Introduction

Parents each have their own parenting styles and beliefs, and often these styles differ between parents, which can lead to conflict (O’Leary and Vidair 2005). Marital conflict, defined as the presentation of negative hostile emotions from parents and poor conflict-resolution strategies between parents, has negative effects on child well being (O’Leary and Vidair 2005). Examining mothers’ and fathers’ perceptions of their own and their partners’ parenting styles as they relate to parental marital conflict and child behavior problems is beneficial to understanding the relationship between these constructs.

Baumrind’s classic model of parenting styles delineates two trait axes: demandingness and responsiveness. Baumrind (1991) proposed four parenting styles, namely, authoritative, authoritarian, permissive, and rejecting-neglectful. These parenting styles relate to child outcomes (Aunola and Nurmi 2005; Darling and Steinberg 1993; Milvesky et al. 2008; Roopnarine et al. 2006). Generally, when mothers and fathers use more democratic, authoritative techniques of parenting, their children tend to experience more positive outcomes (Roopnarine et al. 2006), however parents who are more permissive in their parenting styles tend to raise children who are typically lower in social skills and maturity (Baumrind and Black 1967; Milvesky et al. 2008). From previous research, we also know that when mothers and fathers have similar supportive parenting styles, child outcomes tend to be more beneficial and the quality of overall parenting tends to be better (Meteyer and Perry-Jenkins 2009; Martin et al. 2007). For example, when both parents are supportive at age two, children’s skills in language and math are stronger at age five (Martin et al. 2007).

Bronfenbrenner’s bioecological model, which focuses on proximal processes, can be applied within family dynamics. Parents become part of the microsystem, and have the ability to affect their children’s outcomes (Bronfenbrenner and Morris 2006). More positive parenting styles, coupled with higher parental monitoring tends to lead to better outcomes for children across social and behavioral domains. Parenting styles are also related to more specific parenting practices that are linked with positive and negative outcomes. For example, authoritative parents are more likely to engage in effective scaffolding of their child’s problem solving and foster age-appropriate autonomy in children (Berk and Winlser 1995; Bernier et al. 2010); which from both Vygotskian (Carr and Pike 2012) and self-determination theoretical perspectives (Ryan and Deci 2000) are thought to lead to increased behavioral competence and self-regulation in children. Permissive parents are unlikely to have clear, consistent expectations rules and consequences for children’s behavior (Milvesky et al. 2008), which is known from behavioral and social learning theoretical perspectives (Granic and Patterson 2006) to lead to externalizing behavior problems. Similarly, authoritarian parents are more likely to use overly strict, harsh, and punitive disciplinary practices (Milvesky et al. 2008), which can encourage child aggression according to social learning theory (Chang et al. 2003).

Although there are clearly bidirectional transactional relations between the parent and child, with child behavior and characteristics eliciting certain type of parenting behaviors (Pearl et al. 2014; Zadeh et al. 2010), both longitudinal research and parenting intervention studies show that the direction of parental behavior affecting child behavior is stronger than child effects on parenting style (Choe et al. 2013). Margolin et al. (2001) found that when parenting was characterized by misperception of the partner’s parental styles this led to more marital conflict and poorer parenting. A number of other studies suggest that child behavioral outcomes are driven by parental behavior, be it marital conflict (Schoppe-Sullivan et al. 2007), child rearing disagreements (Chen and Johnston 2012), or differences in parenting styles (Martin et al. 2010).

An important aspect of parenting is the relationship between parents. Children are often exposed to marital conflict and this can dramatically impact their parenting style, parenting behaviors, and the parent–child relationship (Shamir et al. 2001). Marital conflict has been proposed to increase instability in the socioemotional context of the family, resulting in ineffective and inconsistent parenting practices, which in turn leads to child behavior problems (Linville et al. 2010). Children tend to process marital conflict in two ways: they evaluate threat in the environment and attempt to resolve the reason behind the conflict. Ineffective parenting styles can also serve as a mediator between marital conflict and maladaptive child outcomes. That is, the stress and emotional disruption of inter-parental conflict taps parents’ emotional resources, not allowing them to engage in the effective parenting and disciplinary strategies that they would have otherwise implemented, which then leads to child behavior problems (Schoppe-Sullivan et al. 2007).

The Emotional Security theory proposed by Cummings and Davies (2002) suggests that when a child is exposed to destructive parental conflict, this leads to feelings of emotional insecurity, which threatens the child’s confidence and increases the risk for child psychological problems. This insecure parental relationship is hypothesized to mediate the relation between a child’s exposure to marital conflict and child behavior problems (Cummings and Davies 2002).

Chen and Johnston (2012) argued that child rearing disagreements (which reflect the mother’s and father’s perceptions of their disagreement) among parents uniquely predicted child behavior problems, even after controlling for parenting effectiveness. Parenting disagreements on child rearing goals were a stronger predictor of child externalizing behavior problems than child rearing behaviors (Chen and Johnston 2012). Higher levels of parenting conflict also relate to higher levels of externalizing behaviors in children (Schoppe et al. 2001). O’Leary and Vidair (2005) found that parent disagreement over childrearing and over-reactive parenting are mediators between marital adjustment and child behavior problems (O’Leary and Vidair 2005).

Unfortunately, most of the research in this area has been conducted with only mother report, neglecting the important role of fathers. Clearly, to fully understand links between parenting styles, marital conflict, and child behavior, the perceptions of fathers need to be included. Milvesky et al. (2008) found an effect of paternal authoritative parenting style on self-esteem and life-satisfaction in their children. Vera et al. (2012) determined paternal psychopathology was positively associated with their children’s antisocial behavior with parenting style serving a mediating role. Roopnarine et al. (2006) showed that fathers who were more authoritative tended to have children who had more positive social and behavioral outcomes. When mothers score low on supportiveness, and fathers score high on supportiveness, children tend to perform better on school readiness measures (Martin et al. 2010). Father’s supportiveness generally contributed to child’s social competence, but only when maternal supportiveness was slightly below average (Martin et al. 2010). These authors explained that an important contribution of fathers is compensating for what mother’s are lacking in their supportive parenting skills.

Rinaldi and Howe (2012) examined how parenting styles jointly and uniquely predict child behavioral outcomes. Authoritative parents tended to have spouses who were fairly authoritative themselves, and parents’ rating of themselves and their spouses were usually similar. Similarity in parenting style may be a key component of having healthy family relationships (Acitelli et al. 1993) by encouraging marital harmony and supporting and reinforcing similar parenting decisions. Additionally, the combination of maternal permissiveness and paternal authoritarianism was particularly predictive of children’s externalizing behavior problems (Rinaldi and Howe 2012). They conclude that certain pairings of parenting styles by mothers and fathers may result in better or worse child outcomes. Rinaldi and Howe, however, did not include a measure of marital conflict and they did not consider directionality of differences between parents (i.e. when mothers are more authoritative than fathers, and vice versa).

A few studies have assessed both parents’ styles and their perceptions of their partner. Winsler et al. (2005) studied similarities and differences of parenting styles as perceived by spouses. Mothers and fathers of preschool-age children completed the Parenting Styles and Dimensions Questionnaire (PSDQ; Robinson et al. 2001) and they found that parents who self-reported as permissive or authoritarian were likely to be in a relationship with a partner who shared their same parenting style, and they also found that spouses were fairly good at rating each other’s parenting styles (Winsler et al. 2005). While mothers rated themselves as more authoritative than fathers, fathers reported their wives to be more authoritative and permissive but less authoritarian than themselves. Gamble et al. (2007) conducted a similar study with Mexican–American mothers and fathers. They examined similarities and differences between parents on various parenting dimensions using the same instrument (PSDQ). In contrast to the Winsler et al. study, Gamble et al. found more associations between parents on self-reported authoritativeness than on authoritarianism and permissiveness with their Mexican–American sample. These two studies examining perceptions of each other’s parenting styles, however, did not include measures of child behavior or marital conflict as outcomes of parenting styles/patterns. Understanding perceptions of one’s own and one’s partner’s parenting style and how that might differ from the perceptions of one’s partner is important not only for effective parenting, but also for effective marital therapy and couple counseling when couples seek help for marital dissatisfaction (Blais and Renshaw 2014).

The current study evaluates relations between parents’ perceived parenting styles, marital conflict, and child behavior problems. It was expected that increased disagreement in perceived parenting styles would relate to increased marital conflict and child behavior problems. The following research questions were addressed: (1) To what extent are parents self-perceived parenting styles correlated? Are perceptions of their spouses’ style associated with their own self-reported styles? (2) Are disparities between one’s self-reported parenting style(s) and one’s perception of their spouse’s style(s) related to marital conflict and child behavior problems? (3) In the context of disparities between spouse’s perceptions of parenting styles, does directionality matter for marital conflict and child behavior? That is, does it matter, in terms of child behavior and marital conflict, whether the mother or the father is higher or lower than their spouse on a particular style?

Method

Participants

The current study aggregated data from three different samples that contained similar parent self-report data on parenting styles, marital conflict, and child behavior (Madigan 2005; Thorvardarson 2000; Winsler et al. 2005). Merging these three samples was done in order to increase sample size, improve the power of our analyses, and yield better parameter estimates. Since the same or very similar measures were used across the three samples, we combined them to strengthen our analyses.

Sample 1 (Winsler et al. 2005)

The participants of this study were 27 couples (i.e. 27 fathers and 27 mothers) responding about one target child. The children (50 % female, age M = 48 months, SD = 1.71) attended a university-affiliated laboratory preschool in the South that enrolled students from the community, university faculty/staff, and students. Accordingly, a range of socioeconomic levels were represented (Hollingshead M = 49.85, range = 25–66). Families had an average of 2.07 children in the home. The fathers’ age was, M = 37.23 years (SD = 5.22) with an education of M = 17.48 years (SD = 3.0). The mothers’ age was M = 33.57 years (SD = 5.34), with an education of M = 16.26 years (SD = 2.45). As reported in parental reports on preschool enrollment forms, the ethnicity of children was 74.1 % Caucasian, 22.2 % Asian-American, and 3.7 % African-American.

Sample 2 (Thorvardarson 2000)

Participants were 36 fathers and 39 mothers (35 were couples) from an urban and suburban mid-Atlantic region. Participants were included in the study if they met three criteria: (1) Two parents were working full-time, or at least 20 h a week per parent with a total number of hours worked at or above 60 per week, (2) both parents were married or living with a partner of the opposite sex, and (3) there was at least one child and no more than four children living in the home. Adult participants ranged in age from 25 to 50 (M = 39, SD = 6.0 for fathers; M = 36.8, SD = 5.7 for mothers). Ninety percent of participants reported being married while the other 10 % reported living with a partner for an average of 10.8 years (SD = 5.3). Eighty-three percent of the participants reported being a college graduate, and 28 % of those also reported holding a graduate degree or higher. The average number of children between 3 and 10 years old per household was 2 (SD = .64). The children’s average age was 7.2 (SD = 1.4) and 60 % were boys. Couples worked a combined average of 80.1 h per week (SD = 20.3) and had an average income between $101,000 and $125,000.

Sample 3 (Madigan 2005)

The participants of this study were 89 mothers responding about a target child. Participants were recruited from local preschools, community music and movement programs for children, and other child/family organizations in an urban and suburban mid-Atlantic area. Inclusion criteria for participants were: (1) parents were required to be the child’s primary caregiver and must live with the child for the majority of the year; (2) parents did not have to be the child’s biological parent, (3) if previously divorced, parents were ineligible if the divorce had occurred within the last year, and (4) parents and child had to be fluent in English. After screening 100 interested participants for eligibility criteria, four participants were shown to be ineligible due to a child language problem, a developmental disorder, or coming from a single-parent family. Six other families who were eligible either changed their minds or were unable to schedule a time to meet with the researcher for the study. Thus, 89 of the 100 interested families, participated. Children were, on average, 48.47 months old (SD = 7.29), and were 53 % female and 47 % male. Mothers’ age ranged from 29 to 49 years old (M = 36.54, SD = 4.25). Fathers ranged in age from 28 to 50 years old (M = 38.52, SD = 5.15). As indicated by participants, 81 % of the children were White/Caucasian, 3 % African-American, 3 % Hispanic/Latino, 3 % Asian, and 10 % identified as “other.” All parents were married to the child’s biological father (with the exception of one adoptive mother) for an average of 9.41 years (SD = 3.37). Parents’ income was measured on a 9-point Likert scale ranging from 1 ($0–20,000) to 9 (more than $300,000). The families participating had an average income of $81,000–100,000 per year. In this study, 84 % of mothers had completed a bachelor’s degree and 82 % of fathers reported the same. It was reported that 95 % of fathers were employed full time while 62 % of mothers were unemployed at the time of the study.

Procedure

For sample 1, each parent received the questionnaires from school, and each parent completed the surveys independently (Winsler et al. 2005). Upon completion, parents sent the questionnaires back to school with their child. For sample 2, participants were brought onto a University campus and parents were asked to complete questionnaires in a separate room while their child participated in the rest of the study and completed tasks with an experimenter (Madigan 2005). For sample 3, participants were sent a package of questionnaires (each parent received a separate set of similar questionnaires) and were asked to complete them independently from their spouse (Thorvardarson 2000).

Measures

PSDQ

The Parenting Styles and Dimensions Questionnaire (Robinson et al. 1995) is a 62-item survey of self- and spouse-reported parenting practices for parents of preadolescent children and this was used with all three samples. Each item used a 5-point Likert scale ranging from never (1) to always (5). The PSDQ examines the three parenting dimensions of Baumrind (authoritativeness, authoritarianism, and permissiveness; Robinson et al. 1995). Results of the measure are based on a continuous scale for each dimension of parenting. Large numbers for each parenting style indicate a greater use of that style.

The PSDQ has been praised as one of a few instruments available that are psychometrically sound for assessing parenting style (Locke and Prinz 2002; Olivari et al. 2013), and has been used in multiple cultural settings (Daglar et al. 2011; Wu et al. 2002). Typically these multi-cultural studies focus only on the self-report feature, and very rarely use the measure of parents’ perception of their spouse (Porter et al. 2005). In a recent review of the PSDQ, it was determined that the most valuable aspect of this questionnaire is its adaptability (Olivari et al. 2013). It can be administered to parents as a self-report, spousal-report, retrospective report, or given to a child to report on his/her parents; this flexibility increases the validity of the measure (Olivari et al. 2013). Parenting styles and the relationship between parents and children seems to be relatively stable across time (Dallaire and Weinraub 2005). Sensitive and more positive parenting practices are especially consistent across the early years of children’s lives (Dallaire and Weinraub 2005).

The current study used the PSDQ as a self-report and spouse-report measure. Four ratings, thus, are possible. Each parent rated his and her own parenting styles, and also rated their perception of their spouses’ parenting styles on the same items. Therefore, there was a measure of mothers’ and fathers’ self-perceived parenting style, respectively. Additionally, there was a measure of mothers’ and fathers’ perceptions of their spouses’ styles, respectively. For example, a mother would rate her perception of her spouse’s parenting style, as well as her perception of her own parenting style. The father when available would then do the same. Robinson et al. (1995), with a larger (n = 1251 total parents) sample of school-age children measured internal consistency reliabilities (Cronbach alphas) for both mothers and fathers self-report and reported them to be .91, .86, and .75 for the authoritative, authoritarian, and permissive scales, respectively. For the present, smaller sample including younger children, Cronbach alphas were acceptable and ranged from .73 to .89.

For analyses, several new variables were created. First, we took the absolute value of the difference among self-perceived parenting styles of each parent (i.e., mother self-reported permissiveness minus father self-reported permissiveness). These new variables were continuous with higher numbers meaning there were bigger differences in self-perceived parenting styles between the mother and the father. The same absolute value difference scores were calculated between parent’s report of their own style and their perception of their partner’s style (i.e., mother self-reported authoritativeness minus mother’s perception of her spouse’s authoritativeness). For analyses asking about the directionality of the differences between parents, similar difference score variables were calculated but without taking the absolute value. Thus, the values indicated which parent was higher than the other on each parenting dimension.

PKBS

The measure of child behavior problems used in samples 1 and 3 was the Preschool and Kindergarten Behavior Scales (PKBS; Merrell 1994). The PKBS can be used to assess the behavior problems of children between three and 6 years old. Parents were asked to complete the questionnaire which consists of 76 items about the same target child. The items are designed to identify three dimensions of children’s social skills and five dimensions of behavior problems. Items are answered with a 4-point scale ranging from never (0) to often (3). The five dimensions of behavior problems are then calculated to fall into two broad categories (internalizing and externalizing behavior problems). The internalizing scale is computed with a possible range of 0–45, with a higher number indicating more internalizing behavior problems. The externalizing scale is computed with a possible range of 0–81, with higher numbers indicating more externalizing problems. The PKBS is a frequently used measure with exceptional psychometric properties and construct validity that is sensitive to specific differences relating social skills with behavior problems (Merrell 1994). The Cronbach alphas of these subscales are .90 with a test–retest reliability of .80 for the internalizing subscale and .97 with a test–retest reliability of .87 for the externalizing subscale (Merrell 1994). In the original published work, the construct validity of the problem behavior subscale scales ranged from .46 to .80 when correlated with other similar measures, and correlations between the subscales and the total behavior problems score ranging from .90 to .95 (Merrell 1994). Within our sample, Cronbach alphas were .90 for the internalizing subscale and .97 for the externalizing subscale.

CBCL

The Child Behavior Checklist (Achenbach and Edelbrock 1991; Heflinger et al. 2000) was completed by parents in sample 2, referring to a specific target child. There were a total 13 mothers and 12 fathers (12 couples, one mother’s husband did not respond) who completed the CBCL. The CBCL is a very widely used measure of child externalizing and internalizing behavior problems with excellent psychometric properties, validity, and test–retest reliability (Achenbach and Rescorla 2000; Skovgaard et al. 2004; Warnick et al. 2008). Items are rated from not true (0), to very true/often (2). Test–retest reliabilities are .87 and .89 for the Competence and Behavior Problems scales, respectively (Achenbach 1991). For the eight categories, Cronbach’s alphas ranged from .62 to .92 for 4–11 year old boys, and from .66 to .92 for 4–11 year old girls (Achenbach 1991). For our sample, Cronbach alphas, averaged across boys and girls, were .90 for the internalizing subscale, .93 for the externalizing subscale, and .96 for total behavior problems. Overall internalizing and externalizing behavior subscales were used.

When both mothers and fathers reported on child behavior, the two standard scores were averaged. When only mother or father reports were available for a child, that parent report was used. Using combined mother-father reports of child behavior problems increases reliability and reduces the problem of single-source bias (using the same individual to provide both parenting style and child behavior data). Other studies often use one parent’s report with questionnaire data when the other parent report is unavailable (Martin et al. 2007). For the purpose of the current study given that a different scale was used for sample 2, we combined the CBCL and PKBS into master internalizing, externalizing and total behavior problem scores. T-scores from the CBCL were converted to standard scores in order to be easily compatible with standard scores from the PKBS). When both mothers and fathers reported on child behavior, the two standardized scores were averaged and put into a master internalizing, externalizing, and total behavior problems score. When both parents’ reports were not available, either the mother or father report was included in the master internalizing, externalizing, and total behavior problems variable. Combining the PKBS and CBCL is justified given that these two measures of child behavior problems have been shown to be highly correlated (Brassard and Boehm 2007). The CBCL and PKBS measure the same constructs (internalizing, externalizing, and total behavior problems) according to their manuals, and use the same type of standardization procedures, and are highly correlated with one another, lending support to our decision to combine them.

CPS

The Conflicts and Problem-Solving (CPS; Kerig 1996) scale was used to measure marital and interparental conflict for sample 3 (n = 89 mothers). The CPS is a reliable and commonly used measure of marital conflict that is especially accurate in measuring marital conflict through self-report or spousal report (Kerig 1996). The measure consists of 84 items which are comprised by four conflict dimensions (frequency/severity, degree, efficacy, and outcome) and six strategy subscales (cooperation, avoidance/capitulation, stonewalling, verbal aggression, physical aggression, and child involvement). Parents rated the items on a scale from not a problem at all (0) to a severe problem (100). They were then asked to rate the amount of time (percentage) that disagreements were resolved from never (0) to always (100). Furthermore, parents were asked to rate how often they use strategies such as “Talk it out with partner” using a scale ranging from never (0) to often (3). Lastly, parents were asked to rate on a scale from never (0) to usually (3) how often outcomes of disagreements such as “The whole family ends up feeling upset” occur. For the four conflict dimensions, Kerig (1996) found alphas ranging from .75 to .98. Further, the current sample found alphas of .76 to .91. Alphas for the six strategy subscales ranged from .70 to .87 according to Kerig (1996), and .73–.89 in the current sample.

SMAT

The Short Marital Adjustment Test is a 15-item questionnaire measuring martial satisfaction and conflict. It was only used in sample 2 (n = 39 mothers and n = 36 fathers completed the survey, this totals 35 couples, with a few more mother only, and one additional father only, who completed the questionnaire). This measure combines the score of all 15 items resulting in a possible 2–158 points (higher numbers indicating more adjustment/satisfaction with minimal conflict, and lower numbers indicating maladjustment/dissatisfaction within a marriage and higher levels of conflict). Locke and Wallace (1959) found .90 test–retest reliability. Further, Klitzmann (2000) found alphas of .75 for women and .73 for men. Sample Cronbach alphas for this study were .85 for women and .91 for men. As with the child behavior problem measures, we combined the two similar measures of marital conflict across samples. Each of the scores for the CPS and SMAT were converted to standardized Z-scores. Although both measure aspects of satisfaction and conflict, the CPS is scaled as a measure of conflict (with bigger numbers meaning more conflict), and the SMAT is scaled as a satisfaction scale (with bigger numbers meaning more satisfaction/less conflict). So, we reverse coded the SMAT so bigger numbers meant more conflict. Then, we combined the Z scores into one variable.

Since this study combined across multiple samples and measures, it is important to clarify where the data come from for analyses. All three samples used the PSDQ to measure parenting style, so analyses on just the PSDQ include all participants. Since the CBCL and PKBS were combined into one master score, and all three samples used either the PKBS or the CBCL, analyses involving child behavior include all children. However, only samples 2 and 3 had a measure of marital conflict (the SMAT or the CPS), therefore the composite score that combines these two measures was used for these analyses meaning that only samples 2 and 3 are included in analyses involving parental conflict.

Results

Parenting Style, Marital Conflict, and Child Behavior

Before addressing the main research questions, preliminary analyses replicated previous findings that establish relations between parenting styles, marital conflict, and child behavior. Correlations between self-reported parenting styles and child outcomes and marital conflict are found in Table 1. Marital conflict was positively associated with both internalizing and externalizing behavior problems. Authoritarian and permissive parenting in both mother and father were positively associated with child behavior problems. Multiple regressions predicting child behavior problems from self-perceived parenting styles for both mothers and fathers, with control variables (gender and age) included in step 1 are found in Table 2. Externalizing behavior problems were more common among younger children and the combination of child gender and age in step 1 was significant [F(2, 123) = 3.76, p < .05]. We also examined whether child age was associated with parents’ self-reported parenting styles. Fathers’ parenting styles were never associated with child age, and mothers’ authoritative and permissive parenting did not vary by child age. However, there was a small positive association between mothers’ authoritarian parenting and age of target child (r = .25). As to be expected given this, mothers’ authoritarianism was slightly higher in sample 2 with older children than in the other samples with preschool children. To see if this was a problem in our analyses, we also ran partial correlations between maternal authoritarianism and the other child and family outcomes, controlling for child age—all relationships remained the same. Therefore, we are confident that child age did not influence the conclusions made in this paper.

Table 1 Mothers’ and fathers’ parenting style, conflict/satisfaction and child behavior correlations
Full size table
Table 2 Parenting style predicting behavior problems regression
Full size table

The addition of mothers’ three self-reported parenting styles in step 2 explained 10 % additional variance in child externalizing problems [F(5, 120) = 4.58, p < .01], with specifically maternal authoritarian parenting significantly relating to increased child externalizing problems. Similarly, for child internalizing behavior problems, maternal self report of permissive and authoritarian parenting styles were significant [F(5, 120) = 3.76, p < .01]. Finally, similar results were found for total child behavior problems, with 17 % of the variance in child overall problem behavior being predicted by the control variables and maternal parenting styles [F(5, 120) = 4.93, p < .01]. Controlling for child age and gender, increased maternal self-reported authoritarian and permissive parenting styles were related to increased total child behavior problems (β = .31, t = 3.53, p < .05; β = .21, t = 2.41, p < .05, respectively).

Identical regression models were run for fathers’ self-reported parenting styles predicting child externalizing, internalizing, and total behavior problems. Although similar patterns were seen as reported above for mothers, due to the smaller size of the father sample, the regression models were typically only marginally statistically significant. Of note in Table 2 was that paternal authoritative parenting was marginally negatively related to child internalizing problems when all parenting styles and the control variables of gender and age were included in model 2.

Parenting Style Inter-correlations

Using bivariate correlations, the associations between the three self-reported parenting styles of mothers and fathers (authoritative, authoritarian, and permissive) were examined and are listed in Table 3. For example, mothers’ self-perceived authoritative parenting style was correlated with fathers’ self-perceived authoritative parenting style (top panel of Table 3). Further, mothers’ self-perceived authoritative parenting style was correlated with mothers’ perception of fathers’ authoritative parenting style (second panel of Table 3) and fathers’ self-perceived authoritative parenting style was correlated with fathers’ perception of mothers’ authoritative parenting style (bottom panel of Table 3).

Table 3 Self- and spouse-perceived parenting style correlations
Full size table

An overall pattern was that spouses tended to share parenting styles (r’s between mother and father styles between .31 and .43). Interestingly, informant perspectives of their spouse’s parenting style were more correlated with their own style than were their spouse’s self report, perhaps indicating that parents perceive themselves to be more similar to each other than they actually are. Mothers’ self-reported parenting styles were strongly associated with mothers’ report of their spouses’ style (r = .70, r = .77, r = .61 for authoritative, authoritarian, and permissive, respectively, p < .05). The same was true for fathers’ self-report being associated with fathers’ report of their spouse (r = .77, r = .51 r = .64 for authoritative, authoritarian, and permissive styles respectively, p < .05).

Differences in Perceptions of Parenting Styles Related to Marital Conflict and Child Behavior Problems

To analyze whether differences in self-perceived parenting style between mother and father were associated with child behavior, six bivariate correlations were conducted (three for the absolute value of the difference in style and mothers’ perceptions of marital conflict, and three for the absolute value of the difference in style and fathers’ perceptions of marital conflict). Further, analyses were conducted to examine the relationships between parents’ self-perceived parenting style absolute value differences and child internalizing, externalizing, and total behavior problems. Overall, lower levels of marital conflict were associated with smaller differences in self-reported parenting styles and better child behavior (see Table 4).

Table 4 Differences in mothers’ and fathers’ self-perceived parenting style associated with marital conflict and child behavior problems
Full size table

Parenting Style Differences and Marital Conflict

In general, as mothers’ and fathers’ self-reported parenting styles differed more from each other, marital conflict increased. Results showed that greater differences in self-reported authoritativeness were significantly correlated with greater mother-reported marital conflict (r = .38, p < .05), and marginally correlated with father-reported marital conflict (r = .31, p < .10). For both mother and father-reported conflict, differences in authoritarianism and permissiveness did not significantly relate to marital conflict, however, the difference between parents in authoritativeness was marginally related to increased marital conflict reported by mothers (see Table 4).

Parenting Style Differences and Child Behavior Problems

Correlations between the absolute value differences between mothers’ and fathers’ self-reported parenting styles and child behavior problems are also found in Table 4. As the difference in parents’ self-perceived permissiveness increased, child externalizing behaviors also increased (r = .35, p < .05). There were no other statistically significant relationships between differences in self-perceived styles and behavior. All of the above analyses (Table 4) had to do with self-perceived differences in parenting style (i.e., mother’s perception of her style compared to father’s perception of his). Similar correlational analyses were also run to examine differences within the individual parent in their perception of themselves compared to their perception of their spouse (i.e., mother’s perception of her permissiveness compared to her perception of her spouse’s permissiveness). These analyses are discussed next and summarized in Table 5.

Table 5 Differences in mothers’ and fathers’ perceptions of self and other predicting marital conflict and child behavior
Full size table

Conflict

Overall, similar to what was seen in Table 4 for self-perceived differences, as mothers’ perceived differences between her parenting styles and her spouses’ parenting styles increased (for example, the absolute value of mothers’ self-perceived authoritarianism minus mothers’ perception of the fathers’ authoritarianism), her rating of marital conflict increased (Table 5). This relationship was significant for authoritativeness and permissiveness (r = .64 and r = .40, p < .05) for maternal perceptions. Thus, as the difference in how the mother rated her own authoritativeness and how she rated her spouse’s authoritativeness increases (and the same for permissiveness), her perception of marital conflict increased. The overall trend was similar for fathers’ perceived differences.

Child Behavior Problems

Associations between perceived differences in parenting styles and child behavior problems are also found in Table 5. In general, as reported differences between the perception of one’s own parenting style and the perception of the spouse’s style increase, so do most behavior problems. For mothers, a greater difference in how she rates her authoritativeness and how she rates her partner’s authoritativeness was significantly related to internalizing and total behavior problems in children (r = .20 and r = .21, p < .05, respectively). For fathers, there was a significant association between the difference in how fathers rate their own permissiveness and their spouse’s permissiveness and child externalizing and total behavior problems (r = .35 and r = .38, p < .05, respectively).

Does Directionality of Parenting Style Differences Matter for Marital Conflict and Child Behavior?

Until now, all analyses related to differences in parenting style have been conducted with absolute value difference scores which ignore the direction of which parent is higher than the other on that particular parenting dimension. The last set of analyses was conducted on the directional difference scores within informant (mother’s perception of herself minus mother’s perception of the father, with positive scores indicating that the mother was higher on that dimension than the father and negative scores meaning that the father was higher on that dimension than the mother).

Conflict

Table 6 shows the correlations between these directional scores and marital conflict for mothers. When mothers’ self-reported greater authoritative parenting than perceived in their partners, maternal perception of marital conflict was higher (r = .64, p < .05). Interestingly, when mothers reported themselves as more authoritarian than their perception of their spouse, the mother reported less marital conflict (r = −.40, p < .05).

Table 6 Directionality of mothers’ perceptions of self and her perceptions of spouse predicting marital conflict/satisfaction and child behavior problems
Full size table

Child Behavior

Table 6 also includes correlations with child behavior problems. In terms of child behavior, when mothers saw themselves as being more authoritative than she saw fathers, there were increased total child behavior problems (r = .20, p < .05) and increased child internalizing behavior problems (r = .16, p = .08). Table 7 reports these same directional differences in parenting style perceptions and links to marital conflict and child behavior but from the father’s perspective, that is, father’s perception of himself minus father’s perception of the mother. So a large positive number signifies that the father sees himself as being greater on this dimension than the mother and negative scores mean he sees his spouse as being higher on that dimension than himself. For fathers’ reports, there were no statistically significant correlations between directional perceptions of differences in parenting style and father’s reported marital conflict due to the smaller sample size of fathers, however there was a marginal (p = .056) relation for permissiveness, indicating that when fathers saw themselves as more permissive than their partner, they perceived less conflict in the marriage. As for child behavior, when a father reported himself to be more authoritarian than his perception of the spouse, children displayed more externalizing and total behavior problems (r = .34, p < .05; r = .28, p < .10, respectively). Alternatively, when fathers rated themselves as being more permissive than they rated their spouses, the child displayed fewer internalizing and total behavior problems (r = -.30, p < .10; r = -.33, p < .05).

Table 7 Directionality of fathers’ perceptions of self and his perceptions of spouse predicting marital satisfaction and child behavior problems
Full size table

Discussion

The current study examined maternal and paternal perceptions of their own and their partners’ parenting styles, and linked those perceptions to self-reported marital conflict and child behavior problems. As an advantage over much of the previous research, the current study (a) examined fathers’ perspectives in addition to mothers’, (b) examined not only parental self-reports of parenting style but also their perceptions of their partners’ styles, (c) explored whether directional differences in parenting styles (i.e., mom being higher than dad on permissiveness) mattered for child problems beyond just absolute value differences, and (d) combined the constructs of parenting style, marital conflict, and child behavior problems in the same study. Results were consistent with prior findings that the authoritative style of parenting was linked to the best child behavior outcomes overall while authoritarian and permissive parenting styles were correlated with increased behavior problems (Baumrind 1967, 1991; Cohen and Rice 1997). Also consistent with prior research, we found evidence that mothers’ perception of increased marital conflict was associated with increased child behavior problems (Dadds and Powell 1991; Smith and Jenkins 1991). However, this study went further by taking into consideration both parents’ perceptions of their own parenting styles and of their spouses’ parenting styles.

Consistent with Winsler et al. (2005) and Gamble et al. (2007), we showed that parents tend to be in relationships with people who have similar parenting styles, and they tend to rate their partners as having similar styles as themselves. Interestingly, parents perceive themselves to be more similar to their partners in parenting styles than they actually are according to both parents’ self reports. Acitelli et al. (1993) found that perceived similarity in behavior is typically higher for both wives and husbands compared to actual similarity, and perceived similarity is related to marital well being. The authors argue that perceiving oneself as similar to another provides a sense of security and assurance that allows and protects self-esteem. This, in turn, can create a more harmonious and healthy relationship between the wife and husband (Acitelli et al. 1993). Extending this logic to a counseling setting, when counseling couples, it is important to address perceptions of the other, rather than (or in addition to) focusing on the actual behaviors of the other.

Building on Rinaldi and Howe’s work (2012), we focused on how certain combinations of parenting styles are associated with marital conflict and child outcomes. In addition to mothers’ and fathers’ parenting styles being individually associated with child behavior problems, we found that differences between mother and father parenting styles were also uniquely associated with child behavior problems. Our findings suggest that when the father saw himself as more permissive than the mother this was associated with less paternal-perceived marital conflict and fewer behavior problems in children. These findings suggest that marital conflict and perceived parenting differences is associated with child behavior problems, which supports the longitudinal literature proposing that relationship instability and disagreement among parents can lead to increased child behavior problems (Linville et al. 2010).

Also, although absolute value differences between parents in authoritarianism were not associated with child behavior problems, the particular direction of the difference seemed to matter—when mothers were more authoritarian than fathers within the family, mothers experienced less marital conflict, and when fathers were more authoritarian than mothers within the home, children experienced more behavior problems. Finally, the more different parents were in their authoritativeness, the more marital conflict was experienced, especially by the mother, and the more internalizing (and total) behavior problems were present among the children. Direction seemed to be important here as well—it was specifically the situation where mothers saw themselves as more authoritative than their partner that was associated with higher marital conflict and child behavior problems. Existing models of parenting have argued that mothers are more likely to be authoritative than fathers (Winsler et al. 2005) and having at least one parent who is authoritative can help child outcomes when the other parent is not particularly authoritative (Milvesky et al. 2008; Simons and Conger 2007). Father’s supportiveness also seems to have the greatest influence on child behavior when mothers are lower on the supportive scale (Martin et al. 2010). Our findings fall in line with this prior research with older children and builds on it by incorporating marital conflict, examining child outcomes among young children, and exploring agreement between parents on parenting style.

The above results have important implications for prevention/intervention efforts and for marriage and family counseling. Therapists should assess both parents’ parenting styles as well as their perception of each others’ parenting to look for the particular patterns identified here that are associated with problems (fathers being very authoritarian relative to mothers, mothers being more permissive relative to fathers, and mothers being more authoritative than fathers) and/or strengths (fathers more permissive relative to mothers, mothers more authoritarian than fathers). Applied in a different way, when husbands and wives share similar opinions and beliefs about parenting, wives tend to report increased feelings of intimacy over time (O’Brien and Peyton 2002). These results can have clinical applications as agreement in parenting beliefs could be targeted and addressed in therapeutic settings. It also may be good for parents to understand and be aware of their own styles and how their spouse’s parenting style might differ from their own, and how they can work together and collaborate with their parenting styles, rather than having them clash. For therapists who work with families, these results could help point toward potential areas of disagreement that parents would have not thought of before (such as perceived authoritative parenting style disagreements). When parents understand the source of their disagreement, they may be able to adjust in a way that leads to more effective collaborative parenting and this could lead to better child behavior and decreased couple conflict. Accurate perceptions and expectations of the self and partner when it comes to parenting can lead to more effective collaboration and parenting together, which can lead to less disagreement, more marital satisfaction, and better child outcomes (Feinberg 2002). Often, in families, parent perceptions of each other are more important than their actual behaviors. For example, satisfaction with one’s partner is more related to one’s perception of the partner’s behavior than that person’s actual behavior (Cohen et al. 2012). If one parent perceives the other as being more permissive compared to his or herself, and this has been causing problems in the marriage, then this might be a potential target focus for intervention in improving martial satisfaction and decreasing conflict. This is an area in the literature that has not been explored.

Limitations

Although the current study has made some advances over previous research in the area, there are limitations that must be considered. First is the fact that we had a relatively small sample size (and we combined several different samples into one and in some cases not all participants filled out all of the same measures), making it difficult in some cases to find statistically significant effects. However, the fact that we did find reasonable effects sizes and statistical significance even with our small sample suggests that this is an important area for future research with larger and more systematic samples. Also the current study only used parent reports via surveys for all measures. Observational measures and/or teacher reports of child behavior in the future would help avoid single-source bias. We did include combined mother and father reports of child behavior problems when possible in order to have it not always be reported by the exact same single parent, but the data were still limited to parental self-reports. Other studies (Modry-Mandell et al. 2007) have used more direct assessments of child behavior, family emotional expressions, and parental agreement/disagreement using in-home personal interviews and questionnaires administered by researchers, while other studies have used more laboratory based semi-structured assessments of parent–child interactions (Martin et al. 2010); this kind of methodology provides a deeper understanding of these constructs. Since our study was examining individual perceptions of the other’s parenting style, this can perhaps best be completed through self-report instead of direct assessments or observations. Future research should incorporate self-report with direct observations of behavior.

This sample was fairly homogenous in its ethnic makeup and made up of two-parents (one biological mother, one biological father). More studies are needed to examine these questions concerning perceived parenting styles across a variety of different demographic groups including those from different ethnic backgrounds, those in poverty, and those in more diverse types of families (i.e., step families, adopted families, gay/lesbian parents, extended families). Relations between parenting styles and child behavioral outcomes, for example, are known to vary somewhat across different ethnic groups (Chao 2001; Ispa et al. 2004). Thus, it is important to replicate the work done in the present paper with larger and more diverse samples.

Finally, our single point in time, correlational design is not able to inform about direction of effect. Although empirical research and theory support the direction of causality moving from parental behavior to child behavior problems (Chen and Johnston 2012; Martin et al. 2010; Schoppe-Sullivan et al. 2007) which is what we are assuming, it clearly works both ways, with parenting behavior also being affected by child behavior and child characteristics over time (Zadeh et al. 2010). Child behavior can certainly be an important contributor to marital conflict as well (Manning et al. 2014). Also, parenting style interacts with child temperament and personality, with certain types of parenting leading to better outcomes for children depending on the temperament/type of child (Fay-Stammbach et al. 2014). Future work should examine the active role the child herself plays in determining parenting styles and parenting style disagreements between parents.