children and to study the children's adjustment such as that proposed by Musitu and García (2005).
According to this typological approach, the democratic and the indulgent styles are those that seem to best prevent children from getting involved in bullying and cyberbullying (Baldry and Farrington 2005;Gómez-Ortiz et al. 2015;Martínez et al. 2019).On the other hand, the authoritarian style, especially when accompanied by the use of physical punishment and excessive parental control, is closely linked to the likelihood of becoming a bully or victim in these two areas (Dehue et al. 2012;Garaigordobil and Machimbarrena 2017;Low and Espelage 2013;Ybarra and Mitchell 2004).The negligent style is also a risk factor linked to children showing aggressive behavior towards their peers (Cerezo et al. 2018).
Nevertheless, the 'parenting style' construct has recently acquired new dimensions and has been described with greater precision, which enables us to attempt to outline the possible contribution of parents and, in general, of the educational climate in the family to the risk of getting involved in bullying and cyberbullying.The perspective in which the educational style is analyzed is known as a dimensional perspective.It involves studying various facets of the parent-child relationship, which, as well as affection and communication, include a sense of humor, mutual enjoyment between parents and children (Oliva et al. 2007), and the encouragement of autonomy.In other words, this gives children the opportunity to make their own decisions or choose certain behavior (Lewis 1981).Moreover, from this dimensional perspective, the factor of control is examined in a detailed way and the attitudinal and behavioral strategies that are linked to this dimension are diversified.A distinction is, therefore, made between behavioral control, characterized by establishing rules and the supervision of young people's behavior (Gray and Steinberg 1999), and psychological control, meaning the use of manipulative and intrusive strategies, such as generating guilt or the withdrawal of affection when children display undesirable behavior (Silk et al. 2003).In addition, filial disclosure, which refers to the information that young people voluntarily share with their parents about aspects of their private or school life, enables the parents to know more about their child and monitor them (Kerr et al. 1999).
The dimensional approach has also shown that both the use of psychological control and the lack of supervision by parents are risk factors that increase the likelihood of children bullying their peers or being victimized by them (Gómez-Ortiz et al. 2016;Samper-García et al. 2015;Holt et al. 2008;Hong et al. 2017).As for disclosure, it has been found that children who voluntarily confide in their parents and enjoy good communication with them, as well as those that grow up in a positive, warm, loving environment in which the child's autonomy is encouraged, are less likely to get involved in bullying and cyberbullying (Gómez-Ortiz et al. 2014;Larrañaga et al. 2016;Murphy et al. 2017;Offrey and Rinaldi 2017;Stavrinides et al. 2015;Fousiani et al. 2016;Low and Espelage 2013).
The steps that parents take to manage their children's discipline at home constitute a set of parental educational practices closely linked to parenting style.Strauss and Fauchier (2007) identified four procedures related to the exercise of parental discipline.This includes: (a) physical and psychological punishment, including the use of force and even physical or psychological aggression in order to correct a child's behavior, (b) response cost, which involves the withdrawal of privileges and the imposition of other compensatory behavior, (c) inductive discipline, consisting of giving explanations about the right way of doing things, reinforcing the child's correct behavior, and trying to divert their attention from unacceptable behavior, and, lastly, (d) direct supervision, including strategies aimed at checking whether the child is behaving correctly or not and ignoring negative behavior with the aim of suppressing it.
There has also been some research into the link between managing discipline and children becoming involved in bullying even though this has not gone into the subject in much depth.Most studies have focused on the link between punitive discipline and bullying, and have found that parents who exercise punitive discipline at home make it more likely that their children will also become bullies or victims of bullying (Hong et al. 2017;Lansford et al. 2014;Gómez-Ortiz et al. 2016;Lereya et al. 2013).Other risk factors also include disciplinary methods, which involve psychological aggression (Zottis et al. 2014).The relationship between cyberbullying involvement and parental discipline management has been unexplored.
A review of the literature shows that current research is attempting to establish the relationship between parenting practices and their exercise of corrective discipline as well as children's involvement in bullying and cyberbullying.However, the results produced by most of these studies are extremely segmented, focusing on one particular educational practice at the expense of others.This is especially true in the case of discipline, where there are significant gaps in our knowledge regarding everything related to non-punitive discipline, which is still very much uncharted territory, not only in relation to bullying and cyberbullying, but also in the area of competition and the psychosocial adjustment of schoolchildren in general.The original idea of this research was, therefore, to propose categories of parenting styles, which include not only practices linked to the dimension of warm or control, but also others, which have a strong influence over children's psycho-social adjustment, such as the encouragement of autonomy or the parents' sense of humor (Gómez-Ortiz et al. 2015) as well as different disciplinary methods.To the best of our knowledge, no classification of parental educational management styles exists that contains these characteristics and includes all these variables, as well as explores both maternal and paternal behavior.Some authors, however, do seem to have reflected this need, and have proposed alternative classifications to the classic typology devised by Maccoby and Martin (1983), notably Torío et al. (2009), Oliva et al. (2008), andGómez-Ortiz et al. (2015).In fact, even Baumrind (1991), some years after her initial proposal, proposed an alternative typology, which included special features, which typify parent-child relationships in adolescence.In any case, almost all these classifications focus on studying the educational style and do not specifically address discipline management.In addition, they only analyze the individual behavior of each parent rather than the interaction of both, when, in real life, parental coherence has been shown to be an important element in influencing a child's adjustment (Gómez-Ortiz et al. 2015;Oliva et al. 2008).Furthermore, apart from the classification by Gómez-Ortiz et al. (2015), the other classifications differ very little from the classic typologies proposed over 40 years ago.For this reason, we need to reflect the fact that family life has moved on significantly since that time and that the different features of education and parental discipline management may well have expanded and diversified.
The main aim of this research was, therefore, to discover the behavioral patterns displayed by parents while bringing up their children and how this affects the child's adjustment.We assessed this adjustment through their involvement in bullying and cyberbullying, two violent phenomena, which arouse considerable concern in society, given their common occurrence and severe repercussions.
This general objective can be broken down into two specific objectives: 1.
To explore the educational profile of the parents of adolescents, analyzing specific parenting practices they use with their children, evaluated from the children's point of view.2.
To distinguish any possible differences in the degree of involvement in bullying and cyberbullying linked to the parents' educational profile.
The approach and implementation of our research was based on two hypotheses: 1.
We expected to find diverse profiles of parental education, in which the factor of democratic mothers and fathers is attenuated by different styles of discipline management, communication, sense of humor, and other relational patterns (Gómez-Ortiz et al. 2015).

2.
We expected to find that the parents with documented higher levels of affection, sense of humor, promotion of autonomy, and in whom their children spontaneously confide their feelings and ideas (disclosure) and who avoid punitive discipline would show a lower involvement of their children in bullying and cyberbullying (Fousiani et al. 2016;Gómez-Ortiz et al. 2014;Zottis et al. 2014).

Participants
The sample consisted of 2060 students from the Compulsory Secondary Education (ESO, in Spanish) whose ages ranged between 12 and 19 (x ⊥ ¯= 14.34; o = 1.34), of which 52.1% were boys and 47.9% girls.Furthermore, 28.4% were in their first year of secondary school, 28.4% in the second year, 22.1% in their third year, and 21.1% in the fourth year.The students were chosen from public and private schools from different areas around Andalusia.
The sample was considered representative of the population of schoolchildren enrolled in Compulsory Secondary Education (ESO) in Andalusia, and was selected by randomized, single-stage, stratified, random sampling using conglomerates and with proportional affixation.The strata were identified as the geographical area of Andalusia (Eastern or Western Andalusia), the type of school (public or private), and the population of the town/city (under 10,000 inhabitants, between 10,001 and 100,000 inhabitants, and more than 100,000 inhabitants).All the categories of the strata are indexes, which are relevant in Spain.
This study was carried out in accordance with the Helsinki declaration and, before collecting the data, the parents provided informed consent for their sons and daughters to take part in the study.The Ethics Committee of the University of Córdoba authorized the study.

Instruments
The European Bullying Intervention Project Questionnaire (EBIP-Q) was used to evaluate the level of involvement in bullying (Ortega-Ruiz et al. 2016).This self-report is made up of 14 Likert-type items with five response options ('No,' 'Yes, once or twice,' 'Yes, once or twice a month,' 'Yes, about once a week,' and 'Yes, more than once a week'), of which 7 items are related to bullying and the other 7 to victimization.The questionnaire covers aspects of verbal, physical, relational, direct, and indirect harassment.The time interval used to evaluate the frequency was the academic year the pupils were in at the time of answering the questionnaires.The reliability of this scale is adequate (α victimization = 0.80, α bullying = 0.76, total α = 0.83).
Cyberbullying was evaluated using the European Cyberbullying Intervention Project Questionnaire (ECIPQ) (Del Rey et al. 2015), which is made up of 22 Likert-type items, including five response options ('No', 'Yes, once or twice,' 'Yes, once or twice a month,' 'Yes, about once a week,' and 'Yes, more than once a week').This self-report contains two dimensions: cyber-victimization and cyberbullying, with a suitable Cronbach's Alpha score (α cyber-victimization = 0.76, α cyberbullying = 0.68, α total = 0.81).The time interval used to evaluate the frequency was the academic year the pupils were in at the time of answering the questionnaires.
Lastly, discipline was measured using the Inventory of Disciplinary Dimensions (IDD) designed by Strauss and Fauchier (2007) and validated in a Spanish adolescent population by Calvete et al. (2010).This scale consists of a total of 52 items, of which 26 refer to the mother's discipline methods and the other 26 to the father's.The scale has a Likert-type response format with 10 response options ranging from 0 (never) to 9 (twice or more times a day).This instrument measures four factors referring to various disciplinary procedures: Physical and psychological punishment (PPP: psychological aggression and physical punishment), Response Cost (RC: withdrawal of privileges, compensation), Inductive Discipline (ID: distraction, explanation and reward), and Supervision (SUP: ignoring and control).The internal consistency was acceptable for higher order factors: αIDmother = 0.73, αIDfather = 0.74, αSUPmaterial = 0.74, αSUPfather = 0.69, αPPPmother = 0.87, αPPPfather = 0.88, αRCmother = 0.82, αRCfather = 0.82, and total α = 96.

Procedure
After permission was obtained from the parents and school management, we went out to the schools to conduct the survey.The children had previously been informed that the surveys were anonymous and confidential and that they could take part on a voluntary basis.Students filled the questionnaire on the paper.The questionnaires took 45 min maximum to complete.

Data Analysis
There was observed randomness in missing data including a MCAR (missing completely at random) pattern (Little's MCAR test: 6105.86 (6026); p = 0.23).Therefore, we decided to perform the analysis without such data (Bennett 2001).As such, the data included 1096 students who answered all the questions of the survey.N was specified in all analyses.
First, a hierarchical conglomerate analysis was performed on the variables connected with the mother's and father's educational styles, in order to establish the right number of clusters.To achieve this, both the dendogram scores and the distance between clusters observed in the clustering history were taken into account.In the latter case, the cluster was defined by the gaps between the successive values of the agglomeration coefficient, on the basis that adding the next case to the conglomerate would lead to a loss of its homogeneity.Next, a K-media cluster was performed to identify the parents' possible educational profile based on their children's perception of their educational practices and their use of disciplinary procedures.
The factors were then subjected to an analysis of variance (ANOVA) to evaluate the link between the parents' educational profile and the child's involvement in situations of aggression and victimization with their peers either on the Internet (cyberbullying) or in real life (conventional bullying).In this way, the groups defined by the conglomerates were considered as factors and the variables measuring the extent of the (cyber) aggression or (cyber) victimization were taken as dependent variables.Lastly, a Games-Howell post hoc test was carried out to check the difference in means between the groups when each was compared with the other five.The significant differences obtained with the ANOVA were contrasted with the results produced from measuring the size of the effect (Cohen's D).A value below or equal to 0.20 in this index would indicate a minimal effect, while values of more than 0.20 and up to 0.50 would indicate a moderate effect and over 0.80 would indicate a strong effect (Cohen 1992).All the data coding and analyses were carried out using the SPSS statistical package (version 19.0).

Classification of Parental Educational Management
The results of the hierarchical cluster analysis revealed that the solutions from clusters 5, 6, 7, and 8 would be the most suitable.After performing the K-means cluster analysis, it was decided to create six groups, as this appeared to be the most coherent solution and met the objective of providing a classification, which reflected the wide possible variability of the construct, while, at the same time, allowed maximum intra-group homogeneity and inter-group heterogeneity.The 5-cluster solution proved less informative since it reduced the diversification of parental profiles, and solutions with 7 and 8 conglomerates produced groups, which were overly similar to each other, and, therefore, showed little inter-group heterogeneity.
The schoolchildren from group 1 revealed a moderate level of behavioral control by their parents, and this group obtained the lowest means for that dimension.The scores for psychological control were slightly lower, while those for disclosure were about the same.On the other hand, higher scores were produced for items related to affection and communication, as well as the promotion of autonomy and parental sense of humor.Similarly, these students revealed that the use of all types of discipline by their parents was infrequent (this group, in particular, reflected the lowest use of inductive discipline by their parents).It was agreed to label the parents of these pupils as permissive.Group 2 was the group that produced the lowest scores for the dimensions of affection and communication, promotion of autonomy, sense of humor, and disclosure.However, they showed a high level of behavioral, and, above all, psychological control.In addition, the use of all types of discipline by both parents was frequent, and this group obtained the highest scores for physical and psychological punishment.These parents were labeled as authoritarians.The pupils in group 3 had high scores for all the dimensions related to the educational style, except for psychological control, where the scores were moderate.Regarding disciplinary strategies, they showed a frequent use of inductive discipline, while other disciplinary procedures were used much less frequently.Given their frequent use of behavioral control, it was agreed that the most suitable label for this group was normative democratic parents.The boys and girls in group 4 described their parents in a very similar way to the children in group 2, while they revealed a less frequent use of all the disciplinary strategies and slightly higher levels of affection and promotion of autonomy.However, these levels were still low compared with the other groups.These parents were labeled as strict.The parents of the schoolchildren in group 5 produced the highest scores for affection and communication, promotion of autonomy, and sense of humor compared with the other groups.They also revealed the highest levels of disclosure.The scores for behavioral control were also high, but not the highest.The parents of the children of this group stood out for their low scores in coercion.They also obtained the lowest scores for psychological control and for the use of all types of discipline, with the exception of inductive discipline, for which group 1 had the lowest scores.These parents were labeled as indulgent democratic.Lastly, the pupils from group 6 described their parents as affectionate and open to communication, and stressed their tendency to promote their autonomy and have a good sense of humor.However, they also pointed out that these parents commonly used behavioral and psychological control and used the discipline techniques evaluated in this scenario very frequently (this group was the one that scored highest for behavioral and psychological control, inductive discipline, response cost, and supervision).This latter group was labeled punitive democratic parents (see Tables 1 and 2).The answers of the items relatives to affect and communication, behavioral control, psychological control, promotion of autonomy, humor and disclosure ranged from 1 to 6. Therefore, means between 1-2 are considered low, between 3 and 4, moderate (those higher than 4 moderate-high and those lower than 4 moderate-low), and between 5 and 6, high.The answers of the discipline items ranged from 0 to 9, with the maximum mean equal to 6. Therefore, the means lower than 2 indicated an infrequent use, between 2 and 4, a moderate use, and higher than 4, a frequent use.TLM = the lowest mean in this dimension of all groups.THM = the highest mean in this dimension of all groups.

Parental Educational Profile and Involvement in Bullying and Cyberbullying
When analyzing the relationship between the parental educational profile and the aggression and victimization in bullying and cyberbullying, the results of the ANOVA test (see Table 3) showed statistically significant differences between the groups.
The Games-Howell post-hoc test (see Table 2) showed that, in the dimension of victimization in bullying, it was the youngsters with authoritarian parents who obtained the highest scores, which differed significantly, with a moderate effect size, from those whose parents from the indulgent democratic style, which reflected the lowest scores.The latter group also differed significantly in their levels of victimization among peers from the group of schoolchildren with parents from the strict and punitive democratic categories, with a moderate, low effect size.On the other hand, the group of schoolchildren with strict parents also stood out for their levels of victimization, and came second after the children of authoritarian parents.They also differed, therefore, from the children of permissive and normative democratic parents, with a moderate and low effect, respectively.Similarly, the children of authoritarian and strict parents also scored higher in the dimension of aggression in bullying.These two groups differed significantly from those with parents in the normative democratic (high and moderate effect size, respectively), indulgent democratic (high effect size in both cases), and permissive categories (moderate effect size in both comparisons).Differences were also found between the groups of children with parents in the democratic category, with the highest scores in this dimension given by the children of punitive democratics, who differed from the children of the normative democratic and indulgent democratic parents (low and moderate effect size, respectively).It was the latter who obtained the lowest scores in this dimension, with moderate differences between them and the children of permissive parents and low differences with the children whose parents are in the normative democratic category.
Regarding victimization in cyberbullying, differences were revealed between the group of schoolchildren with strict parents (second highest scores, after the group with parents in the authoritarian category) and those with permissive, indulgent democratic and normative democratic parents.The effect size of these differences was moderate in the former two cases and low in the latter.
As for cyberbullying aggression, the groups that reflected the highest scores were the pupils with parents in the authoritarian and strict categories.This latter group, in fact, differed significantly from those with parents classified as permissive or normative democratic (low and moderate effect size, respectively).In contrast, the lowest scores for this dimension were found in the group of children with indulgent democratic parents.This group differed from the group of schoolchildren with parents in the permissive, democratic, normative, strict, democratic, punitive, and authoritarian categories (low size effect for the differences in the first two cases, moderate in the third and fourth cases, and high in the last).

Discussion
The main aim of this research was to examine the relationship between the parents' educational profile and likelihood of their children becoming involved in the phenomena of bullying and cyberbullying, as well as to reassess the profiles of parents' educational styles to include the way parents use and handle discipline strategies as part of this relationship.
With regard to parenting styles, in agreement with the information in the first hypothesis, including the use and management of parental discipline techniques, levels of affection and communication, behavioral and psychological control, promotion of autonomy, children disclosure, and sense of humor shown by both parents, the overall picture of parental educational styles has become more diverse and it is clear that a more accurate description of parents' educational practices is required to reflect this wider variability.The results obtained produced a typology, which distinguishes six styles of parental educational management, which have been labeled: (1) permissive, (2) authoritarian, (3) normative democratic, (4) strict, (5) indulgent democratic, and (6) punitive democratic.The three styles we have termed 'democratic' were consistent with the traditional description of this style, and obtained high scores for the dimensions of affection and communication, behavioral control, promotion of autonomy, sense of humor, and disclosure (Gómez-Ortiz et al. 2015;Maccoby and Martin 1983;Musitu and García 2005;Oliva et al. 2008).However, it became evident that there were major differences between them regarding the use of discipline and psychological control.Thus, while indulgent democratic parents used psychological control and most disciplinary methods less frequently (except inductive discipline, where permissive parents obtained the lowest scores), the punitive democratic profile, as its name suggests, stood out for obtaining the highest scores in the use of most disciplinary and psychological control strategies.Normative democratic parents, on the other hand, fell halfway between these two styles, showing a moderate use of psychological control and non-punitive discipline techniques, while resorting fairly often to punitive measures, although to a lesser extent than the other groups (except indulgent democratic and permissive parents).Their main persuasive tool, therefore, was behavioral control, which entailed obtaining information by which they could control their children and establish rules (hence, the term 'normative').In another dimension, there are two parental prototypes, which seem to tie in consistently with the characteristics of the fathers and mothers described in the classic typologies (Maccoby and Martin 1983;Baumrind 1968).The permissive parents group scored lower for affection and communication, promotion of autonomy, sense of humor, and disclosure than the democratic parents, and they differed considerably from the latter, especially in their infrequent use of behavioral control, and, in general, of all methods of coercion and discipline.The authoritarian parents group obtained the lowest scores for the dimensions linked to personal involvement (affection and communication, sense of humor) and for promoting filial autonomy, while they scored the highest for physical and psychological punishment.These parents also reflected high levels of psychological control.In other words, they were described by their children as punitive, coercive, and with a personal involvement, which fell short of their expectations, as described in the classification by Baumrind (1968) and Maccoby and Martin (1983).Lastly, the parents we termed 'strict' were, in fact, a moderate version of the previous style, although their profile included slightly higher (although not excessive) levels of involvement (affection, communication, and sense of humor).The profile also included a more moderate use of harsh disciplinary procedures such as physical and psychological punishment, which is why the label is given (previously used in the study by Oliva et al. (2008).In all cases, strict parents and authoritarian parents also shared the fact that they were the ones who least encouraged, and, therefore, least obtained, disclosure from their children.This fact may have its own particular relevance and logic.Since these types of parents put emphasis on controlling their children behaviorally and psychologically, one would expect that they do not gain the mutual trust that children need to take part in processes such as disclosure, telling their parents about their daily affairs or feelings, and so on.Disclosure, as well as a feeling of security in the affection, seems to require warm, receptive communicative styles, which are not often evident in these types of parents (Kerr et al. 1999).
The second main aim of this research was to explore the possible differences in the degree of involvement in bullying and cyberbullying of adolescents depending on their parents' educational profile.The results obtained seem to confirm clearly that children with authoritarian and strict parents are those who tend to admit that they are involved in the phenomena of bullying and cyberbullying, both in situations of aggression and victimization.Similarly, the group of schoolchildren with punitive democratic parents is also linked to high levels of involvement in both phenomena, although to a rather lesser degree than the previous groups.At the opposite extreme, schoolchildren who describe their parents as permissive, normative democratic, or indulgent democratic are the ones who showed less involvement in bullying.In particular, the group with the lowest levels of involvement in the dynamics of bullying and cyberbullying in any of its manifestations or behavior is that of the schoolchildren with parents with an indulgent democratic profile.
These results confirm our second hypothesis, and appear to imply that adolescents who have a negative perception of their family environment recognize implicitly that they have less ability and resources to face up to common social problems, of which bullying and cyberbullying are among the most frequent.In particular, the greatest risk seems to be found in the combination of parenting, which overuses coercion and offers little involvement with their children.Both the authoritarian and strict parenting styles lead to rigid, restrictive parent-child dynamics in terms of the freedom and autonomy needed by the adolescent, combined with attitudes that offer very little emotional warmth.Likewise, these profiles tend to make an excessive use of punishments and psychological and behavioral restrictions that do nothing to stimulate the child's sense of security and maturity: on the contrary, they make them more socially vulnerable (Baumrind 1968;Maccoby and Martin 1983;Oliva et al. 2008).All this, together with social conditions that directly affect them (interpersonal peer relationships, managing the balance of power in the group, teachers who supervise children's social life and peaceful coexistence, etc.) may affect the adolescent's ability to face up to these conflicts and increase the tendency to become victims or cyber-victims as well as bullies or cyberbullies of their peers (Dehue et al. 2012;Garaigordobil and Machimbarrena 2017;Low and Espelage 2013).
On the other hand, when a family atmosphere breeds affection and a warm, fluid communication where the child can talk freely to their parents about their feelings, mentioning both their successes and their failures or problems, a greater stimulus is given to security and social competence.This enables the child to acquire the behavioral and attitudinal resources, which may reduce the possibility of becoming victims of bullying and cyberbullying or of using aggression and harassment against their peers (Larrañaga et al. 2016;Martínez et al. 2019;Fousiani et al. 2016;Gómez-Ortiz et al. 2014).As our results show, this pattern of close involvement is especially effective when accompanied by a positive, sensible management of discipline issues, in which induction and reasoning are preferred as a way of guiding the child's behavior.Punishments that use humiliation and manipulation to re-conduct the child's behaviour are avoided, as reflected in the profiles of indulgent democratic and normative democratic parents.This fact highlights how useful it is to have a parental typology, which looks in detail at the patterns of parent-child interaction and not just at general notions of involvement or coercion.It is clear that the classic generalized profile of the 'democratic' parent does not convey sufficient information: our study has shown different levels of involvement in bullying and cyberbullying depending on the sub-category analyzed.In any case, the impact of unsuitable discipline management on the child's involvement in these violent phenomena (Gómez-Ortiz et al. 2016;Zottis et al. 2014) seems to be attenuated when warm, affectionate attitudes and behavior are shown by parents, which promote a healthy degree of autonomy and independence in the child, as evidenced by previous studies (Gámez-Guadix et al. 2010;Ma et al. 2012).For this reason, the levels of involvement of children with punitive democratic parents are lower than those with authoritarian or strict parents.However, to draw a clearer conclusion about these tendencies, more research is needed into educational profiles and their impact on involvement in bullying and cyberbullying in children.
This study has certain limitations, such as the use of self-report measurements to collect the data.Self-report surveys tend to elicit subjective and somewhat spontaneous answers from the respondents, which means we must be rather cautious and reserved when drawing our conclusions.Furthermore, parents have not been included as participants, taking only into account the teenagers' perception about parenting styles.In addition, the cross-sectional design of the study prevents us from establishing causal relationships.Future research should conduct further longitudinal studies into showing the sustainability or changes in the filial perception of family education, as well as the inclusion of other variables related to other individuals and contextual factors linked to the phenomena of bullying and cyberbullying.Moreover, it could be interesting to include the parental reporting about their own parenting styles to compare with that of the adolescents.

Conclusions
This research confirms the major role that parenting play in their children's involvement in bullying and cyberbullying.Abusive practices by the parents, such as physical punishment and psychological aggression, accompanied by a lack of affection and poor communication, little promotion of autonomy, and excessive behavioral control, which all combine to increase their children's vulnerability and make them more prone to victimization or bullying at school or behind a screen (Gómez-Ortiz et al. 2015).These findings highlight the importance to include parents as a key element of intervention and prevention programs not only to increase their awareness about their role in the development of these violent phenomena and improve their parenting skills, but also to teach them the best way to help their children cope with bullying and cyberbullying involvement (Larrañaga et al. 2018).In this sense, the intervention should guide parents to become in "indulgent democratics," which means to promote affect, communication, and autonomy showing also a positive humor.Parents should also be able to fix fair and enough rules and to monitor children behavior avoiding manipulative strategies such as those included in psychological control.Regarding discipline, the best seems to be to use inductive procedures based on explanations, rewards, and the use of attention to avoid children's misbehavior.In any case, these practices should be accompanied by positive and close attitudes that allow parents to be empathetic with their children while they are also able to make them understand that they have to respect parents (as parents respect them) and that their behavior has consequences.However, this study only includes the assessment of parental practices and not attitudes, as seen in most parenting studies.This is because the most of the parenting scales are focused on practices and ignore values and attitudes, which are important aspects of the parenting styles (Darling and Laurence 1993).Therefore, it would be necessary to design instruments

Table 1 .
Final areas of conglomerates of groups defined according to the scores on parenting practices.

Table 2 .
Abstract of qualitative description of each style of parental education management.

Table 3 .
Differences in involvement in bullying and cyberbullying, according to the parents' educational profile.