A Deeper Understanding of Child to Parent Violence (CPV): Personal Traits, Family Context, and Parenting

The increase in the rate of child to parent violence is a concern for society in various countries. Different psychological and personal characteristics tend to define the profile of the minors who commit this type of offense. Various social factors have been also related to this type of violence, including exposure to violence, the family environment and parenting. The relationship between child to parent violence and previous exposure to violence has yet to be clarified. Comparatively little research on this crime involving samples from juvenile court has been undertaken. This study uses a standardized measure (YLS/CMI) to determine the extent to which three of the most extensively studied groups of factors in child to parent violence—the family context, parenting and the adolescents’ psychological characteristics are relevant in the predicting this type of violence. The sample consisted of a total of 342 juveniles from a Juvenile Court in Eastern Spain, dealt with under the terms of Organic Law 5/2000 regulating the Criminal Responsibility of Minors. A child to parent violence group is compared with a control group committing the entire range of offenses. Personal variables, antisocial personality and exposure to violence play a clear role in the commission of this type of crime. Parenting has a determinant influence even when compared with family characteristics. What affects the commission of this type of violence in the most serious cases is therefore not having been exposed to it, but instead the individual’s upbringing and their current relationship with their parents.


Introduction
Child to parent violence is one of the types of violence that causes the most concern in our society today. Child to parent violence can be defined as any reiterative act committed by a son or daughter toward his or her parents, (or to the people performing that role) which causes physical, psychological, or financial harm (Pereira et al., 2017). One of the most used definition is the one provided by Cottrell (2003) who defines child to parent violence as "any harmful act by a child intended to gain power and control over a parent. The abuse can be physical, psychological or financial." Other authors such as Patterson and Fisher (2002), consider the CPV when others in the family feel threatened, intimidated, or controlled by their son or daughter and if they believe that must adjust their own behavior due to threats or violence.
Meanwhile, "battered parent syndrome" was described by Harbin and Madden (1979) as physical assaults or verbal and nonverbal threats of physical harm and distortions in the hierarchy of generational authority. This involves various acts, such as using real harm, threats of harm, intimidation, or different manipulative techniques, which the child uses to consequently gradually gain power and control (Cottrell, 2001).
It is difficult to provide accurate data about the prevalence of this crime. The different methodologies, procedures, and samples used in studies and even the definition of the violence taken into account in each one hinders comparisons of results and rates, and hampers accurate measurement of this phenomenon (Gallagher, 2008;Holt, 2012). Nevertheless, the increase in the rate of occurrence of this violence in different countries seems to be a concern to society (Gallagher, 2008;Kennair & Mellor, 2007). The rates range from 7% to 29% in the United States, Spain, Canada, France, and Australia in some studies of its prevalence (Bobic, 2002;Cottrell & Monk, 2004;Romero et al., 2007;Selwyn & Meakings, 2016). The differences in rates may be due to the difficulties for parents in reporting the assaults. The differences in rates may be due to the difficulties for parents in reporting the assaults, the differences in procedures mentioned above (methodologies, procedures, and samples), but also to differences in cultures and the penal systems of the different countries. The minor has criminal responsibility at a different age depending on the CPV legislation of each country. Spain has a specific legislation that includes CPV (Circular 1/2010 from the Chief Prosecutor on crimes by children toward relatives). Minors are criminally responsible from the age of 14 as in Bulgaria for example, whereas in England and Ireland these ages are 10 and 12 years, respectively. According to Condry and Miles (2014), in the UK, adolescent to parent violence lacks of appropriate legislation or policies. If the adolescent was under the age of 18, adolescent to parent violence was not officially defined as domestic violence. This fact led to a lack of official figures regarding the prevalence of the problem as well as to a deficiency in the support for families. In 2013 this age was extended to 16 years old, making visible the perpetrators of these age groups.
The reluctance has been related to the parents' feelings of stigma, shame and guilt at having a child who abuses them. Parents tend then to avoid seeking help from social services or legal authorities, and as a result they suffer from the aggression until the situation is untenable (Downey, 1997;Edenborough et al., 2008;Selwyn & Meakings, 2016). Current research is restricted by reported cases, which are also thought to represent the most severe cases perhaps cases which are particularly dangerous, high risk or when this behavior has been longstanding (Méndez, 2020;Moulds et al., 2019a).
The gender proportion in general rates of delinquency without selecting a specific crime generally varies between 80% and 90% of males and 10% and 20% of females. Nevertheless, this percentage of females is higher for CPV samples (when is compared to the general offender) (Cuervo et al., 2017;Ibabe & Jaureguizar, 2010). It varies between 20% and 30%, and appears to increase with age Edenborough et al., 2008;Gallagher, 2008;Laurent & Derry, 1999;Romero et al., 2005;Walsh & Krienert, 2007). However, the types of sample and methodology used have little influence on these differences. Meanwhile, epidemiological studies have found no differences according to gender (Ibabe & Jaureguizar, 2010;Pagani et al., 2004). In more specific terms, the proviso could be made that there are differences according to gender depending on whether the type of violence is committed is physical, emotional, or verbal, as boys tend to be more physically aggressive, while girls tend to be more emotionally and verbally abusive and manipulative (Boxer et al., 2009;Calvete et al., 2011;Kennair & Mellor, 2007). The age of the aggressors in this type of violence varies. Studies have found different ranges between 10 and 21 years old (Edenborough et al., 2008;Laurent & Derry, 1999;Walsh & Krienert, 2007). Some studies have identified ages from 14 to 17 years (Walsh & Krienert, 2007) and others from 12 to 14 years (Cottrell, 2001). Due to the Spanish Criminal Responsibility Law, different authors place the age range of the aggressors between 14 and 18 years old (Garza, 2019;Ibabe et al., 2007;Romero et al., 2007;Sempere et al., 2006). However, if we keep the law aside, studies point to the range between 12 and 17 years as the most prevalent ages (Cottrell & Monk, 2004;Romero et al., 2005;Routt & Anderson, 2011;Sheehan, 1997;Walsh & Krienert, 2007). Indeed, Bailín-Perarnau et al. (2016) indicate that the average age at the onset of violence is 11 years old. It appears that a specific age for committing this type of violence cannot be established.

Psychological Profile of Juveniles
Various psychological and personal characteristics tend to define the psychological profile of these minors. These adolescents are usually characterized by feelings of unhappiness, low self-esteem, low empathy, poorer tolerance of frustration, problems with affect regulation, and impulse control, making them more demanding, oppositional and aggressive than other adolescents (Calvete et al., 2011;Cottrell & Monk, 2004;Nock & Kazdin, 2002). Some of them also may present some symptoms of depression, attention deficit, hyperactivity disorder, learning disorders, behavior disorder, bipolar disorder, and even schizophrenia. However, the majority did not present any mental illness (Ibabe & Jauregizar, 2009;Ibabe & Jaureguizar, 2010;Paulson et al., 1990). However, some cases may indeed present some callous-unemotional traits related to the beginning of maltreatment at an early age, which may have worse intervention outcomes (Curtis et al., 2019). Actually, CPV offenders present differences in relation to the kind of charges committed. They tend to commit more violent offenses and significantly fewer property offenses (Kennedy et al., 2010). Even when gender was considered, the biggest group was the one presenting violent offending, that is to say, juveniles who committed also other violent offenses (Moulds et al., 2019a). This research concludes that the juveniles who use this type of violence are more likely to belong to a violent antisocial trajectory, rather than a non-violent offending trajectory.
On the other hand, low self-esteem is a trait which has been present in various studies of juveniles who abused their parents (Calvete et al., 2011). These juveniles presented lower self-esteem compared to juveniles who did not assault their parents. This difference was present in all types of violence toward parents: physical, emotional, and psychological (Ibabe & Jaureguizar, 2010;Ibabe et al., 2013).

Family Context
Different social factors have been found to be related to this type of violence: different types of exposure to violence at home, the type of neighborhood, peers, school, community, and violence on television. Other factors related to CPV are related to the characteristics of the family: inconsistent parenting, the type of family, drug use, adoption, single-parent families, among others, some of which are described below. Among the different types of exposure to violence that have been related to CPV (exposure to violence in school, community, TV), the strongest factor related to CPV appears to be exposure to violence at home and the quality of the home environment. It seems having been a direct victim of the parents predicted subsequent aggression toward them. Children who abuse their parents were more likely to have previously experienced violence at their hands (Boxer et al., 2009;Langhinrichsen-Rohling & Neidig, 1995;Mahoney & Donnelly, 2000;Straus & Hotaling, 1980). Similarly, aggression from the parent to the child increased cases of CPV (Boxer et al., 2009;Cottrell & Monk, 2004;. A Court comparison between juveniles who committed CPV and juveniles who had not committed this crime, the CPV group was significantly more exposed to both violence and victimization at home (Contreras & Cano, 2016). Victimization could therefore play a crucial role in the appearance of violent behaviors by children toward parents (Contreras & Cano, 2016). In fact, exposure to violence in the home has been related to other types of violence, such as violence toward intimate partners in adulthood (Browne & Hamilton, 1998;Cottrell & Monk, 2004;Gebo, 2007;Ibabe et al., 2020). On relation to this, Condry and Miles (2014) found that among cases reported to the police, CPV is a largely gendered phenomenon, although not exclusively. Since sons form the majority of perpetrators and mothers most victims. It has also been shown that parents feel more physically threatened by sons than by daughters, and perceive that it is more socially acceptable to report violence from a son than from a daughter. However, this fact could also be influenced by mothers feeling more threatened, fearful, or vulnerable than fathers (Condry & Miles, 2014). In the same way, in compliance with Holt and Lewis (2021), CPV is a "subset of domestic violence, where domestic violence is the collective name for a wide range of behaviors" such as elder abuse, "honor" based violence, female genital mutilation (FGM) and forced marriage, (Home Office, 2015).
Experiencing direct or indirect violence at home and being a victim of maltreatment could be one of the most explanatory factors for CPV according to social learning (Bandura, 1977). Children who have live with violence at home might consider that this way of dealing with other people is tolerable and acceptable, and may repeat the maladaptive patterns (Bandura, 1977;Calvete, 2007). Children construct their model for interpersonal relationships based on their close caregivers using internal working models, which may lead them to have relationship problems (Bowlby, 1980). The child may interiorize the aggression as being appropriate and justified, or attribute righteousness or positive consequences to it, in order to gain control (Calvete et al., 2015;Ibabe et al., 2013). Furthermore, these children and adolescents also lack the positive role models and the adaptive patterns of appropriate behavior (Gómez, 2011). Children exposed to violence at home may therefore lack the ability to manage escalating emotions, and the skills required to resolve conflicts and problems in constructive and nonviolent ways (Izaguirre & Calvete, 2017;Siegel, 2013).
The adolescent's traits influence the relationship with the parents, and the parent's characteristics influence the behavior of the child. The child's temperament means that he or she finds it more difficult to obey rules and hence parents may fail to respond effectively. Explosive, irritable and inconsistent behaviors lead to a noncompliant child and difficulties in discipline that may create a kind of rejection by the parents, which simultaneously acts as a factor in the conflicts persisting. Negative paternal reactions are higher in problematic families and abusive families than in asymptomatic samples (Patterson, 1982). When a child is difficult, a non-contingent parent may lead to a compliance failure (Bornstein, 2002). The family environment appears to be characterized by high levels of conflict, low levels of cohesion and the presence of gender violence (Ibabe, 2015). Studies show that many mothers present lower rates of prosocial behavior and inappropriate upbringing (Agnew & Huguley, 1989;Bousha & Twentyman, 1984;Ibabe et al., 2013;Paulson et al., 1990). In addition, some parents express dissatisfaction, saying that their lives are empty, meaningless (Pereira & Bertino, 2009). In terms of communication and attachment models, a weak parentchild bond seems to be present with a lack of communication or emotional connection between parents and adolescents (Agnew & Huguley, 1989;Duffy & Momirov, 1997;González et al., 2013;Paulson et al., 1990), which may lead to feelings of rejection by the parents (Aluja et al., 2005;Ibabe, 2015). Indeed, children who experienced rejection were related to this type of violence on three levels (physical, psychological, and emotional violence toward parents (Ibabe et al., 2013). In other words, the evidence for the Bi-directionality of family violence and aggressive behavior has been increasingly prominent in the literature (Juan et al., 2020;Kennedy et al., 2010;Mahoney & Donnelly, 2000;Maxwell & Maxwell, 2003;Pagani et al., 2009). One of the most commonly related factors to this type of violence is therefore exposure to violence, but family characteristics also play an important role in this interactive cycle, together with the characteristics of the juvenile.
These mechanisms described above are aggravated or influenced by various problems or conflicts related to stress, such as breakups, divorce, violence, mental health, drug use, and poverty. The increase in marital conflict or economic stressors is associated with disruptions in discipline and monitoring practices. The situation is of course aggravated when the conflict leads to violence (Calvete et al., , 2014Ibabe & Jaureguizar, 2009;Villanueva, 2020). These are crucial points when dysfunctional parenting styles may arise or deteriorate. Low socioeconomic levels may therefore be related to this type of violence (Cottrell & Monk, 2004). However, there is no agreement in this regard, since other studies report that socioeconomic levels do not influence results (Calvete et al., 2011;Romero et al., 2005). Other research concludes that child aggression toward parents occurs in every social class (Agnew & Huguley, 1989;Pelletier et al., 1999), while other studies state that around 75% take place in middle -middle and upper-middle socioeconomic level families (Aroca-Montolío et al., 2014;Ibabe et al., 2007;Rechea et al., 2008;Romero et al., 2007). Bronfenbrenner (1979), defines violence as a relational phenomenon or interpersonal process in which different variables converge, from the individual to the converge of different levels of impact, from the individual to the macro-social. Therefore, if we apply the ecological perspective to this area, we obtain that the context, including the emotional and affective climate, affects the individual exchange.
The structure of the family is an important factor in CPV. It was found to a greater extent in single-parent families, single mothers, separated/divorced old, or young parents and increasingly, adoptive parents (Barbolla et al., 2011;Pérez & Pereira, 2006;Tremblay et al., 2004). In terms of their family structure, some studies find that a small percentage of adolescents who commit this crime live with both parents (Ibabe & Jaureguizar, 2010;Kethineni, 2004;Purcell et al., 2014;Walsh & Krienert, 2009). This may be related to a lack of supervision from the parents, and the added difficulties involved in a more limited ability to communicate and impose discipline with a volatile juvenile by a mother who is usually the principal caregiver. Some biological conditions and adverse experiences such as abandonment, insecure attachment, family breakdowns and violence may lead to abuse of the parent. Adoption may therefore be related to CPV families in this respect. However, others found that the problem was more common in two parent families (Weinblatt & Omer, 2008). (Weinblatt & Omer, 2008).
It is possible that when there are problems in the relationship, one of the parents forms an alliance with the son or daughter against the other parent. These arguments may involve criticizing and undermining the other partner.
Later, when the breakup occurs, the custodial parent may adopt inappropriate parental roles. In certain cases, the parent-child relationship might be too symmetrical. In these specific cases, when the mother or father tries to exert authority or to impose discipline, and some of the previous risk factors take place, the son or daughter may not obey and respond with violence and disobedience (Agustina & Romero, 2013).
The coercion theory explains how a negative relationship is created, due to hostile and aggressive interactions influenced by harsh punishment (Granic & Patterson, 2006;Patterson, 1982Patterson, , 1986. Parental control strategies such as verbal punishment (Pagani et al., 2009) and emotional deprivation (Agnew & Huguley, 1989) have been related to CPV (Brezina, 1999;Patterson et al., 1984). Other more recent research has also found negative consequences of permissive parental styles, or in other words, parents who give the child whatever they want or who are inconsistent with their rules (Cottrell, 2005;Paulson et al., 1990;Robinson et al., 2004). As a result, both types of parenting, negative parental discipline and supervision, as well as hostile and punitive parenting, are related to antisocial behavior in adolescence. It seems that depends on research which factor is the key for child to parent violence (Gallagher, 2008). Some studies have focused on permissiveness, and on the consequences of applying inconsistent rules (Cottrell, 2005) while other studies highlight control and authoritarianism, defined by humiliation, infantility, and resentment in the treatment of adolescents (Gallagher, 2004;Garrido, 2005;Straus & Stewart, 1999). In both cases, there would be general difficulties in child-parent relationships, and in particular among parents with unrealistic expectations, and a lack of adequate communication skills (Kennedy et al., 2010;Paulson et al., 1990;Peek et al., 1985).
The relationship between child to parent violence and previous exposure to violence has yet to be clarified (Contreras & Cano, 2016), and comparatively little research has been undertaken with samples from juveniles from Court (Purcell et al., 2014).
As seen in the previous sections, some personality and family factors may influence this problem. These individual factors could be depression, low self-esteem, as well as factors related to low levels of empathy and anti-social characteristics. Whereas the relevant family factors refer to those related to family structure, type of parenting and exposure to violence. Even though, theses individual and family factors are also present among young people no related to CPV.

The Present Study
This study compares three of the most relevant and extensively studied groups of factors in the prediction of CPV: psychological characteristics of adolescents, family context and parenting. The main aim is to analyze to which extent these features are relevant in predicting CPV. The analysis will be undertaken using standardized measures, as this is required when studying this specific type of violence (Ibabe & Jaureguizar, 2010).
The first objective was to analyze which are the differences between the profiles of juveniles who committed CPV and the juveniles who committed other types of crimes. Second, this study aims to analyze the specific influence of 3 groups of variables in CPV: personal/psychological traits of the juvenile, family context and parenting. According to the hypothesis of the bi-directionality of violence (Mahoney & Donnelly, 2000), it is expected that exposure to family violence will be more decisive in predicting this type of violence.

Sample
The sample was composed of a total of 342 juveniles from the Juvenile Court of Castellón (in the East of Spain) under the terms of Organic Law 5/2000 regulating the Criminal Responsibility of Minors. In Spain, minors must be from 14 to 17 years old to be charged under the juvenile legal system. A young person under the age of 14 remains outside this law. According to the Organic Law 5/2000, Spain has a specialized system for juveniles. A disciplinary record is created in the Juvenile Court when a juvenile is accused of committing a crime.
The sample contained records from 2010 to 2017, covering 7 years of follow-up. The experimental group contained 154 juveniles, who were all the juveniles who committed this type of crime within the previous period (mean age 15.81), of whom 36.4% were girls and 63.6% were boys. The child to parent violence is included in the Spanish Criminal Code as an "intra-family violence" or "domestic violence" crime. For the realization of the study, it was confirmed with the Judicial Psychosocial Team that all cases in the sample were CPV files. These files gathered all the cases of adolescent-toparent violence committed in this period of time.
The comparison group was selected by deleting one juvenile in every 4, listed by the case number within the same period of time. The participants in this group were 188 juveniles charged with different types of crimes, 19.7% of them were girls and 80.3% were boys (mean age 15.81 years old). The age was calculated when the offense took place in all cases. The juveniles included in this group include those committing all types of offenses, ranging from minor offenses to crimes committed against persons and against property, such as shoplifting, fraud, reckless driving, robbery, as well as assault and sexual assaults.

Instrument
The instrument used to carry out the evaluations of the minors in the present study was the Youth Level of Service/Case Management Inventory (YLS/CMI) Hoge and Andrews (2006) (Garrido et al., 2006). This instrument involves hetero evaluation of the risk of recidivism by 42 items grouped into 8 factors. The instrument provides a total score for assessing the risk level of recidivism: low risk, from 0 to 8 points; moderate, from 9 to 22; high, from 23 to 32; and very high, from 33 to 42. Each item can be marked as present (1 point) or absent (0 point). The items in various domains are added together to provide a total score ranging from 0 to 42. The eight factors were: Prior and current offenses/adjudications (five items), (2) Family circumstances/ parenting (six items), (3) Education/employment (seven items), (4) Peer relations (four items), (5) Substance abuse (five items), (6) Leisure/recreation (three items), (7) Personality/behavior (seven items), and (8) Attitudes/orientation (five items). Part III of the Inventory also takes other needs and special considerations into account. The items in this section are in two groups: those related to the youth's family circumstances (marital conflict, abusive father, etc.) and those related to the youth himself or herself (low self-esteem, depressed state, etc.). The items in this section are not included in the risk assessment, but might be relevant for disposition, case planning or intervention.
The instrument is completed by a member of the Youth Offending Team of the Juvenile Court, combining data from multiple informants such as interviews with the youth and his or her family, previous court records, and information from other social centers, high school and others. These professionals interview both the minor and his/ her legal representatives about individual, educational, family, and social aspects in the youth's environment. The authorization to apply the YLS/CMI to this sample was obtained from Multi-Health Systems (MHS).
The YLS/CMI is one of the most widely used risk assessment for youth. It assesses the static and dynamic risk factors that predict the risk of recidivism and risk of violence in adolescents. The YLS/CMI it is widely used in forensic populations. It has been designed to aid probation officers, youth workers, psychologists, and social workers in assessing adolescent offenders. The instrument was developed from a theory of criminal behavior (Hoge & Andrews, 2006). The authors designed the YLS/ CMI with the pre-sentencing and post-sentencing evaluations of juvenile justice in mind, thus the instrument allows its use in the complete management of the judicial case (Wenger & Andrés Pueyo, 2016). There have been many field studies concerning the reliability and predictive validity of the Youth Level of Service/Case Management Inventory (YLS/CMI) (e.g., Olver et al., 2009;Schmidt et al., 2005). It is a Fourth-Generation risk assessment, that is to say, emphasizes the link between assessment and case management. Other studies have been conducted without standardized tools at all and arbitrarily selecting information about the personal files of the minors. Therefore, we believe that this instrument would be related to the object of study.

Procedure
The data for this study were obtained from an analysis of the records of minors in the Juvenile Court of a Spanish province (Castellón). These charges were filed in 135 municipalities, covering a total of 604,344 inhabitants. This study was carried out using reports drawn up by the Judicial Psychosocial Team, the Police and the Social Services at the Juvenile Court. The Judicial Psychosocial Team obtained information about the juveniles and their family circumstances by means of interviews with the adolescents and their parents, filling in the part III of the Inventory. Data were anonymously gathered with the permission of the Juvenile Court of Castellón Spain. Verbal informed consent was obtained prior to the interview from all individual participants included in the study.
(2) Family context: Adoption (yes/no). Exposure to violence (abusive father, abusive mother and marital conflict; from 0 to 3), Financial problems (yes/no), Traditional family (yes/no). (3) Family circumstances/parenting (Subscale 2; from 0 to 6). This subscale is related to inappropriate supervision, inconsistent discipline, and negative relations with the mother and father.

Data Analysis
The data were entered into the IBM SPSS Statistics 26 package in order to compare the two groups, and contingency tables were drawn up. The discrete variables were analyzed using the Chi-square test. ANOVA was used to examine significant group differences in continuous variables. Logistic regression was conducted in order to analyze the predictive value of each of the three groups of variables for the type of crime committed. The presence of CPV was given a value of 1 in the parameter coding, and 0 was assigned to the control group. The Wald test was used to check the significance of model parameters. A significance level of α < .05 was considered.

Results
The demographic and other personal traits of juveniles are presented in Table 1, with the percentages for the chi-squared test and mean and standard deviation by ANOVA. All the differences between groups were significant. The juveniles in the CPV group presented more cases of Depressed states, Callousness, Low self-esteem and higher mean scores in Antisocial personality. By gender, the percentage of girls was higher in the CPV group, at 36.4% versus 19.7% observed in the control group. The family context and structure factors are presented in Table 2. All the differences were significant except for financial problems. Adoption, exposure to violence and Family circumstances/parenting were more common in the CPV group, and traditional family structures were more common in the control group. Table 3 presents the results of the bivariate correlation (N = 342). Family circumstances/parenting was the strongest factor related to the commission of CPV (r = .55, p < .001), followed by Antisocial personality (r = .43, p < .001), Depressed state (r = .36, p < .001), and Exposure to violence (r = .24, p < .001). A negative relation of CPV was found with traditional family (r = −.17, p < .001). Low self-esteem (r = .24, p < .001), Callousness (r = .15, p < .005) and adoption (r = .13, p < .005) were significantly related to CPV. Being a boy was negatively related to CPV (r = .18, p < .001).
The following binary Logistic regression analyses are conducted to check the prediction of CPV. These analyses were carried out with the crime type group of juveniles as dependent variables (CPV = 1; Non CPV = 0) with the three groups of predictor variables. The variables were presented in two groups: personal factors (Table 4) and family factors (Tables 5 and 6).
In the personal variables, the significant predictors of CPV were the Depressed state and Antisocial personality/behavior subscales (Nagelkerke R 2 = .368).
Variables related to the family context were also analyzed (Table 5). Significant predictors were Exposure to violence and Adoption (Nagelkerke R 2 = .115). Adoption was negatively related to CPV, but the odds ratio was nevertheless very weak (.09), and as such its involvement in predicting CPV is minimal.
Family circumstances/parenting was added to the previous group of factors related to the family context. This factor was significantly associated to CPV, whereas the Exposure to violence factor was no longer significant in the new model. Adoption was still significant, but with the same low odds ratio value. The highest Negelkerke value  was obtained when the factor of Family circumstances/parenting is entered in the model, and has a stronger predictive power than the other family variables (Nagelkerke R 2 = .412).

Discussion
The primary objective of this article was to analyze the differences in the juvenile profile in terms of the type of crime committed (CPV vs. other crimes). The CPV group showed higher rates of Emotional deficits, Callousness, higher scores in   Antisocial personality and Low self-esteem. In relation to their family context, the juveniles in the CPV group presented higher percentages of adoption and non-traditional family structures, and higher scores for Exposure to violence and Family circumstances/parenting. Low self-esteem in juveniles is a consistent result in this type of violence. Low self-esteem has traditionally been related to antisocial conducts in general (Ibabe & Jaureguizar, 2010;Ibabe et al., 2013). However, the percentage of minors who presented low self-esteem was considerably higher in the CPV group than in the control group. This factor was also a predictor of CPV in several studies (Contreras & Cano, 2015;Ibabe & Jaureguizar, 2010;Ibabe et al., 2013). The Depressed state was also present in the differentiation between two groups. This state was considered to be present when the juvenile showed clear signs of emotional instability or signs of depression. Minors committing this type of offense who are undergoing psychological family treatment but not for serious pathologies are common (Ibabe & Jaureguizar, 2010;Paulson et al., 1990). The mean scores for antisocial and aggressive personality/behavior were higher in the CPV group. Violent behavior usually becomes apparent at early ages, and the longitudinal studies show signs of this type of behavior. For instance, the degree of risk of aggression toward fathers increases in proportion to the severity and chronicity of violent behavior from grades 1 through 6 (Pagani et al., 2009). On the other hand, as adolescents, those rated as chronically aggressive by their primary school teachers were between six and three times more likely to engage in verbal and physical aggression toward their parents than their nonaggressive peers (Pagani et al., 2009). Disruptive behavior in school and high school has also been related to CPV (Ibabe & Jaureguizar, 2010;Pagani et al., 2004). Other research with 1,400 adolescents found a relationship between CPV and the use of violence and hostile social perceptions. In other words, the use of violence was justified by statements such as: "because he or she deserves it" (Calvete et al., 2011;Contreras & Cano, 2016). This may be related to the higher percentage of callousness in the CPV group (Curtis et al., 2019).
As for gender, 36.4% were girls in the CPV group, whereas 19.7% were girls in the control group. A very similar percentage of girls, 34.4%, was reported by Ibabe and Jaureguizar (2010). Compared to the general profile of minors in Juvenile Courts, girls commit CPV more often than other types of crimes (Calvete & Veytia, 2018;Cuervo et al., 2017). When variables of family context are analyzed, there were no differences between the groups for financial problems. When variables of family context are analyzed, there were no differences between the groups for financial problems. As it was argued on the introduction section, there was no agreement on the typical socioeconomic level of these minors (Calvete et al., 2011;Romero et al., 2005). The diversity of results found in this respect in the different investigations, as well as in this one, may be due to the minors that compose the group of CVP. In this case, the criterion for the differentiation of the groups was to have committed CPV. Therefore, many of the minors who committed this type of violence also committed other types of crimes, which could have influenced the similarity of the socioeconomic levels in both groups. For future research, it would be interesting to explicitly analyze the minors who committed only CPV in their criminal trajectory.
The proportion of adopted minors who perpetrated child to parent violence is higher compared to the minors who committed other types of crimes (Alemany, 2019;Del Moral Arroyo et al., 2015). Adoptive parents may report a sudden onset of violence related to the beginning of puberty (Selwyn & Meakings, 2016;Selwyn et al., 2014;Urra & Urra, 2015). There may be bonding problems and differences between the parents' expectations and reality in adoptive families, especially for older children (Agustina & Romero, 2013;Barbolla et al., 2011;March Ortega, 2017), although this prevalence may be a problem related to social support and coping strategies rather than to the family structure (Gil, 1998). Likewise, in juvenile centers, the proportion of adopted minors who have committed assaults against their parents is much higher than that of young people who have committed other types of crimes. Little research has been done on this topic, although it would be important to investigate whether attachment problems and early experiences of abuse or mistreatment lead to situations of child to parent violence (Alemany, 2019;Del Moral Arroyo et al., 2015).
One-parent families were more common in the CPV group than in the control group. This single parent usually is the mother, as discussed in previous research (Ibabe & Jaureguizar, 2010;Pagani et al., 2003;Walsh & Krienert, 2007). Exposure to violence was significantly higher in the CPV group. CPV had double the score for exposure to violence in the home compared to the control group. Exposure to violence, considered as a conflict of interactions between the parents, was also related to CPV by Contreras and Cano (2016). This study concludes that being exposed to violence in the family is specifically important for the development of a social hostile perception. The score for the Family circumstances/parenting subscale was also higher in the CPV group. Inadequate discipline and poor relationships are more common in the CPV group than in the general offenders group (Calvete et al., , 2014Ibabe & Jaureguizar, 2009;Villanueva, 2020). A high proportion of parents in which child to parent violence occurs have problems and needs such as difficulties with drugs, alcohol, higher rates of trauma or mental health concerns. These complicated profiles for both victims and perpetrators may influence the behavior of each other which complicates to define the causes for this violence (Diaz, 2016;Maroto & Cortés, 2018;Moulds et al., 2019b). As a result, juveniles who abuse their parents present a clearly riskier psychosocial profile according to the YLS/CMI items compared to the general offenders' profile.
This study aims to analyze the specific influence of three groups of variables that have been found to be particularly relevant in CPV (Personal/psychological traits of the juvenile, Family context and Family circumstances/parenting). According to the hypothesis of the bi-directionality of violence (Mahoney & Donnelly, 2000) it is to be expected that exposure to family violence will be more decisive in predicting child to parent violence. However, this hypothesis was not supported by the results. Psychological factors and in particular, Depressed state and the Antisocial personality/ behavior subscale, presented a major impact on the prediction of CPV (Nagelkerke .36). Depressive symptoms predicted an increase in CPV. Adolescents who assault their parents tend to have symptoms of distress and depression (Kennedy et al., 2010). Gender was not significant when introduced with the personal variables. The juvenile's psychological characteristics seem to be more important in predicting CPV, rather than being a boy or a girl. CPV may be one of the few crimes where the proportion of boys and girls is not a determinant in the prediction of the crime. Boys usually tend to commit most of the crimes coming before the Juvenile Court (Cuervo et al., 2017). As regards family structure, being adopted has been ruled out, even when significant. Economic difficulties were not significant. In fact, in some studies this type of violence has been related to middle and upper class families (Cottrell & Monk, 2004;Paulson et al., 1990). CPV does not seem to be related to financial status. It may be more closely related to stress and lack of support issues in one-parent families.
Exposure to violence was composed of an abusive father, abusive mother and marital conflict. This factor was a predictor for CPV. The fact that growing up in a violent family environment leads to parental abuse has been gradually proven. The etiology of antisocial conducts in children usually has its origins in parent-child interactions, and as such CPV could follow the same path (Patterson, 1982), which would be consistent with the hypothesis of bi-directionality. CPV would therefore be a logical consequence of early hostile and aggressive interactions between the caregivers and their children. Exposure to violence at home was also found to be a predictor in another Spanish sample of 1,400 adolescents, in which physical, verbal and total violence predicted CPV (Calvete et al., 2011). Of all the different types of exposure to violence, the crimes which a general offender commits are more closely related to the exposure to violence in the community, whereas child to parent violence is related to the exposure to violence at home. Exposure to violence at home was found to be the best predictor of CPV (Contreras & Cano, 2016). The bi-directionality hypothesis is therefore supported by results when it is analyzed with family structure variables. Violence is thus returned to the parents, with an intergenerational transmission of violence (Harbin & Madden, 1979). An outstanding but equally important question is whether these juveniles really abused their parents, or of this behavior is a response to a mutually offensive or defensive behavior. Some individuals who are violent to their parents have been also victims of bullying in school, showing another kind of transmission of aggression (Izaguirre & Calvete, 2017). However, having witnessed violence at home was not a significant factor between the CPV and non-CPV groups (Contreras & Cano, 2016). Witnessing interpersonal violence against the mother was less relevant than direct victimization for the subsequent commission of CPV (Izaguirre & Calvete, 2017). A key factor which could explain this phenomenon is therefore the extent to which the minor experiences this violence. In fact, the perception of victimization was reported by Contreras and Cano (2016) as being the best predictor.
However, according to the results, a violent context had a lower predictive value than Family circumstances/parenting. The Family circumstances/parenting subscale on the YLS/CMI seems to be the key factor in CPV in this study. This parenting involves a lack of supervision, harsh punishment discipline, and relationships with parents. It is therefore a mixture of coercion, permissive styles and poor bonding which leads to this crime. When this is considered, the exposure to violence is no longer significant in the regression for prediction of CPV. The predictive power of exposure to violence is apparent, but when parenting enters in the equation its effects are conclusive, with exposure to violence no longer relevant. The child's compliance depends on the interaction of both parents and the child. A difficult child with a noncontingent parent may lead to the child's compliance failure (Shaw & Winslow, 1997). A failure in this initial socialization may lead to a delay in social development or antisocial behavior. Changes in parenting practices over time are associated with changes in the child's level of antisocial behavior (coercion model) (Patterson & Bank, 1986). The effects are bidirectional, and as such it also modifies the relation with the caregiver (Patterson & Fisher, 2002). This effect is explained by the Coercion theory (Granic & Patterson, 2006;Patterson, 1982;Patterson & Fisher, 2002). This theory focuses on coercive education, and assumes that CPV is a logical consequence of earlier hostile and aggressive interactions between parents and children. This familiar cycle of conflict during childhood is involved in the evolution of the adolescent's aggression toward their parents (Ulman & Straus, 2003). Coercive parental practices were predictive of teacher-and peer-reported aggression in the short term (6 months later) (Dodge et al., 1990) and violent and criminal offenses in the long-term (Farrington, 1991). The influence of harsh punishment of children by parents has raised the question of the importance of parenting styles in the development of violent behaviors in children. The results highlight Family circumstances/parenting such as harsh punishment, inconsistent parenting, inadequate supervision, as well as a poor relationship with parents as the main precursors for CPV. Parental styles with excessive control over minors and frequent use of punishment and even physical bodily harm as a strategy for discipline have been related to CPV (Gallagher, 2004;Garrido, 2005).
These practices were also related to lower levels of moral internalization and overall mental health (Gershoff, 2002;MacMillan et al., 1999;Pagani et al., 2009). Studies conclude that abuse of parents may occur when the parents are excessively controlling or authoritarian (Cottrell & Monk, 2004;Gallagher, 2004;Pagani et al., 2004). This strict control, without taking into account the adolescent's age may cause feelings of humiliation, infantility and resentment (Straus & Stewart, 1999). Aggressive adolescents describe their parents as less emotionally warm, and they feel neglected by them (Aluja et al., 2005). When the relationship is damaged, communication becomes more difficult and the boundaries for each role are blurred. This is linked to the lack of authority that authoritative parents appear to lose when nurturing children Etxebarria et al., 2009;Harbin & Madden, 1979). These findings highlight risk factors common to both verbal and physical aggression by adolescents, and suggest problematic parent-child interactions that probably originated in early and middle childhood (Pagani et al., 2009). This may explain the predictive influence of the factor of parenting, which also includes items related to the relationship with the parents. Likewise, some research indicates that it is the inconsistent use of punishment rather than the punishment itself which contributes to abuse of parents (Ibabe et al., 2013).
One possible explanation that arises from the above is that aggression toward parents could be retaliation by the children if harsh and parenting styles are perceived as demeaning and degrading, and especially by those who have never developed appropriate anger management or self-regulation strategies (Pagani et al., 2009). Permissive styles have recently also begun to be considered as negative as authoritarian styles. These types of styles, which have no fixed boundaries, seem to be related to a lack of hierarchical structure between the parental and subsidiary levels, creating shortcomings in setting standards and in supervision Pereira, 2011). This lack of discipline has been related to other problems of behavior in adolescence (Dodge et al., 2008). Today, it seems that excessively permissive parents are overprotective for various reasons, and in some cases give their children more benefits than obligations. As a result, on the rare occasions they are faced with an event that may cause them frustration. A classic clinical study by Barcai and Rosenthal (1974) considers a tyrannical child whose temperament is influenced by a sense of grandiosity and self-entitlement. As a consequence, their demands increase and they adopt tyrannical behavior (Garrido, 2007). These tyrannical behaviors appear to be a habitual pattern in subsequent social relationships (Gelles, 1997;Harbin & Madden, 1979;Pagani et al., 2009). The effectiveness of aggression helps to create a tendency to dominance, which becomes universally applied across multiple social contexts (Patterson, 1995). In these cases, the parents have not been able to establish a limit on the child's emotional blackmail, and so the child increasingly makes greater demands accompanied by greater violence, and gains power and control over the family (Coogan, 2011). This type of upbringing without any limits fosters self-centeredness, a lack of empathy and low tolerance for frustration when the children do not immediately get what they want (Etxebarria et al., 2009;Urra, 2006). All this is influenced by an environment in which a great deal of time is spent away from home. There is little dialog, and the communications that take place may trigger violent situations or reinforce indulgent styles to avoid confrontation Urra, 2006). In fact, according to Baumrind (2005), children with nonauthoritarian-directive parents were less distressed and more competent than children with authoritariandirective parents. This author concluded that intrusiveness and low levels of parental support were associated with maladjustment, rather than high levels of behavioral control.
It is considered necessary to try to modify the dysfunctional family patterns that have allowed the emergence and maintenance of violent behavior (Pereira, 2020). The family history, family structure, and the educational styles and patterns used, will be crucial factors to take into account when designing the therapeutic intervention, specifically for each case (Pereira, 2020). The detection of risk factors through the YLS/ CMI on a Spanish population may play a critical role in predicting CPV in juvenile populations. The identification and detection of a particular need related to a specific type of crime could be useful for intervention purposes, thus increasing the appropriateness of the programs. This may be useful to regulate the intensity and urgency of the supervision.
A limitation of our study was our reliance on Juvenile Court data, which does not enable general inferences to be made for the general population of minors, less serious cases, or different penal systems of the different countries. Our data were also gathered from one specific Spanish province, and as such conclusions should not be generalized to other Spanish territories or other countries. The information needed to complete the YLS/CMI is compiled from several sources, including an interview with the youth and their family, their previous charges, social services involvement, educational institutions attended, and so on. Hence, all the assessment was performed without a formal diagnosis which may limit the conclusion of the study. Another important issue within this sample is the unwillingness to report assaults due to feelings of shame and guilt, as discussed in the introduction section, which makes ascertaining the scale of the problem with precision difficult. Even if the intention of this research was to understand deeply this phenomenon, causal explanations are a complex issue. It cannot be concluded that a single factor is responsible for this type of violence, but rather a complexity of factors that interact with each other to a greater or lesser extent. Trying to do so would under-estimate the complexity this social problem (Moulds et al., 2019a).
As a future line of research, it would be necessary to clarify precisely which patterns of upbringing lead to this violence, and to determine the role of attachment styles in this type of aggression related to parental relationships (Ibabe & Jaureguizar, 2010;Laible et al., 2004). This could be done with the use of qualitative and quantitative mixed method, which could yield deeper insights into the experiences of parents. This will provide a better and more accurate picture of CPV, since bonding may play an important role in this crime. Another line of research could also cover a specific analysis of the role of gender. It may be important to analyze the predictive variables in the study when mediatized for gender, and to analyze the factor of Exposure to violence in-depth, according to whether the violence is experienced directly or observed.

Conclusion
As a final conclusion, the factors of personal variables, antisocial personality and exposure to violence may play a role in the commission of this type of crime, although the latter are determinant when family characteristics are compared with parenting. As a result, what affects the commission of this type of violence in the most serious cases, that is those discussed in court, is not having been exposed to it, but instead what seems to be decisive are parental guidelines and the offender's current relationship with the parents (Beckmann, 2020). In other words, it seems that the results do not support the bi-directionality of violence hypothesis. When the juvenile has been exposed to family violence, but the bond can be restored, there may be a decrease in this type of violence (Juan et al., 2020). And consequently, perhaps low self-esteem and traits related to depression could also be improved.
A key factor in interventions to modify this type of aggression would be to train the parents on how to regain their power and control over their children's behavior, and how to reinforce their ability to establish norms and boundaries Calvete et al., 2011). Training on preventing emotional deficits and coping with traits related to an antisocial personality would also be useful. Helping to modify attitudes toward violence and engender better self-control in individuals is therefore positive in preventing crime, and violent offenses in particular. However, external and internal factors must be taken into account when predicting the wide range of CPV, since the commission of a crime involves both, in order to help to design appropriate interventions for these offenses.

Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author disclosed receipt of the following financial support for the research, authorship, and/ or publication of this article: