Psychological maltreatment affects an estimated 1.1 million children and adolescents every year (National Center on Child Abuse and Neglect, 1996). Nevertheless, when compared to research on physical and sexual abuse, little attention has been given to the role of psychological maltreatment as a risk factor for children's psychological distress. One proposed aspect of psychological maltreatment is degrading parenting behavior (Hart et al., 1987; Kairys et al., 2002) including verbal abuse, name-calling, belittling, ridicule, hostility, sarcasm, unjustified criticism, and humiliation (Garbarino et al., 1986; Iwaniec, 2003; Navarre, 1987).

Despite interest in the effects of emotional abuse, much is still unknown about the impact of degrading parenting practices. A number of possible maladaptive outcomes for children have been theorized, including impaired emotional awareness, anxiety, depression, lowered self-esteem, aggression, poor peer relations, and academic failure (see Brassard and Gelardo, 1987; Thompson and Kaplan, 1996). One study found that adolescents’ self-reported lifetime experience of having been criticized and treated unfairly by parents predicted internalizing and externalizing symptoms (McGee et al., 1995). Another study found that parental criticism and hostility predicted aggression and anxiety in children 6–17 years old (Crittenden et al., 1994).

Associations between degrading parenting and adolescent internalizing symptoms were also suggested by Stone (1994), who reviewed case files for adolescents and found that a history of “emotional abuse”, including degrading behavior, predicted depressive symptoms. Engfer and Schneewind (1982) found that adolescents’ internalizing problems were related to their perception of harsh parental punishment (including yelling and physical abuse). A longitudinal study found that hostile parenting attitudes in parents of school-age children predicted development of adolescent depression (Katainen et al., 1999). Several retrospective studies have also found that self-reported frequency of parental verbal abuse is predictive of internalizing problems in early adulthood (e.g., Duncan, 1999; Yamamoto et al., 1999).

Relations between degrading parenting and externalizing problems in adolescence have been similarly suggested. Several studies by Gerald Patterson and his colleagues have documented that coercive parenting, characterized in part by scolding, threatening, and hostile parenting behaviors, predicts the later development of adolescent aggression and conduct problems (e.g., Patterson, 1982, 1986, 1995). Similarly, Roehling et al. (1996) found that harsh parenting practices (including degrading and physically abusive parenting) were associated with greater levels of conduct disorder in adolescence, particularly for boys. Other researchers have documented an association between hostile parenting and adolescent externalizing (e.g., Webster-Stratton and Hammond, 1999). While there is conceptual overlap between degrading parenting and coercive, harsh, or hostile parenting, definitions of these constructs are not interchangeable, and physically abusive behaviors have traditionally been included within the latter constructs.

Fig. 1.
figure 1

Hypothetical model.

Full size image

While research to date has been suggestive of relations between degrading parenting and adolescent outcomes, a clear consensus in the literature is limited by heterogeneous definitions of parenting behavior (i.e., measures that do not distinguish between verbally degrading, other psychologically maltreating, and physically abusive parenting) and methodological limitations (e.g., use of retrospective data, inclusion of children's outcomes in assessment of parenting behavior). In addition, studies that show relations between psychological maltreatment and children's distress often fall short of investigating the mechanisms (i.e., mediators) that might account for those relations. The present study used a specific measure of degrading parenting that avoided overlap with physical abuse and proposed three mediators of degrading parenting's association with internalizing and externalizing symptoms.

As shown in Fig. 1, hypothesized mediators were avoidant coping, mother-adolescent conflict, and perceived maternal support. A theoretical model by Miller (1983) identified avoidant coping as a potential route by which degrading parenting behavior might negatively impact children's functioning. Miller proposed that children have specific needs that must be met for normal emotional development to occur, including being respected, understood, and accepted. Miller suggested that these basic needs cannot be met when a parent engages in various types of degrading behavior. When a child's basic needs are not met, Miller theorized that the child is likely to engage in avoidant coping by repressing thoughts and feelings of frustration, pain, and helplessness in order to avoid losing parental love. Furthermore, a child's avoidant coping would lead to psychological distress, including depression and anxiety (Miller, 1983).

There is little or no empirical research that shows degrading parenting as a risk factor for reliance on avoidant coping; however, case studies do support cognitive avoidance as an outcome for emotionally maltreated children (see Jacobsen and Miller, 1998). Furthermore, studies of adolescents and adults with other types of abuse histories have found that the use of avoidant coping elevates the risk for psychological distress (Gold et al., 1994; Spaccarelli and Fuchs, 1997).

Parent-adolescent conflict is another potential mediator in the relation between degrading parenting behavior and adolescent adjustment. Aversive parenting practices, including name-calling and humiliation, might disrupt conflict resolution strategies, leading to sustained and more frequent conflicts (see Patterson, 1982). Furthermore, attachment theorists have predicted that some children might respond to degrading parenting with anger and overt resistance (see Crittenden and Ainsworth, 1989), suggesting greater conflict frequency in parent-child relationships. This theory is supported by findings that parenting styles characterized by hostility and physical punishment are associated with elevated levels of adolescent irritability and hostility toward parents (Conger and Ge, 1999; Snyder et al., 1997). In addition, there is suggestive evidence that greater frequency of parent-adolescent conflict might place children at risk for conduct problems (see reviews by Robins et al., 1999; Rutter, 1994). Longitudinal work by Patterson and his colleagues identified a pathway whereby coercive parents who favor scolding, threatening, ignoring, and aggressive responses toward their children tend to elicit greater frequency of noncompliant and impulsive behaviors from their children, increasing risk for conduct disorder (Patterson, 1982, Patterson and Reid, 1984). It is predicted that results from the present study will complement Patterson's findings by identifying mother-adolescent conflict as a mediator in the relation between mothers’ degrading parenting and adolescent conduct problems.

Maternal support is another proposed mediator in the relation between mothers’ degrading parenting behavior and adolescent adjustment. Although our review of the literature did not reveal any studies that have examined the relation between degrading parenting and perceived parental support, it has been theorized that degrading parenting practices might affect a child's perception of being respected and supported by his or her parent (Navarre et al., 1987). If mothers’ degrading parenting does indeed decrease adolescents’ sense of maternal support, this perceived lack of support might, in turn, increase adolescents’ risk for psychological distress. There is considerable evidence that parental support is related to psychological distress in adolescence (see review by Barrera and Li, 1996). What is lacking, however, is research that shows that parental support mediates the relation between degrading parenting and adolescents’ psychological distress.

To summarize, this study tested a model of relations between mothers’ degrading parenting and adolescent adjustment. It was hypothesized that mothers’ degrading parenting would be positively related to adolescent internalizing and externalizing symptoms through three mediating variables: adolescents’ reliance on avoidant coping, mother-adolescent conflict, and perceived maternal support (see Fig. 1). This study extends previous research by clarifying the specific predictive value of degrading parenting (separate from other forms of maltreatment) for adolescent internalizing and conduct problems, evaluating predictions from the psychological maltreatment literature, and investigating potential mediators of these relations.

An important methodological feature of the study was the use of both mothers and adolescents as reporters of degrading parenting and adolescent adjustment. This feature allowed us to determine if the hypothetical model was specific to the perceptions of just one set of reporters, or if the model fit the perceptions of both mothers and adolescents. Although gender and ethnic group differences in model fit were not hypothesized, such differences were tested to evaluate the external validity of the model.

METHOD

Participants

Participants were 232 adolescents (115 boys and 117 girls) and their mothers. Adolescents ranged in age from 11 to 15 years, with a mean age of 13.0 years. The sample included 62 Caucasian mothers (26.7%) and 170 Mexican-American mothers (73.3%). In the Mexican-American group, 69 mothers completed the assessment in English and 101 mothers completed it in Spanish. Mothers’ mean age was 37.37 years, with mean educational level for mothers at 10.0 years. The sample included 69 single-parent families and 163 two-parent families. Median annual income was $15,000–$20,000 overall; $20,000–$25,000 for two-parent families and $10,000–$15,000 for single-parent homes.

Procedure

Participants were recruited from school rosters of 7th and 8th grade students in two school districts in a large southwestern city. One of the school districts had an ethnic makeup of 80.6% Hispanic; the other school district had an ethnic makeup of 71.0% Hispanic. Families were recruited with telephone calls and with letters (for those who lacked telephones). In the overall study, 74.3% of the families who were eligible for the study were successfully recruited into the study.

All data were collected in home interviews that required approximately 2 hr. Each participant was interviewed individually by a professional interviewer who used the participant's preferred language. Prior to assessment interviews, parents gave written consent to participate in the study and gave written consent granting permission for their child's participation. In addition, children gave their written assent. One-parent families received $30 for participation; two-parent families received $45. Parents and children were interviewed concurrently by separate interviewers. Interviewers read all questions from a laptop computer screen. To maintain privacy, participants entered their responses using the computer keyboard without verbalizing their responses. In most cases, parents and children were interviewed in separate rooms.

Measures

Degrading Parenting Behavior (Mother and Adolescent Report)

A seven-item scale was constructed to measure frequency of degrading parenting behavior without including items concerned with physical abuse. Content validity considerations led to the selection of items that assessed hostile and demeaning parenting practices. Items were developed from data obtained during a qualitative study of ethnically diverse, low-income parents (Gonzales et al., 2000), from harsh-parenting scales used by Conger and Elder (1994), and from the hostile-control subscale of the Children's Report of Parent Behavior Inventory (CRPBI; Schaefer, 1965). Adolescents and their mothers used a 5-point response scale ranging from “almost never” to “almost always” to indicate the frequency with which each behavior had occurred in the past 3 months. Scores for each reporter were calculated by taking the mean of the seven items. Higher scores indicated greater frequency of degrading parenting behavior.

Criterion validity for the degrading parenting measure was established by examining the relations between it and a subscale of the Parent-Adolescent Conflict Scale termed “Power Assertion” (Gonzales et al., 2000). The four Power Assertion items assess whether threats, intimidation, aggression, or criticism occur during conflict between mothers and their adolescent children. The zero-order correlations between mothers’ degrading parenting and mothers’ use of power assertion ranged from .31 to .41 (within-reporter, p < .001) provided additional evidence for the validity of the degrading parenting measure.

Mother-Adolescent Conflict (Adolescent Report)

Seven items from the Parent-Adolescent Conflict Scale (PACS) were used to assess the frequency of conflict between adolescents and their mothers (Gonzales et al., 2000). These items measure the presence of general disagreements and conflicts, both minor and serious (e.g., “You and your mother became frustrated with each other”, “…had a serious argument or fight”). Adolescents rated the degree of conflict they had with their mothers in the past 3 months using a 5-point response scale that ranged from “never” to “always.” This measure was scored by taking the mean of seven items. High scores indicated greater conflict frequency.

The PACS was designed to capture parent-adolescent conflict that occurs in low-income African-, European- and Mexican- American families. Item content was based on qualitative interviews conducted with 32 low-income, ethnically diverse families (Gonzales et al., 2000). In the present sample, the 7-item PACS showed an internal consistency reliability of .88. As an indication of its concurrent validity, it was correlated .50 with a 20-item version of the Issues Checklist, a well-known parent-child conflict measure (Prinz et al., 1979). The convergence of mothers’ and adolescents’ reports of conflict on the PACS (r=.49) was higher than the convergence of mothers’ and adolescents’ reports of conflict on the Issues Checklist (r = .29).

Perception of Maternal Support (Adolescent Report)

Nine items based on the Inventory of Parent and Peer Attachment (IPPA; Armsden and Greenberg, 1987) were used to assess adolescents’ perception of maternal support. The original IPPA included 25 items. For the present study, a 9-item version was developed from factor analysis of a separate multi-ethnic sample of 450 7th and 8th grade students (Gonzales and Jackson, 1996). In the Gonzales and Jackson study, reliabilities for mother-adolescent support ranged from .89 to .92 for the ethnic subgroups (African American, Mexican American, and Caucasian). Furthermore, the scale was positively correlated with maternal acceptance (.66) and maternal monitoring (.54) which were validity indicators. Adolescents rated how often each item (e.g., “My mother helped me to talk about my difficulties”) occurred in the last 3 months on a 5-point response scale that ranged from “almost never” to “almost always.” A score for this scale was derived by taking the mean of 9 items. High scores indicated greater perceived maternal support.

Avoidant Coping (Adolescent Report)

Adolescent reliance on avoidant coping was an 11-item scale. It included four items from Avoidant Actions, four items from Repression, and three items from Wishful Thinking subscales of the Children's Coping Strategies Checklist (CCSC) (Ayers et al., 1996). Items on the CCSC included items written by Ayers and his colleagues and items derived from the Behavior-based Coping Inventory (Wills et al., 1985). Adolescents were asked to report on how often they usually used each strategy (e.g., “you just forgot about it”) to solve their problems or make themselves feel better within the past month. A four-point response scale ranged from “never” to “almost always.” A mean score is calculated for this scale. Higher scores reflected greater reliance on avoidant coping strategies.

Adolescent Internalizing (Mother Report)

An internalizing score was computed by adding the standardized mean of 19 items on the depression subscale (e.g., “[my child] seemed to feel worthless or inferior”) and the standardized mean of 17 items on the anxiety subscale (e.g., “[my child] was nervous, high strung, or tense”) of the Child Behavior Checklist (Achenbach and Edelbrock, 1991). The zero-order correlation for mother’ ratings of anxiety and depression was .63 (p < .001). Mothers used a 3-point response scale ranging from “not true” to “often true” to rate the frequency of occurrence for each item within the past month. Higher scores indicate greater levels of internalizing symptoms.

Adolescent Internalizing (Adolescent Report)

Internalizing symptoms were assessed with adolescents’ reports on two scales, the Children's Depression Inventory (CDI) and the Revised Children's Manifest Anxiety Scale (RCMAS). The CDI is a 27-item, multiple-choice scale that assesses affective, cognitive and behavioral symptoms of childhood depression (Kovacs, 1981). Each CDI item assesses one symptom by having the child select one of three statements, each representing different levels of symptom severity over the past month.

Castaneda et al. (1956) developed the CMAS as a measure of a child's chronic state of anxiety. Reynolds and Richard (1978) revised the scale into the RCMAS by reordering items, adding and deleting items, and developing new norms. The 28-item RCMAS yields a total score by counting all symptoms that were experienced in the past month.

A total internalizing score was computed for each adolescent by adding the standardized means of the RCMAS and CDI scores. The zero-order correlation for adolescents’ ratings of anxiety and depression was .68 (p < .001). Higher scores indicate greater levels of internalizing symptoms.

Adolescent Conduct Problems (Mother Report)

Mothers’ perceptions of adolescents’ conduct problems were assessed with externalizing behavior subscales (Delinquent and Aggressive Behavior) from the Child Behavior Checklist (CBCL; Achenbach and Edelbrock, 1991). One item from the CBCL (“argued a lot”) was deleted in order to eliminate content overlap between this measure and the measure of mother-adolescent conflict frequency. Six additional delinquency items were added to supplement the CBCL scale to increase the content validity of conduct problems for the present sample. One item (“use of force to get something”) was included from the Denver Youth Survey (Esbensen et al., 1999). Two items were derived from focus groups conducted with mental health professionals who worked in communities similar to the one sampled (“sneaking out of the house in the middle of the night without parents’ permission” and “leaving home for more than one day without parents’ permission”). Three additional items (“participation in gang activity”, “spreading lies and rumors”, and “lied about his/her age to buy or do something”) were added by the research team as indicators of conduct disorder/delinquency. Mothers used a 3-point response scale ranging from “not true” to “often true” to rate the frequency of occurrence for each item within the last month. A mean score was computed for a total of 29 items. High scores indicate greater conduct problems.

Table I. Measurement Descriptives
Full size table
Table II. Correlation Matrix for Variables Included in Model
Full size table

Adolescent Conduct Problems (Adolescent Report)

Adolescents responded to 22 items from the externalizing scale of the Youth Self-Report Scale (YSR) developed by Achenbach (1991). One item from the YSR (“argued a lot”) was eliminated from this measure in order to minimize content overlap between measures. In addition, six delinquency items that paralleled the six additional items on the parent ratings of conduct problems (see above) were included in this measure of adolescent conduct problems. Adolescents rated the frequency with which each item occurred in the past month on a 3-point scale. A score for conduct problems was derived by taking the mean of 27 items. Higher scores indicate greater conduct problems.

Fig. 2.
figure 2

Final model for mothers’ reports of degrading parenting and adolescent outcomes (M-M model). Single-headed arrows denote direct paths. Paths with statistically significant unstandardized loadings (maximum likelihood) are shown. Characters d1-d5 refer to structural disturbances. Psi (ψ) parameters reflect error covariance.

Full size image

RESULTS

Table I shows descriptive statistics and internal consistency reliabilities for each measure. Table II gives zero-order correlations between variables included in the model. The proposed model (see Fig. 1) was evaluated using path analysis to estimate direct effects, predicted mediating relations, and the overall goodness-of-fit of the model. All of the measures were determined to have skew and kurtosis in acceptable ranges (skew < 2.0 and kurtosis < 3.0). The ratio of participants (232) to number of parameters to be estimated (15) was 15.47, suggesting that the sample size was large enough to provide accurate estimates of parameters and goodness of fit (Bollen, 1989). Parameters in the proposed model were estimated using the LISREL 8.3 program (Joreskog and Sorbom, 1989). Several demographic variables were considered for inclusion in the model as “covariates,” but they were either unrelated to the criteria (adolescent age, per capita income, and number of parents in the family) or did not affect model fit (mother's education). For parsimony, those demographic variables were not included in subsequent model testing.

Estimation of the Model

The model shown in Fig. 1 was tested first using mother's report of degrading parenting and mother's report of adolescent conduct problems and internalizing. This was named the “M-M model” because mothers reported on both degrading parenting and adolescents’ symptoms. The three mediators were adolescent report variables. The initial analysis indicated that the model did not fit well; χ2(6)=190.88, p < .001; Goodness-of-Fit Index (GFI)=.82, Adjusted GFI=.39, Normative Fit Index (NFI)=.34, Comparative Fit Index (CFI)=.33, Root Mean Square Residual (RMSR)=.09.

From inspection of modification indices, error covariances were freed for mother's report of adolescent conduct problems and internalizing (ψ54), mother-adolescent conflict and perceived maternal support (ψ21), and perceived maternal support and avoidant coping (ψ32). Also, the path loading (β51) was freed for mother-adolescent conflict frequency on adolescent internalizing symptoms. The revised model resulted in χ2(2)=4.00, ns; GFI=.99, Adjusted GFI=.94, NFI=.99, CFI=.99, RMSR=.01, indicating an excellent fit to the observed data. The freed β51 parameter resulted in a significant improvement in fit over the previous model specification Δχ2=6.91, Δdf=1, p < .01).

Subsequently, the revised model was estimated with adolescent report of both degrading parenting and the criterion variables (A-A model). As with the M-M model, the A-A model fit the data very well (χ2(2)=4.26, ns; GFI=.99, Adjusted GFI=.99, NFI=.99, CFI=.99, RMSR=.01). The ratio of participants (232) to number of parameters to be estimated (19) in this refined model was still sufficient to provide accurate estimates of parameters and goodness of fit (Bollen, 1989).

Direct Effects

Figure 2 shows the significant direct paths in the M-M model. As predicted, mother's degrading parenting was positively related to mother-adolescent conflict frequency, adolescent conduct problems, and adolescent internalizing. Moreover, mother's degrading parenting was negatively related to adolescent perception of maternal support. Mother-adolescent conflict frequency was positively related to adolescent conduct problems and internalizing. Predicted positive relations between avoidant coping and adolescent internalizing were also supported. Relations between mother's degrading parenting and avoidant coping, and between perceived maternal support and internalizing, were not significant.

Table III. Summary of Significant Direct Effects Across Reporter Groupings
Full size table
Table IV. Indirect Effects
Full size table

Figure 3 shows the final model using adolescent report of mother's degrading parenting, conduct problems, and internalizing (A-A). Consistent with the hypothesized model, mother's degrading parenting was positively related to both mother-adolescent conflict frequency and avoidant coping, and negatively related to perceived maternal support. A hypothesized positive relation between conflict frequency and adolescent conduct problems also was supported. As expected, perceived maternal support was negatively related to internalizing, while avoidant coping was positively related to internalizing. As with the M-M model, a positive relation between mother-adolescent conflict frequency and adolescent internalizing was significant. Predicted direct relations between mother's degrading parenting and adolescent outcome variables (conduct problems and internalizing) were not supported in the A-A model.

Fig. 3.
figure 3

Final model for adolescents’ reports of degrading parenting and adolescent outcomes (A-A model). Single-headed arrows denote direct paths. Paths with statistically significant unstandardized loadings (maximum likelihood) are shown. Characters d1-d5 refer to structural disturbances. Psi (ψ) parameters reflect error covariance.

Full size image

A hypothesized direct relation between perceived maternal support and adolescent conduct problems was not supported for either model. Direct loadings for Figs. 2 and 3 are presented as unstandardized loadings (rather than standardized loadings) in order to allow for accurate comparison across models (Bollen, 1989). See Table III for a summary of the significant direct paths for the M-M and A-A models.

Indirect Effects

Estimates of total indirect effects and their standard errors were used to compute t ratios for each mediational pathway. Those t ratios were then tested for statistical significance. Indirect path coefficients, standard errors, t values, and significance levels for each of the M-M and A-A models are presented in Table IV. Full mediation was indicated when the indirect effect was found to be statistically significant, but the direct effect was nonsignificant. Partial mediation was demonstrated when the indirect and direct effects were both statistically significant (Bollen, 1989). Results of mediation tests are presented in Table V.

In the M-M model, mother-adolescent conflict partially mediated degrading parenting's relations to both internalizing and conduct disorder symptoms.

In the A-A model, the effect of mother's degrading parenting on adolescent conduct problems was fully mediated by mother-adolescent conflict. The effect of mother's degrading parenting on adolescent internalizing symptoms was fully mediated by conflict frequency, perceived support, and avoidant coping.

Table V. Summary of Significant Mediational Pathways Across Reporter Groupings
Full size table

Multi-Sample Analyses: Gender

Multi-sample analyses investigating the comparability of path loadings in the model across samples of boys (n=115) and girls (n=117) were conducted (Widaman and Reise, 1997). Estimation of a baseline M-M model showed a similar pattern of path loadings for boys and girls (χ2(4)=6.47, ns; CFI=.99). In a second step, the model was estimated with path loadings that were constrained to be equal for boys and girls (χ2(23)=32.44, ns; CFI=.97). When the baseline model and the constrained model were compared, they did not differ significantly (Δχ2=25.97, Δdf=19, ns), indicating that the path loadings for boys and girls were equivalent.

Those steps were repeated for estimates of the A-A model. Those tests showed that the path loadings for boys and girls were equivalent with one exception. The path loading from degrading parenting to maternal support was −.50 for boys, but −.07 for girls.

Multi-Sample Analyses: Ethnicity

Multi-sample analyses also were conducted to explore potential differences in the model between Caucasian and Mexican-American participants. With the M-M model there was a similar pattern of path loadings for Caucasians and Mexican Americans (χ2(4)=4.98, ns; CFI=1.00). In a second step, the model was estimated with path loadings that were constrained to be equal for the two ethnic groups (χ2(23)=30.31, ns; CFI=.97). When the baseline model and the constrained model were compared, they did not differ significantly (Δχ2=25.33, Δdf=19, ns), indicating that the path loadings for the two ethnic groups were equivalent.

The tests of equivalence were repeated for estimates of the A-A model. Those tests showed that the path loadings for Caucasian and Mexican American adolescents differed in several ways. First, the relation of mother-adolescent conflict and conduct problems was stronger for Mexican American adolescents (.27, p < .001) than it was for Caucasian adolescents (.09, ns). Also, the relation between degrading parenting and mother-adolescent conflict was stronger for Caucasians (.82, p < .001) than it was for Mexican Americans (.43, p < .001). Finally, the relation of maternal support to internalizing symptoms was stronger for Mexican Americans (−.84, p < .001) than it was for Caucasians (−.16, ns).

DISCUSSION

This study aimed to add to our understanding about the effects of degrading parenting behavior in adolescence in several important ways. First, the model empirically evaluated predictions from the psychological maltreatment literature regarding relations between maternal degrading behavior and adolescent adjustment. Second, this study clarified the specific usefulness of mothers’ verbal hostility, belittling, and humiliating parenting, separate from other types of adverse parenting practices, in predicting adolescent conduct problems and internalizing. Third, this study contributed to greater understanding about parent-adolescent conflict, perceived maternal support, and avoidant coping as mediators in the relation between maternal degrading parenting and adolescent adjustment. This study found that mothers’ degrading parenting was positively related to adolescent conduct problems and internalizing through many of the hypothesized mediational pathways. It was an important step in documenting a relation between degrading parenting and problematic adolescent adjustment, and in identifying some of the mechanisms that could explain that relation.

In this study, mother-adolescent conflict mediated the relation between mother's degrading parenting and adolescent internalizing. Families characterized by mothers’ degrading parenting were more likely to have mother-adolescent conflict, which, in turn, was associated with higher risk for adolescent internalizing problems. This is consistent with previous work that found an association between daughters’ frequency of conflict with their mothers and internalizing symptoms (Powers and Welsh, 1999) and an association between hostile parenting and adolescent depression (Kaitainen et al., 1999). The current study's findings add to the literature by identifying the unique predictive value of mother's degrading parenting behaviors for adolescent internalizing, as well as a mediational pathway by which degrading affects adolescent outcomes.

Perceived maternal support was also found to be a significant mediator of degrading parenting's relation to internalizing problems. Although numerous studies have demonstrated a relation between perceived parental support and adolescents’ internalizing problems (see Barrera and Li, 1996), the links between degrading parenting, parental support, and internalizing symptoms had not been established adequately. The identified mediational pathway is consistent with theory that degrading parenting increases risk for adolescent internalizing in part because it thwarts adolescents’ needs for belongingness (Brassard and Gelardo, 1987) and conveys a message that the degrading parent will not be available or supportive to the adolescent (Navarre, 1987).

Similarly, the positive relation between avoidant coping and adolescent internalizing had been fairly well established in previous studies (Fields and Prinz, 1997; Herman-Stahl et al., 1995; Moos, 1997; Sandler et al., 1994; Wills, 1997). The present study, however, adds to this literature by linking degrading parenting and adolescents’ avoidant coping, as well as establishing avoidant coping as a pathway whereby mother's degrading parenting may affect adolescent risk for internalizing symptoms. That mediational pathway lends some empirical support to Miller's (1983, 1997) theory that adolescents who experience degrading parenting are more likely to rely on repression and cognitive avoidance, thus increasing risk for depression and anxiety.

The mediational pathways for perceived maternal support and avoidant coping were only found for internalizing symptoms and only for adolescents’ reports. It is possible that when adolescent reports of degrading parenting or internalizing are used, relations between variables in the model are inflated due to shared reporter variance. However, it also could be that adolescents are more accurate reporters of their mothers’ degrading behavior and their own internalizing symptoms than are their mothers. In support of this stance, Schwarz et al. (1985) found that parents tended to be biased in presenting a more favorable image of their own parenting behavior, and that adolescents were more accurate reporters of child-rearing behavior. Also, because internalizing problems include the personal experience of negative affect and private cognitions, adolescents would be in a better position to report on those problems than would their mothers.

Although all three mediators showed the predicted relations with internalizing symptoms, only mother-adolescent conflict was a significant mediator of conduct problems. That mediational pathway complemented past research efforts that showed positive relations between harsh parenting and parent-adolescent conflict (e.g., Reuter and Conger, 1995), as well as Patterson's (1982, 1986, 1995) work that demonstrated how coercive parenting leads to adolescent conduct disorder. As with coercive parenting, degrading parenting behaviors (such as name-calling, sarcasm, and humiliation) might increase mother-adolescent conflict frequency by disrupting effective conflict resolution strategies (Patterson, 1982). Parent-adolescent conflict, in turn, increases risk for adolescent conduct problems (Barrera et al., 1993; Formoso et al., 2000; Patterson and Reid, 1984; Rutter, 1994). This study's findings enhance the current literature by clarifying the specific predictive value of degrading parenting, separate from parenting constructs that include physical abuse, in considering risk for mother-adolescent conflict and adolescent conduct problems.

Results from the current study suggest that mothers’ degrading parenting is associated with risk for internalizing and conduct problems, regardless of adolescent gender or ethnicity (Caucasian and Mexican-American descent). Nevertheless, some unhypothesized ethnic group differences were found when adolescents were reporters of parenting, the mediators, and the outcomes. Degrading parenting was more highly related to mother-adolescent conflict for Caucasians than it was for Mexican Americans. That finding might be due to the influence of familism and simpatia for Mexican American families that would stress positive relations within the family and the avoidance of overt conflict (Castro and Hernandez, 2004). If Mexican American families do, in fact, value the avoidance of conflict, conflict might be more damaging when it does occur between Mexican American mothers and their children. That would explain why the relation between mother-adolescent conflict and conduct problems was stronger for Mexican American adolescents than it was for Caucasian adolescents. The relation between maternal support and internalizing problems also was stronger for Mexican American adolescents than it was for European American adolescents. Formoso et al. (2000) noted that Hispanic families have been characterized by close emotional ties between parents and children. They speculated that, “Because there are strong cultural incentives to maintain supportive parent-child bonds, even in the face of family adversity, these children may be more likely to derive protective benefits from these bonds” (pp. 179–180). Even though the results of the present study are consistent with that speculation, research has not found greater benefits of parental support for Hispanic adolescents relative to other ethnic groups (e.g., Formoso et al., 2000; Hill et al., 2003). Research has yet to provide a compelling explanation of the variability in this finding.

Strengths of this study include the use of a large community sample that included both male and female adolescents from single- and two-parent families, which enhances the generalizability of this study's findings. Furthermore, this study's design permitted an investigation of whether the relations in the model differed depending on adolescent gender or ethnicity. The use of multiple reporters for mothers’ degrading parenting, adolescent conduct problems, and adolescent internalizing allowed for the examination of reporter effects.

However, several limitations should be noted. First, the cross-sectional design of this study did not allow for estimates of prospective relations between degrading parenting and the other model constructs. Another limitation of this study was its exclusive focus on the mother-adolescent relationship, without simultaneous examination of the influence of father-adolescent relationships within two-parent families. From this study we do not know how fathers’ parenting practices might influence the risk associated with mothers’ degrading parenting. Previous studies have found that mothers and fathers influence adolescent adjustment differently (Collins and Russell, 1991; el-Guebaly et al., 1978; Formoso et al., 2000; Steinhausen, 1984; Werner, 1986). It would be interesting to examine both mothers’ and fathers’ degrading parenting simultaneously, in order to investigate the possibility of asymmetrical effects, interactions between parenting behaviors, and the degree to which one partner's parenting behaviors affect the other's parenting quality.