Introduction

Previous research has studied the association between maternal and paternal parenting styles and found low levels of similarity between parents’ parenting styles [1, 2]. However, the research suggests that dissimilarity between parenting styles is an important factor in the development of emotional and behavioral problems in children. For example, Dwairy [3, 4] investigated the effect of dissimilarity in parental authoritarianism on the psychological well-being of adolescents and found that parental dissimilarity was associated with higher scores on a global measure of psychological problems. In addition, Jaursch et al. [5] found that parental dissimilarity in emotional warmth and rejection was correlated with children’s internalizing and externalizing problems. Findings of Lindsey and Caldera [6] also stress the importance of similarity in parenting styles; their results reveal that more similarity between parents in the use of control is associated with more compliant behavior in children.

Thus, the dissimilarity between parenting styles is associated with mental health problems in children. What mechanisms possibly play a role in this association? One possible mechanism is that children who perceive their parents as dissimilar have more trouble predicting their parents’ behavior. Difficulties in predictability have been associated with children’s depression and anxiety symptoms [7]. Unpredictability of the parents’ behavior constitutes a risk for children to develop learned helplessness, which is a chronic belief that external forces and persons determine one’s destiny [8, 9], and is associated with internalizing and externalizing problems [10, 11]. See Ross and Hill [12] for an extensive description of the literature on family unpredictability.

In previous research on the effects of parenting there are a couple of limitations that we will address. First, the majority of previous research on the effects of parenting on children is focused on maternal parenting styles, because mothers are primarily responsible for the child-rearing process. However, fathers increasingly take part in family life [13], and therefore, fathers’ roles in parenting become increasingly important. Because of this increased involvement of fathers, the relationship between both parents’ parenting styles also becomes more important in studying the effects of parenting [13]. Thus, instead of studying maternal parenting alone, paternal parenting and the relation between maternal and paternal parenting styles should be taken into account. There seems to be a gap on this topic in the existing literature.

Second, studies involving parenting styles differ in the methods of assessing parenting. In research based on questionnaires, parents’ as well as children’s perceptions of parenting can be assessed. Several researchers studied the association between parent and child reports of parenting, and found that these were only low to moderately correlated [14, 15]. Frampton et al. [16] found that children’s negative perception of the relationship with their mother predicted internalizing problems. Moreover, Demo et al. [17] found that children are influenced by their perceptions of parental behaviors, rather than by actual parental behaviors or those reported by the parents. Barry et al. [14] and Tein et al. [15] therefore suggest not denying the inclusion of children’s perception of parenting in examining effects of parenting.

Two important statements follow from the results of previous studies. First, higher levels of dissimilarity in parenting are associated with emotional and behavior problems in children [5, 6]. Second, children are important informants on parenting styles, because previous research has shown that children’s perceptions of parenting are associated with problems [16, 17]. These two statements support research on dissimilarity between parenting styles based on the perception of children.

To date, a limited number of studies have been conducted on the children’s perception of dissimilarity between their mothers’ and fathers’ parenting styles. Dwairy [3, 4] examined the effect of parental inconsistency in parenting and psychological disorder symptoms. However, these studies cover only one specific aspect of parenting (i.e., authoritative parenting), whereas other parenting styles are associated with children’s problems as well. Johnson et al. [18] found that perceived dissimilarity in parenting styles characterized by emotional warmth and control was associated with lower self-esteem, school adaptation, and school achievement. These results showed associations between dissimilarity in parenting and negative outcomes in children. However, research on the effects of dissimilarity in parenting should be expanded to examining the association between perceived dissimilarity in several parenting styles and the effect on internalizing as well as externalizing problems. In the current study we examine dissimilarity in functional (emotional warmth) as well as dysfunctional (rejection and overprotection) parenting styles, and the relationship with problems in children.

The aim of the current study is to investigate the relationship of perceived dissimilarity in parenting with mother-reported internalizing and externalizing problems in children. We examine children’s perception of dissimilarity in parental emotional warmth, rejection, and overprotection, and mothers’ reports on internalizing and externalizing problems are used. We hypothesize that lower levels of perceived maternal and paternal emotional warmth, and higher levels of perceived maternal and paternal rejection, and overprotection are associated with higher internalizing and externalizing problem scores. Higher levels of perceived dissimilarity in emotional warmth, rejection, and overprotection are hypothesized to be associated with higher levels of internalizing and externalizing problems in children.

Methods

Participants

The present study is part of a longitudinal study that started in 2003, in which 1,710 families participated, with children in the age between 6 and 18 years and living in the Dutch province of South Holland. For details on the initial data collection, see the work by Tick et al. [19].

Of the participants of time 1, we selected 1,160 children to participate in the present study (time 2). These children were selected based on their birth dates (between 1 January 1988 and 31 August 1997), their participation at time 1, and their permission to be contacted for follow-up research. This selection based on birth dates resulted in a sample of children between the ages of 8 and 20 years. The reason for this selection was the age range of the questionnaires that were used. Of these 1,160 children, 216 children refused participation and 15 were excluded because of a severe illness or handicap, or because they live outside the Dutch province South Holland. A total of 271 participants agreed to participate, but did not fill-out one of the questionnaires that are used for the current paper (e.g., one-parent families).

The analyses were performed on the information of the remaining 658 participants. The mean age of the 355 girls was 13.7 years and the mean age for the 303 boys was 13.4 years. The median age is 12 for both boys and girls. The national origin of the child is based on the country of birth of the parents. In accordance with Statistics Netherlands [20] we classified a child as non-western if one of the parents (and/or the child itself) was born in a non-western country. Almost 90% of the children have a western national origin, and 10% has a non-western national origin. Information on parental occupational level was derived from questionnaires. The occupational level is categorized into five categories from 1 (low occupation), to 5 (high occupation), following the definition of Statistics Netherlands [21]. The mean occupational level is 3.1. Approximately 76% of the mothers and 97% of the fathers were employed.

When analyzing the attrition, results showed that participants and non-participants did not differ in their internalizing scores at time 1 (t = −0.38, NS). However, they differed in the externalizing scores; children who did not participate at time 2 had higher externalizing scores than children who participated (t = 3.85, p < 0.001). The mean age of the children who participated was lower than the mean age of the children who did not participate (t = 2.03, p < 0.05).

Procedure

The parents received a telephone call in which the study was explained and in which they were asked to participate. The families, who agreed to participate, were invited at the outpatient department Child and Adolescent Psychiatry and Psychology of the Erasmus Medical Center. During their visit to the outpatient department, children and their parents were in separate rooms, assuring that they filled out the questionnaires independently. Furthermore, the confidentiality of their responses was stressed. When families did not agree to visit the outpatient department, but did agree to participate, they received the questionnaires in home by mail, and were asked to send them back. Parents and adolescents received a gift-coupon for their participation and children received a present.

Instruments

Parenting styles: We used the Egna Minnen Beträffande Uppfostran-child version questionnaire (EMBU-C, my memories of upbringing) [22] for assessing parenting styles. The EMBU-C was adapted from the originally developed EMBU [23], to assess the children’s perception of the parents’ rearing behavior. The EMBU-C consists of 57 questions. Markus et al. [22] examined the factor structure of the EMBU-C resulting in four scales: emotional warmth, rejection, overprotection, and favoring subject. There were 19 items included in the emotional warmth scale. The main concepts of this scale were: giving special attention, praising, unconditional love, and being supportive and affectionate. For rejection, there were 17 items included. This scale is characterized by hostility, punishment, derogation, and blaming of the subject. There were 12 items that were included in the dimension overprotection. The aspects fearful and being anxious for the child’s safety, guilt engendering, and intrusiveness were represented in this scale. We did not use the favoring subject scale, due to an internal consistency below 0.60. Each item of the EMBU-C is scored on a 4-point scale (1 = no, never, 2 = yes, sometimes, 3 = yes, often, and 4 = yes, always), and is answered for the father and the mother separately. Cronbach’s alphas for the scales ranged from 0.67 to 0.92.

Internalizing and externalizing behavior: We used the child behavior checklist (CBCL) [24] to assess internalizing and externalizing behavior in children. The CBCL contains 120 items, covering children’s emotional and behavioral problems in the preceding 6 months. The response format is 0 = not true, 1 = somewhat or sometimes true and 2 = very true or often true. The internalizing scale (anxiety, depression) consists of 32 items, and the externalizing scale (aggression and rule-breaking behavior) consists of 35 items. We used raw scores in the analyses of the current paper. Good reliability and validity of the CBCL have been reported for the Dutch version [25]. Cronbach’s alpha for the internalizing scale was 0.86 and for the externalizing scale 0.87.

Data analysis

First, we will give descriptive statistics, such as means, standard deviations, and percentages of the variables. We will also give demographic information, such as occupational level and employment of the parents, and national origin of the child.

To obtain a measure for dissimilarity between parents’ parenting styles, we computed intraclass correlations (ICCs) between the children’s responses on the EMBU-C about their mother and father [26, 27]. The ICC describes how similar children perceive their parents’ parenting styles. An ICC takes account of both level differences and score mismatch. For each EMBU-C scale and for each child, an ICC was calculated between the responses for his father and his mother. The ICCs for the three EMBU-C scales were used as continuous predictor variables in the regression analyses with internalizing and externalizing problems as outcome measures. The ICC scores range from −1 to 1, with 1 indicating perfect perceived similarity between the mother and father of a child, and −1 indicating perfect dissimilarity. Higher ICCs indicate that children perceive their parents as more similar [26, 27]. However, for ease of interpretation, we mirrored the ICCs in the analyses. As a result, higher ICC values reflect higher perceived dissimilarity between the parents.

To examine whether dissimilarity in parenting styles is associated with higher levels of internalizing and externalizing problems in children, hierarchical multiple regression analyses were conducted with the ICCs of the three parenting styles as independent variables, and mother-reported internalizing and externalizing problems in children as outcomes. In the first step, fathers’ and mothers’ parenting scores were entered into the model. In the second step, the ICCs were added. In this way, the results of the regression analyses will be corrected for the maternal and paternal parenting scores. We investigated whether dissimilarity in parenting styles has an effect on internalizing and externalizing problems in addition to the level of the maternal and paternal parenting styles. Thus, significant effects of dissimilarity in parenting on internalizing or externalizing behavior will exist over and above the effect of the individual parenting styles of the parents.

Previous research has shown age differences in perceived parenting and in perceived dissimilarity in parenting; adolescents perceive less authoritative parenting and more dissimilarity in parenting than pre-adolescents [18]. Sentse et al. [28] found gender differences for boys and girls. Boys perceive more overprotection and rejection than girls, and girls perceive more emotional warmth than boys. Therefore, in all analyses, age and gender were included as covariates.

Results

The mean and standard deviation of internalizing and externalizing scores, perceived parenting, and perceived dissimilarity in parenting are displayed in Table 1.

Table 1 Means, standard deviations (SD), and ranges of problem behavior, perceived parenting, and perceived dissimilarity in parenting (n = 658)
Full size table

Correlations between the variables are displayed in Table 2. The results of the correlation analyses revealed that rejection and overprotection are correlated with internalizing problems and that all parenting styles are correlated with externalizing problems. Age is negatively correlated with emotional warmth, rejection, and overprotection, and age is positively correlated with perceived dissimilarity in emotional warmth and overprotection. The results further revealed that girls perceive more emotional warmth from their parents than boys, whereas boys perceive more maternal rejection than girls. Girls had higher internalizing problem scores than boys. There were no gender differences in perceived dissimilarity in parenting.

Table 2 Correlations among perceived parenting, problem behavior, dissimilarity in parenting, age, and gender (n = 658)
Full size table

Perceived dissimilarity in emotional warmth and perceived dissimilarity in overprotection are significantly associated with internalizing and externalizing problems. Perceived dissimilarity in rejection is associated with internalizing problems. Higher levels of dissimilarity are associated with higher problem levels.

The question is, do these effects of perceived dissimilarity in parenting on problems behavior in children, exist over and above the effect of perceived parenting? Hierarchical multiple regression analyses were performed with internalizing and externalizing scores as outcomes and perceived maternal and paternal parenting styles entered in the first step. In the second step the ICCs for the parenting styles were added. Age and gender were entered in the model as covariates. The results are presented in Table 3.

Table 3 Results of regression analyses of perceived maternal and paternal parenting and dissimilarity between parenting styles, associated with internalizing and externalizing problems, adjusted for age and gender
Full size table

Results revealed that higher levels of perceived dissimilarity in emotional warmth were associated with higher levels of internalizing as well as externalizing problems, over and above the effect of the parents’ individual levels of emotional warmth. Perceived dissimilarity between parents in the level of rejection is associated with neither internalizing nor externalizing problems. The association between perceived rejection and externalizing problems remained significant. Higher levels of perceived dissimilarity in overprotection were associated with higher levels of externalizing problems, over and above the effect of the parents’ individual levels of overprotection. The range of R 2 coefficients suggests that the regression models explained between 2.3 and 8.7% of the variance in internalizing and externalizing problems.

Discussion

The results of the current study indicate that higher levels of perceived dissimilarity between maternal and paternal parenting have a negative effect on children’s emotional and behavioral well-being, over and above the effect of parents’ individual parenting styles on children’s well-being. Children who perceive more dissimilarity in emotional warmth have higher levels of internalizing problems and children who perceive more dissimilarity in overprotection have more internalizing and externalizing problems. Perceived dissimilarity in rejection was not significantly related to internalizing or externalizing behavior problems. Our results partly support the hypothesis that there is a relation between perceived dissimilarity in parenting and psychopathology in children. However, the data do not support all hypotheses. Striking in the results of our study is that dissimilarity in dysfunctional as well as in functional parenting has negative effects on children.

Gruner et al. [29] found that rejection and anxiety are associated, which is partly comparable with our study in which rejection is associated with externalizing problems. However, in our study, we did not find an association between dissimilarity in rejection and externalizing problems in children. This is striking, because of the clear associations that have been found between rejecting parenting and psychopathology [7, 14]. A possible explanation could be that rejection may have a powerful influence on children, regardless of whether one or both parenting styles are perceived as rejecting. There is little additional effect of dissimilarity in rejection on children’s outcomes, over and above the rejecting parenting style of one of the parents.

A couple of limitations of our study can be addressed. First, there is little variance in the level of perceived similarity. Over 50% of the parental dyads have high ICCs (i.e., ICC higher than 0.80), and approximately one-third of the children perceive perfect similarity between parents. In previous studies low to moderate agreement between parenting styles has been found. For example, Winsler et al. [1] reported correlations from −0.07 to 0.51 between mother- and father-reported parenting. The low variance in current study makes it difficult to detect differences in the sample. To further examine the association between perceived similarity in parenting styles and psychopathology in children, it may be useful to use a sample of children in which one can expect more variance, like a clinical sample or a high-risk group.

Second, children who did not participate in the current data collection wave of the study (time 2), have higher externalizing scores at the first data collection wave (time 1) than children who participated. The results regarding externalizing problems should be treated with caution.

Third, a limitation of the study is the wide age-range of participating children. Although the questionnaires are designed for a broad age-range, one cannot be sure that children of age 8 are as capable in reporting on their parents’ parenting styles as 20-year-olds. Johnson et al. [18] found furthermore that adolescents perceive more dissimilarity in parenting than pre-adolescents, which is in accordance with the results of our study, in which age is positively correlated with dissimilarity between parenting in emotional warmth and overprotection. Therefore, we corrected for the age of the children in the analyses. Several studies found that children benefit from congruent parenting [3–6], However, Mc Kinney and Renk [30] examined similarity in parenting from the perspective of late adolescents (aged 18–22 years) and found that perceived similarity in parenting alone is not related to beneficial outcomes. Therefore, Mc Kinney and Renk [30] suggest that younger children benefit from similarity in parenting, despite the type of parenting style that is used by fathers and mothers, whereas late adolescents do benefit from a supportive parenting style, but are not as much influenced by dissimilarity between the parenting styles.

In studying the association between perceived parenting, and internalizing and externalizing problems, the following question raises. Do children with problems perceive their parents as displaying less-favorable parenting styles, or do their parents objectively have less-favorable parenting styles? Yahav [31] studied the relationship between children’s internalizing and externalizing problems and their perceptions of parenting behavior. Siblings without symptoms were included as a control group in an attempt to differentiate between subjective perception and parenting style. Children with problems and their siblings were asked to report on their perceptions of parenting toward themselves and toward their sibling. The findings suggests that children with problems tend to see their parents as generally negative toward themselves and their siblings. The opposite seems to be true for the healthy siblings, who perceive their parents as more positive towards themselves and their siblings. It is possible that this is also the case in current study on dissimilarity in parenting, and internalizing and externalizing problems in children. Possibly children with higher levels of internalizing and/or externalizing problems may perceive higher levels of dissimilarity between parents, while their non-symptomatic siblings would not report this. In further research on this topic, it would be recommended to include non-symptomatic sibling as controls or find another way to objectify the perceived parenting. On the other hand, as Demo, et al. [17], as well as Yahav [31] suggested, children seem to be more affected by their perceptions of parental behaviors, than by actual parental behaviors or those reported by the parents.

It is commonly assumed that parenting influences the well-being of children. Therefore, in clinical work with children with different kinds of problems, clinicians already focus on the parents as well as on the child itself. In several treatment programs, as for example the FRIENDS treatment-program for children with anxiety disorders, attention is paid to improvement of parenting practices [32, 33]. However, little attention is paid nowadays to the degree of dissimilarity between the parenting styles, which is, as the results of our study reveal, also important. We therefore recommend paying attention to the dissimilarity between parenting styles in treatment programs for children with internalizing and externalizing problems. The results of the study highlight the importance of the perception of the child of dissimilarity between parents. Therefore, it is recommended to assess children’s perception of dissimilarity between their parents, before treatment starts. If the child reports high levels of dissimilarity, the clinician should try to lower this dissimilarity in the therapy sessions for parents. For establishing this goal, it is important that both parents are encouraged to attend to the therapy and therefore, it would be recommended to give the opportunity to visit the departments in evenings or weekends. When the therapist succeeds in having both parents in the therapy sessions, attention can be paid to dissimilarity between the parents by means of discussions and role-plays. In this way, parents will become more aware of the differences between them and of the importance of similarity in parenting.

To conclude, the results of the study highlight the negative consequences of perceived dissimilarity in parenting styles between parents; children have higher problem scores when they perceive their parents as more dissimilar in their parenting styles.