Various studies have reported that children can present externalizing and internalizing behavior problems and difficulties of psychosocial adjustment as a result of dysfunctional parenting styles (Alt, 2016; Chen et al., 2015; Franco-Nerín, Pérez-Nieto, & De Dios-Pérez, 2014). Due to this, it becomes necessary to have a valid and reliable instrument for the Colombian population, that allows the assessment of parental styles in order to predict such behavioral problems in children.

In the Colombian context, there has been a change on the structure and family dynamics. Currently, up to 14 % of families are made up of extended families, and up to 10% are made of single families, while it is also noticeable a decrease in nuclear families by up to 30 % based on previous years. This in turn has generated changes in parental styles (Ministerio de Salud, Colciencias, & Pontificia Universidad Javeriana, 2015).

There have been some international studies that have evaluated the psychometric properties of instruments on parenting styles (Ang, Tan, & Fam, 2016; Chakawa & Hoglund, 2016; Kerns, Eating, & Zeman, 2014; Young et al., 2013). These authors also concluded that the factorial structure of a categorical type of parenting styles based on Baumrind’s taxonomic perspective (1971) does not provide enough clarity about parental styles; whilst a dimensional conceptualization of parental styles based on the numerical dimensions of warmth and nurturing behavior management can more accurately capture the construct.

In Spain, Gómez, Martínez de Salazar, García Barranco, Cobos and Filippo (2012) conducted a pilot study of the Educational Parenting Styles Questionnaire (CEEP), They aimed to identify parenting styles, that can be risk or protective factors for clinical problems in children aged between 6 and 12. This instrument showed high reliability of 0.92, general content validity kappa index of 0.92, and construct validity in each of the five scales with factorial structures whose percentages explained between 56.5 to 65.05 of variance. Gómez et al. (2012) suggested to extend the CEEP psychometric analysis to a representative sample of normal population in order to have greater predictive validity and an overall higher level of reliability.

In Latin America, other researchers have also evaluated the psychometric properties of parental styles instruments (Ballesteros, 2001; Manrique, Ghesquière, & Van Leeuwen, 2014; Molina, Raimundi, & Burgallo, 2017). The instruments reviewed assess certain characteristics about parenting, however, none of them allow for sufficient determination of the capacities required by parents in order to meet the needs of their children, what highlights the need to standardize those instruments with larger samples and refine their measurements.

Several studies have focused on studying the prediction of externalizing and internalizing behaviors from parental styles, albeit with diverse results. Fuentes, Salas, Bernedo and García-Martín (2014), found that the authoritarian style was related to thisbehaviors; Tan, Gelley and Dedrick (2015) found that permissive and authoritarian styles were associated, and Lakin, LeRoy, Boxer, and Mahoney (2014) concluded that a negative parental alliance predicted further internalization problems.

Evidence has been found of how the same component of parental style can be a predictor of different internalization and externalization problems in children, for instance, the use of physical punishment (Chen et al., 2015), lack of communication (Cerezo, Ruiz-Esteban, Sánchez-Lacasa, & Arense-Gonzalo, 2018; Fite, Rubens, Preddy, Raine, & Pardini, 2014), inconsistency and ambiguity in the rules and limits (Jabagchourian, Sorkhabi, Quach, & Strage, 2014). Contrarily, some of these might be protective factors for those problems, such as cohesion, promotion of positive behaviors, emotional competencies, establishment of firm limits, communication and affection (Cenk & Demir, 2016; Gómez et al., 2012; Lakin et al., 2014; Newland, 2015).

Therefore, the theoretical position on paternity that underpins the present study is proposed by Duncan, Coastworth and Greenberg (2009). From a dimensional perspective, the aforementioned authors emphasize the conscious attention on the part of parents towards their children. From this model, it is argued that the quality of relationships between parents and their children can be strengthened when parents become aware of the upbringing of their children.

In light of the above, the objectives were as follows: Firstly, to perform a psychometric analysis of the Educational Parenting Styles Questionnaire for a Colombian population. Secondly, to evaluate the predictive validity of the questionnaire on parental educational styles on internalizing and externalizing psychological problems in Colombian children.

Method

Participants

The sample included 680 parents with a mean age of 37.34 years (SD = 9.02) of children between 8 and 12 years old. 70.1 % of the parents were mothers, whereas only 29.9 % were fathers. 35.3 % were in common law marriage, 27.7 % were single, 21.6 % were married, 5.9 % were separated and, to a lesser extent, 0.9 % were widowers. With regard to occupation, 54.6 % of the parents were employed, 16.8 % of the father and 23.1 % of the mothers were employed while 14.1 % of the mothers were housewives, none of the fathers were stay-at-home dads, and 10 % were independent. The participants belonged to a low social economic status, and were selected for convenience.

Instruments

Parental Educational Styles Questionnaire (CEEP - Reduced version; Gómez et al., 2012). This instrument has five scales with corresponding factors as follows: Parental Practices (Promotion of positive behaviors, dysfunctional reactions to disobedience, inconsistency, proper use of corrections); Educational Styles (Democratic, permissive, ambiguous, authoritarian); Emotional Competences (Impulsiveness, emotional expression, recognitions of emotions, emotional management); Parental Role (Parental satisfaction, collaboration with school, parental overload) and Family Dynamics (family cohesion, communication, shared leisure, conflicts). The questionnaire has of 66 Likert items with four answer options (1 most of the times, 2 frequently, 3 sometimes, 4 rarely). According with Cronbach’s Alpha, the total of the questionnaire showed a high reliability of 0.92 while the subscales between medium and high between 0.64 and 0.84; high content validity with an overall Kappa index of 0.92. Regarding construct validity for all factors and scales, assumptions of sphericity, adequacy ratio of the Spanish sample and the value of the determinant have been optimally met.

Child Behavior Checklist (CBCL), parent format. Achenbach and Edelbrock’s (2001) The CBCL allows to evaluate internalizing and externalizing psychological problems in children aged 4 to 16. This checklist has a rating as follows: (0) if it is not true, (1) if sometimes it is true and (2) if it is true very often. The direct scores are transformed into t scores according to the following ranges: Clinical: between 64 and 100; Borderline: between 60 and 63; Normal: between 59 and 33. Regarding content validity, 116 of the 118 items of behavioral problems are significantly associated with clinical status – p < 0.01. It has been adapted for the Colombian population where it obtained a reliability coefficient Cronbach’s Alpha of 0.83 and an internal consistency of 0.94 (Rescorla et al., 2012).

Procedure

The study was approved by the Ethics Committee of the University to which the authors belong. School principals were asked to authorize access to the participants. The research objectives were then explained to parents in order to obtain an informed consent to participate in the study. Subsequently, the instruments were administered to the parents collectively in the school by 6 psychologists who were previously trained by the researchers on the test and its administration. The resulting data was analyzed, and the institution was given the results of the research. Parents were given psycho-educational primers to strengthen parenting styles with their children for participation in the study.

Statistical Analyses

Based on the contribution of each original variable in each of the factors, a new model fitted to the Colombian population was obtained, and was subjected to confirmatory factor analysis (CFA) using the software AMOS 23. Other indexes were used, such as CMIN / df, which refers to the c2 ratio, over the degrees of freedom, GFI (goodness of global adjustment index), the CFI (comparative fit index) and RMSEA (square error approach to medium roots), AIC (Akaike information criterion). To determine the fit of the data to the model, the value of c2 should be meaningful, the CMIN / df, requires less than 3.0, the CFI and GFI vary between 0 and 1, where 0 means no adjustment and one means optimal fit, values higher than 0.90 represent an acceptable fit; while values of 0.95 or higher are considered excellent. The RMSEA needs to be less than 0.08, and preferably less than 0.06, because if the model is greater than 0.10, it must be rejected (Pilatti, Godoy, & Brussino, 2012).

Correlations between scales of de CEEP and CBCL were analyzed using the Pearson coefficient. Finally, a multiple regression analysis was conducted in order to establish the predictive validity of the CEEP. The parental styles in each of the dimensions were used as predictors, and the scales of the CBCL of the internalizing and externalizing behaviors were used as criterion variables.

Results

The results of the confirmatory factor analysis (CFA) presented in Table 1 show that the indexes in the fit model are better than the rates in the original model and the one-dimensional. The resulting model presents the best indicators of favorable fit according to recommendations suggested by CFA. These values can show an internal consistence that also favors results for the scales of the instrument.

The fit model founded for the Colombian population was one-dimensional for family dynamics (c2 = 2.78; df = 3; c2/ df = 0.93; GFI = 0.99; CFI = 1.00; TLI = 1.00; RMSEA = 0.00; AIC = 26.78), two factors for parental role (c2 = 3.95; df = 4; c2/ df = 0.99; GFI = 0.99; CFI = 1.00; TLI = 1.00; RMSEA = 0.00; AIC = 26.78), and parental practices (c2 = 29.59; df = 12; c2/ df = 2.47; GFI = 0.99; CFI = 0.97; TLI = 0.94; RMSEA = 0.04; AIC = 61.59); three factors for emotional competences (c2 = 35.33; df = 19; c2/ df = 1.86; GFI = 0.98; CFI = 0.94; TLI = 0.91; RMSEA = 0.04; AIC = 69.33), and educational styles (c2 = 53.63; df = 31; c2/ df = 1.73; GFI = 0.98; CFI = 0.95; TLI = 0.94; RMSEA = 0.03; AIC = 101.63).

According to Cronbach alpha coefficient, the reliability coefficients found were the following: family dynamics (a = 0.65), parental role (a = 0.53), emotional competences (a = 0.66), educational styles (a = 0.63), parental practices (a = 0.60) and total parental educational styles (a = 0.78).

Table 2 presents the results corresponding to the multiple regression analysis. As derived from the results obtained, dysfunctional reaction to disobedience, family cohesion, inconsistency, communication and conflicts predict internalizing behavior problems. It is important to note that a dysfunctional reaction to disobedience in parents significantly predicts anxious-depressive, isolated-depressive behavior, somatic complaints, and total internalizing behaviors in children. Conflicts is another variable that allows to predict isolated-depressive, somatic complaints and total internalizing behaviors.

In Table 3, the results on the multiple regression analysis show that the dysfunctional reaction to disobedience, communication, conflicts, overload and parental satisfaction predicts externalizing problems. Specifically, the dysfunctional reaction to disobedience and conflicts in parents predict both rule-breaking behavior and aggressive behavior in children.

Discussion

One purpose of this study was to establish the psychometric properties of the Parental Educational Styles questionnaire (CEEP) for a Colombian population. From the confirmatory analysis, it was found that the CEEP model for Colombia had values indicating an optimal fit for all subscales and the total scale. The total scale is superior to the Spanish model, which showed acceptable levels; just as the one evaluated by Manrique et al. (2014) in Peru. The latter had a moderate adjustment, and as well as the instrument analyzed by Penelo, Viladrich and Domènech (2010) which was acceptable. Similar to the research by Sheffield, Waller, Emanuelli, Murray, & Meyer (2006), the AFC, showed a shorter version of the original instrument.

The fit model founded for the Colombian population was one-dimensional for family dynamics, two factors for parental role, and parental practices; three factors for emotional competences, and educational styles. The findings of the present study provide a numerical dimensional perspective of parental styles that permit to capture this construct more accurately that structure of a categorical type of parenting styles (Ang et al., 2016; Chakawa & Hoglund, 2016; Kerns et al., 2014; Young et al., 2013).

The results also show that the instrument allows evaluating with more determination parents skills about parenting in order to meet the needs of their children, as previous studies have indicated (Ballesteros, 2001; Manrique et al., 2014; Molina et al., 2017).

Another purpose was to evaluate its predictive validity on externalizing and internalizing problems in Colombian children. The factors that constituted risk predictors of internalizing behaviors were dysfunctional reaction to disobedience, inconsistency, conflicts, and difficulties in family cohesion and communication. In turn, the results indicate that a dysfunctional reaction to disobedience, conflicts, parental overload, problems in communication and lack of parental satisfaction predicted externalizing behavior problems. One dysfunctional reaction to disobedience, is the use of punishment by parents which has an impact in the manifestation of externalizing behaviors in children. This can be explained by the association between coercion and imposition, when they are not accompanied by dialogue and affection have an effect more harmful than adaptive (Uji, Sakamoto, Adachi, & Kitamura, 2014).

When conflicts arise within the family characterized by the existence of discussions and lack of communication among their members, they predict internalizing behavior patterns, as well as externalizing behavior in children (Estévez, Jiménez, & Moreno, 2018; Livas & Polo, 2014; Fernández-Daza & Fernández-Parra, 2012; Percy, Creswell, Garner, O’Brien, & Murray, 2016).

For their part, Pereira, Negrao, Soares and Mesman (2013) point out that paternal stress and family conflict are predictors of severe discipline in mothers. Additionally, Malczyk and Lawson (2017) conclude that parents who are overwhelmed with paternity tend to control their children less. Both influences have a significant negative impact on the stability and emotional behavior of children. Chappel, Suldo and Ogg (2014) argue that family conflict is inversely related to children's emotional satisfaction and safety.

Communication has been the protective factor encountered for both externalizing and internalizing behaviors. This can be explained by families that have firm parenting styles, sustained by criteria that instill autonomy for all members and promote adequate, affection-based relationships within the family. These families present the best results in the education of children (Cerezo, et al., 2018; Duncan et al., 2009; Jabagchourian et al., 2014) and it generates between parents and children a high positive effect on the relationship (Gómez et al., 2012; Newland, 2015).

In conclusion, it can be said that this study contributes to having a reliable and valid instrument for the dimensional evaluation of parenting styles with a Colombian sample. This is underpinned by the fact that most of the studies have been conducted with North American, Asian and European populations (Alt, 2016; Budd, Danko, Li, & Henriquez, 2015; Cenk & Demir, 2016; Kerns et al., 2014; Young, et al., 2013).

This research has important clinical implications at the level of the evaluation of parental educational styles in Colombian population and their impact on the internalizing and externalizing behavior problems in the period of middle childhood. Research also solves one limitation of the study carried out by Gómez et al. (2012), wherein data was provisional for the negligible sample; in addition, the CBCL scales were ultimately deemed invalid.

At the same time, this paper provides information of interest for the prevention of psychological problems in children and for parent counseling on childrearing. This is due to the fact that different family dynamics and parental roles were identified, as were styles and practices that constitute risk factors / protective factors for the behavioral and emotional problems of children.

Limitations and Future Directions

It was found that this instrument has not been used with clinical samples, and the findings can only be generalized for a sample with characteristics similar to the one evaluated in the present study. In this regard, it is important to extend this study with samples from different socioeconomic status and cultures and with different clinical characteristics because parental styles could have a different predictive role in European, Asian or North American countries where the rules and regulations are set differently from Colombia. Likewise, it is suggested to include another instrument to account for the association of parental styles with the adaptive behavior of children on a personal, family, school and social level, given the implications of the styles of aging on children’s different areas of adjustment.

Acknowledgments

The authors sincerely thank the participating parents for their cooperation with this project.

References

Notes

* Research article.

Author notes

^a Corresponding author: E-mail: jaimehm@yahoo.com

Additional information

How to cite: Moreno Méndez, J. H., Espada Sánchez, J. P., & Gómez Becerra, I. (2019). Advances on parental educational styles predictors of internalizing and externalizing behaviors in children. Universitas Psychologica, 18(5), 1-10. https://doi.org/10.11144/Javeriana.upsy18-5.apes

Descargar