Parental control is not unconditionally detrimental for externalizing behaviors in early childhood

Abstract

The association of three different strategies of maternal control (behavioral, psychological, and physical), and maternal warmth with children’s externalizing behaviors were analyzed in an observational study of 3-year-old children in Turkey (N = 123). The results indicated that (i) mothers exercised all three types of control simultaneously; (ii) behavioral control had a curvilinear association with child externalizing behaviors, suggesting the existence of an optimum level of behavioral control; and (iii) the negative effects of behavioral and psychological control could be moderated by parental warmth. These findings highlighted the importance of studying samples from diverse cultural contexts in order to validate and enrich theoretical models of behavioral development.

Keywords

behavioral control culture externalizing behavior observation study parental control physical control psychological control warmth

The present research identifies the linkages between three different strategies of maternal control (physical, behavioral, and psychological) and externalizing behaviors in 3-year-old children. Externalizing behaviors refer to acting-out behaviors including aggressive, impulsive, hostile, defiant, oppositional, and destructive behaviors (Achenbach, Edelbrock, & Howell, 1987). A number of studies showed that externalizing behaviors in early childhood predicted problems in peer and family relationships in childhood, and academic and conduct problems in adolescence (e.g., substance abuse, school dropout and delinquency; Deater-Deckard & Dodge, 1997; Joussemet et al., 2008). This study was conducted in Turkey, in a cultural context where few children attend preschool at the age of 3, where few mothers of preschool children are employed, and where both maternal control and maternal warmth is high.

Parental control refers to all parental behaviors with the intention to inhibit, direct or modify an ongoing behavior of the child to elicit a desired behavior (Schaffer & Crook, 1980). Since nearly all parents attempt to control their children’s behaviors with a variety of strategies, one must study the outcomes of different degrees and strategies of control on children’s behaviors. Here, we provide definitions and distinctions between physical, behavioral, and psychological control. Next, we provide a brief review of empirical research linking parental control to externalizing behaviors of children in diverse cultural contexts. Then, we present the social and cultural context of the current study. We offer our theoretical arguments, conceptual model, and hypotheses in the final subsection.

Physical, behavioral, and psychological control

The definition of physical control has been well established by previous research. Physical control is a strategy of forceful, coercive or power-assertive control in parenting (Strassberg, Dodge, Petit, & Bates, 1994). Physical control is operationally defined to include behaviors such as yelling, grabbing, pushing, hitting, physically hurting, shouting, using overt expressions of anger, and using physical threats (Patterson, 1982; Strassberg et al., 1994). In the current study, physical control is defined as the parental effort to physically punish or threaten to punish, restrict, or physically redirect the child.

Many previous studies, including the current study, have defined behavioral control as the set of attempts to manage or control the child’s behavior via expectations, guidelines, limits, clear and consistent rules, restrictions, and structures (Aunola & Nurmi, 2005; Barber, 1996; Gurland & Grolnick, 2005; Hasebe, Nucci, & Nucci, 2004; Joussemet et al., 2008; Mills & Rubin, 1998; Shek, 2007). Observational measures of behavioral control include explicit or implicit commands, directives, prohibitions, rules, maternal verbalizations of expectations, unsolicited checking, taking over the child’s task, imposing a particular structure to a task on which parent and child were working together, and over-involvement (Gurland & Grolnick, 2005; Mills & Rubin, 1998).

In the present study and numerous other studies, psychological control is defined as parenting behaviors that constrain and manipulate children’s emotional experiences such as making the child feel guilty, anxious, and worthless through manipulating the parent–child bond (e.g., love-withdrawal, guilt induction, and emotionally isolating the child), expressing negative affect (e.g., through demonstration of anger, impatience), and directing negative criticism in an attempt to induce shame (Aunola & Nurmi, 2005; Hasebe et al., 2004; Mills & Rubin, 1998; Shek, 2007). There have been relatively few studies that distinguish between the effects of behavioral and psychological control. Nevertheless, factor analytic studies support the distinction between these two types of control (Aunola & Nurmi, 2005; Barber, Olsen, & Shagle, 1994; Hasebe et al., 2004; Shek, 2007). Operational definitions of psychological control refer to parental attempts to induce negative emotions in the child subsequent to misbehavior (e.g., disappointment, shame, guilt, not being worthy), parental distance and non-responsiveness toward the child, parental insults, humiliation, blame or generalized criticism (Aunola & Nurmi, 2005; Barber, 1996; Barber et al., 1994; Gurland & Grolnick, 2005; Mills & Rubin, 1998; Shek, 2007; Verhoeven, Junger, van Aken, Deković, & van Aken, 2010).

Consequences of parental control in diverse cultural contexts

Setting some limits and establishing rules (“structure”) for young children reduce externalizing problems (Barber et al., 1994). However, overprotective and overly concerned mothers and highly controlling mothers tend to have children who are more likely to be socially withdrawn or aggressive than children of mothers who use normative levels of control.

The link between parental control and child externalizing behaviors has been explained by three mechanisms. The first mechanism appears to be culture-neutral. High degrees of control provide few opportunities for children to self-regulate, and results in an inability to rely on themselves for emotional and behavioral regulation due to an intensive and restrictive approach to discipline (Barber, Stolz, & Olsen, 2005; Chen, Liu, & Li, 2000; Gurland & Grolnick 2005; Rubin & Mills, 1990). It has also been found that children of parents who use too little behavioral control have externalizing behavior problems, difficulties in self-regulation, and lower competence (Aunola & Nurmi, 2005; Barber et al., 1994; Baumrind, Larzelere, & Owens, 2010). The totality of this body of research points to a U-shaped association of behavioral control with externalizing problems. Physical control has also been linked to aggressive and defiant behaviors in children who have been forced to exhibit a desirable behavior without learning to regulate their own behaviors, without internalizing the rules of conduct, and without opportunities to express their feelings and thoughts (Strassberg et al., 1994). A large body of research supports the strong link between physical control and externalizing behaviors (Gershoff, 2002). Although some recent studies suggested that discriminating and occasional use of physical punishment could be effective in instilling desired behaviors (Larzelere & Kuhn, 2005), the current study was not designed to capture those rare occasions when physical control could be called for.

The second mechanism that links parental control and child externalizing problems is likely to be culture-dependent. Different sets of parental behaviors coexist in different cultural contexts leading to culture-specific parenting “styles.” Those styles that are frequently observed in Western European and Anglo-American cultures are not common in other cultural contexts. Previous studies that used data from Anglo-American samples established a few distinct combinations of strategy of control, degree of control, and degree of warmth (“styles” such as authoritarian, authoritative, and permissive styles). These styles tend to predict children’s externalizing behaviors (Deater-Deckard & Dodge, 1997). However, parenting “styles” that are culture specific could lead to distinct behavioral consequences for a child, because the effects of different types of parenting behaviors could be multiplicative rather than additive. For example, parents exercising control at the levels generally associated with an authoritarian parenting style did not generally lack warmth in Turkey and in other non-Western cultures such as China, Korea, and Japan, and in minority populations within the Anglo-American culture (Deater-Deckard & Dodge, 1997; Dekovic, Pels, & Model, 2006; Hughes, Blom, Rohner, & Britner, 2005; Kagitcibasi, 1996; Pomerantz & Wang, 2009; Rohner, Khaleque, & Cournoyer, 2005; Rudy & Grusec, 2006; Wu et al., 2002).

The third mechanism linking parental control to externalizing behaviors may also be culture-specific. Previous research indicated that the meaning and importance given to parental control and expectations of control levels are culturally-influenced (Dwairy & Achoui, 2010; Hughes et al., 2005; Rohner et al., 2005; Stern, Rohner, & Sacks-Stern, 2007). The meaning attributed to control shaped its effects on externalizing behaviors. High levels of parental control, when exercised concurrently with high levels of parental warmth and support (Chen et al., 2000; Hughes et al., 2005; Stern et al., 2007) might not have detrimental consequences (Kagitcibasi, 1996; Parpal & Maccoby, 1985; Rudy & Grusec, 2006; Wu et al., 2002). The buffering effects of parental warmth on the association between high control and externalizing behaviors were indicated in some previous empirical studies (Aunola & Nurmi, 2005; Erkman & Rohner, 2006; Kagitcibasi, 1996; Kim & Rohner, 2002; Parpal & Maccoby, 1985).

Previous studies posited that children of parents who exercised psychological control learned to perform desirable behaviors because of the fear of parental love-withdrawal without understanding the general rules of conduct (Barber et al., 1994). Furthermore, they found that psychological control was highly likely to be interpreted as rejection by children (Rohner et al., 2005). A number of empirical studies have linked parental psychological control to externalizing problems (Aunola & Nurmi, 2005; Barber et al., 2005; Joussemet et al., 2008; Verhoeven et al., 2010).

The context of the current study

The current study was conducted in Turkey, a “majority world” (a term initially introduced by Kagitcibasi, 1990) country, located between Europe and the Middle East. Turkish society has experienced a rapid economic, demographic, and social transition accompanied by rapid urbanization (25% rural in 2010) and a drop in the fertility rate (total fertility rate of 2.1 in 2010; World Bank, 2010). Despite substantial cultural change, Turkey is ranked 37th out of 93 countries in collectivism (Hofstede, Hofstede, & Minkov, 2010).

Historically, obedience to family rules have been an important parenting goal for Turkish parents (Kagitcibasi, 1990). However, recent studies have found that the importance placed on obedience has substantially diminished (Kagitcibasi & Ataca, 2005). Nevertheless, it has been repeatedly shown that Turkish parents use high levels of control and power-assertive strategies in disciplining their children, accompanied by a high level of warmth (Kagitcibasi, 1996; Kircaali-Iftar, 2005). This cultural context provides an excellent opportunity to study the separate and interacting influences of parental control and warmth.

In a cross-cultural comparison, the average number of commands (an indicator of behavioral control) issued by Turkish mothers was 14.5 per minute in a community sample, whereas this number ranged between 1.5 and 9 per minute among the mothers of Head Start children, the mothers of children with ADHD, and the mothers who were referred to family protective services in the US (Akcinar & Baydar, 2011; Arslan, 2010; Baydar, Reid, & Webster-Stratton, 2003). The number of behaviors expressing warmth in the Turkish sample was 2.7, compared to 0.9 per minute among the mothers of Head Start children (Akcinar & Baydar, 2011). In this context, a high level of maternal control could be perceived as an expression of acceptance rather than rejection, because it was exercised simultaneously with a high level of warmth.

Theoretical framework and hypotheses

Three major developmental theories provide insight into the processes through which different control strategies may be associated with children’s externalizing behaviors: the coercion theory, the Parental Acceptance-Rejection Theory (PART), and the social learning theory. The coercion theory (Patterson, 1982) provides a micro interactional perspective; that is, if a child responds to parental control with negative or oppositional behavior (whining, shouting, etc.), and if this aversion wards off further parental control, then the child’s oppositional behavior is reinforced. Such exchanges are expected to result in an escalation of negative or power-assertive behaviors displayed by both the parent and the child.

Rohner’s PART offers another process that may account for the negative consequences of high levels of parental control (Rohner et al., 2005). According to the PART, children are active recipients of parental behaviors and the way a child interprets these behaviors will shape their consequences. Thus, parental control may have negative consequences when the child perceives a high level of control as an expression of rejection.

The social cognitive approach suggests that parental control may influence a child’s behaviors through two mechanisms: (i) it may deprive the child of experiences that could promote social problem solving and self-regulation and, (ii) it may provide negative behavioral models (Bandura, 1977). Children of parents who exercise high levels of behavioral, psychological, and physical control do not get opportunities to learn to regulate their own behaviors, solve problems in their interactions with others, and participate actively in their own social relationships with their peers. The inability to solve problems in social interactions may result in deviant and aggressive behaviors that may be partly modeled after parental behaviors, and may be expressions of an inability to self-regulate (Rubin & Mills, 1990).

It is likely that distinct strategies of control lead to distinct patterns of behavior because of (i) the differences in the behavioral models that they present to children; (ii) the differences in the patterns of interaction that they lead to; (iii) the differences in which they are interpreted by the children in a given situation, in a given context of other parental behaviors, and in a given cultural context; and (iv) the differences in the degree to which they allow opportunities for self-regulation and social problem-solving.

The following hypotheses were tested in the current study. First, high levels of maternal control (behavioral, psychological and physical control) would lead to high levels of externalizing behaviors in children due to their negative influence on self-regulatory skills. Second, there would be a curvilinear effect of behavioral control on externalizing behaviors, because both low and high levels of behavioral control will promote the inability to self-regulate. Third, maternal warmth would have a buffering effect on child externalizing behaviors when displayed simultaneously with maternal control because it will moderate the interpretation of control as rejection by the child. Fourth, socioeconomic status (SES) would potentially influence parental control behaviors and child externalizing behaviors.

Method

Participants

Children and mothers of this study were selected among the participants of the Study of Early Childhood Developmental Ecologies in Turkey (ECDET). The ECDET study had a nationally representative sample of 1,052 children aged 36–47 months, and their mothers. Both qualitative and quantitative data were collected through home visits that lasted 2–3 hours. The sample for this study consisted of 123 children and their biological mothers who were randomly selected from the ECDET study sample in the four largest metropolitan areas of Turkey to participate in an observational protocol. Observational procedure could not be implemented with the larger sample because of the high cost. The comparison of the characteristics of the study sample to the characteristics of the larger sample is presented in Table 1. The significant differences between the observational sample and the larger sample were having an urban origin, χ²(1, 123) = 1412.25, p < .01, being a married mother, χ²(1, 123) = 6.63, p < .01, and having a second child, χ²(1, 123) = 7.45, p < .01.

Table 1.

Characteristics of the study sample.

Characteristics	Total Sample (N = 1,052)	Observational sample (N = 123)
Age of child (in months)	41.53 (3.65)	41.59 (3.64)
Male	55.4%	47.2%
Age of mother (in years)	30.08 (5.74)	29.54 (5.68)
Birth order of the child
First child	51.8%	57.0%
Second child	27.9%	29.2%
Third child or higher parity	20.3%	13.8%
Number of children
One child	28.2%	30.9%
Two children	41.3%	48.0%
More than three children	30.5%	21.1%
Family size	5.31 (2.13)	4.53 (1.45)
Mother’s education level
Did not complete elementary school (%)	15.6%	11.4%
Completed from elementary school (%)	53.2%	46.3%
Did not complete high school (%)	11.2%	15.4%
Completed from high school (%)	14.4%	19.5%
Completed from college or higher (%)	5.2%	7.3%
Mother’s SES level
Low SES (%)	36.4%	26.0%
Middle SES (%)	35.9%	32.5%
High SES (%)	26.0%	38.2%
Urban origin (%)	54.0%	78.9%
Married mothers (%)	98.2%	95.9%
Working mothers (%)	12.5%	15.4%

Note. The values are the means with the standard deviations in parenthesis.

A factor score was computed to quantify the SES of the participants. The factor score was based on the levels of maternal and paternal education, a measure of material possessions of the family, and an estimate of the total per-person monthly expenditures of the family reported by the mothers. This factor score was constructed using data from the full sample of ECDET (N = 1,052), thus zero represents the national mean. The cutoff scores of −0.5 and 0.5 were used to define low, medium and high SES families. All indicators of SES significantly and substantially differed across the three categories of SES. Of the mothers with low SES, 39% did not complete 5 years of primary school compared to 4.9% of the middle-SES mothers, and 1.5% of the high-SES mothers. The families with low SES had a mean level of per person monthly expenditures of $68, compared to $107 for middle-SES families and $207 for high-SES families.

Procedure

The participants were video-recorded in their homes during a 10-minute semi-structured observation on the same day when the maternal interviews were conducted. A play task with 45 Lego Duplo^TM pieces was used to create a semi-structured situation to stimulate interaction between the mother and the child. The task and the figures were age-appropriate. Three pictures of simple Lego^TM figures (a tree, a house, and a boat) were given to the children to replicate. Mothers were instructed that the children would build the figures and that the mothers could help their children in this task. The instructions were carefully worded to avoid misinterpretations that could lead to mothers’ perceiving a speed- or performance-oriented task. An additional model was available (a bird) if the children completed the task in less than 10 minutes. Few dyads who completed the three figured task sooner than 10 minutes started the additional figure, but none could complete the fourth figure. The camera was handled by the interviewer. Both the mother and the child could see the camera. The interviewers were instructed not to intervene in the play activity and to record for exactly 10 minutes.

Measures

Parental and child behaviors can be measured both by parent reports and observations. These two different methods are often used together but may yield inconsistent results (Robson, 2002). An observational method may lack validity and temporal reliability because it assesses behaviors in a limited time interval in a specific situation. Nevertheless, observational measures may help understand behavioral triggers and their consequences in micro-level interactions. In contrast, parental reports may measure general patterns of behaviors. They may have temporal stability and predictive validity. However, parental reports may not be valid or reliable sources of information on micro-level interactions because these interactions may occur without the awareness of the parent (Stafford & Bayer, 1993). Furthermore, parental reports may be partly influenced by parental psychological well-being (Gartstein, Bridgett, Dishion, & Kaufman, 2009). The present study relies on the observational method to assess the parent behaviors, and both the observational method and parent reports to assess the children’s behaviors. Only the externalizing behavior as a child outcome, not the internalizing behaviors, was our interest, because internalizing behaviors are difficult to measure by observational methods.

Dyadic parent-child interaction coding system

The original Dyadic Parent-Child Interaction Coding System (DPICS; Robinson & Eyberg, 1981) is a behavioral observation system that assesses the quality of parent–child interactions. The DPICS assesses the frequency of specific types of verbal exchanges and behaviors during parent–child interactions. The Turkish version of the Dyadic Parent-Child Interaction Coding System (DPICS-TR) was developed for the present study (Baydar, Akcinar, & Arslan, 2007) to score video recorded interactions. The original DPICS manual was translated to Turkish and adapted to assess relevant interactions in Turkish families.

The videos were coded by two trained coders whose inter-rater reliability calculated as the proportion of codes that were in agreement. It was 0.90 for all parent behaviors, and 0.79 for all child behaviors (Arslan, 2010). Behaviors coded by the DPICS were grouped into four composite indicators of parenting behaviors: behavioral control, psychological control, physical control, and maternal warmth. The composite scores were calculated by summing the occurrences of targeted behaviors that indicated each parenting construct. All observational frequencies were standardized to express the number of relevant behaviors per minute. All behaviors, regardless of their brevity, were included in the specific behavior category if the mother displayed it.

Behavioral control

The composite behavioral control measure included: (1) direct and indirect commands and play directives (e.g., “Give it to me,” “Let’s put it here”) and (2) rules and warnings about the play behaviors (e.g., “If you build this figure then you can play with whatever you want”).

Psychological control

The composite psychological control score was obtained by combining: (1) parental attempts to influence the emotions of the child through threats that were not related to the natural consequences of the child’s undesirable behaviors through guilt induction and through withdrawal of love (e.g., “If you don’t do this I will not love you anymore.”); (2) negative talk that expressed verbal disapproval aiming to induce guilt and remorse (e.g., “Stop already!”); (3) critical statements or demeaning comments such as insulting, blaming, and providing a generalized criticism of the child’s pattern of behavior (e.g., “You are not good at this.”); (4) being non-responsive to the child’s activity or verbalization aiming to emotionally isolate the child.

Physical control

The composite physical control behavior measure consisted of: (1) intentionally inflicting pain on the child (e.g., spanking, hitting, pinching); (2) physically restraining, physically forcing, pulling or pushing the child; (3) physically intruding in the play or physically interrupting the ongoing activity of the child without the consent of the child (e.g., grabbing an object from the child); (4) threats of physical or hurtful punishment that were not related to the natural consequences of the child’s negative behaviors (e.g. “If you act up, I will spank you”).

Maternal warmth

The composite parental warmth measure consisted of: (1) positive touch (e.g., mother hugs, kisses); (2) parent utterances that communicate positive affect (e.g., “My sweetheart!”); (3) labeled praises (e.g., “You did a very good job completing the model in this picture.”); (4) unlabeled praises (e.g., “Well done!”); (5) utterances of acknowledgment of a behavior or suggestion of the child (e.g., “Yes, sure!”). These behaviors represented the extent of maternal love, approval, acceptance, and responsiveness to the child.

Child externalizing behaviors

The composite child externalizing behavior measure included (1) smart talk (e.g., “No, I am not going to do it!”); (2) oppositional behaviors (e.g., shrugging his/her shoulder); (3) destructive behaviors (e.g., throwing the Lego^TM pieces at the wall); (4) physically negative-child behaviors (e.g., hitting, kicking).

Eyberg child behavior inventory-TR

The original Eyberg Child Behavior Inventory (ECBI, Eyberg & Robinson, 1983) assesses the frequency and severity of externalizing behaviors in children between the ages of 2 and 17 years based on maternal reports. In the current study, an adapted version of the ECBI was used (Batum & Yagmurlu, 2007). In order to ensure the validity of ECBI-TR, factor analyses and validity checks with other scales used in the ECDET study were conducted. The ECBI-TR included 36 items (e.g., “Fights with the peers,” “Argues with the parents about rules”) and maintained the original structure except that the frequencies of behaviors were rated on 5-point Likert scales instead of 7 to facilitate comprehension and reliable reporting by participants who had limited education. A total intensity scale score was computed. The internal reliability of the ECBI-TR total intensity scale for the larger ECDET sample was 0.93 (Baydar, Kuntay, Goksen, Yagmurlu, & Cemalcilar, 2007).

Statistical methods

Regression analyses were used to estimate the effects of the three measures of control and warmth on observed and maternally-reported child externalizing behaviors. A quadratic term for behavioral control was included in the regression model to test our hypothesis of a non-linear association of behavioral control with externalizing behaviors. Interaction effects of each one of the measures of control with warmth were tested. In order to reduce multicollinearity, due to the presence of a linear and a quadratic term, and the presence of interaction terms, these higher order terms were computed with centered variables (Belsley, Kuh, & Welsch, 1980).

Results

The means, standard deviations and correlations of DPICS composite parenting measures for behavioral, psychological and physical control and maternal warmth are presented in Table 2. The mothers displayed almost five times as many attempts at behavioral control as psychological control and they used physical control the least frequently, about half as often as psychological control. There was approximately one behavior communicating warmth for almost every four attempts to control through any type. Maternal control measures were associated, but they did not overlap. Furthermore, behavioral and psychological control measures were positively associated with warmth, and physical control was orthogonal to warmth.

Table 2.

Correlations between composite DPICS parenting measures and child externalizing behavior measures.

	Range	Mean	SD	2	3	4	5	6	7
1. Socioeconomic status	−1.55−3.33	0.25	1.05	−.27**	−.26**	−.33**	.02	.00	−.25**
2. Behavioral control	2.77−29.17	14.52	5.36		.42**	.27**	.50**	.13	.28**
3. Psychological control	0.20−6.74	2.80	1.47			.45**	.27**	.18*	.23*
4. Physical control	0.00−4.23	1.45	1.08				−.04	.20*	.10
5. Maternal warmth	0.20−13.58	4.91	2.69					−.04	.20*
6. DPICS Externalizing behaviors	0.00−2.54	0.59	0.69						.23*
7. ECBI Externalizing behaviors	2.78−92.36	35.21	17.82
N		123

Note. DPICS scores are the number of behavior occurrences per minute.

** p < .01; * p < .05.

Predictors of externalizing behaviors

Regression analyses indicated that there was a U-shaped association between behavioral control and observed (DPICS) child externalizing behaviors (Table 3, Panel 1, Model 1). Mothers’ psychological control and physical control behaviors did not significantly predict DPICS externalizing behaviors; however, as expected, behaviors indicating warmth were predictive of low levels of externalizing behaviors (Table 3, Panel 1, Model 2). The total effect size of behavioral control and warmth were of equal absolute value (0.21, result not presented in Table 2).

Table 3.

The unstandardized regression coefficients of the hierarchical regression models predicting DPICS and ECBI externalizing problems with additive and multiplicative effects of maternal control and maternal warmth (N = 123).

Predictors	Model 1	Model 2	Model 3	Model 4	Model 5	Model 1	Model 2	Model 3	Model 4	Model 5
	DPICS child externalizing problems					ECBI child externalizing problems
SES	−.005	.057	.053	.040	.053	−2.922^†	−2.954^†	−3.587*	−3.277*	−3.102^†
Behavioral control	.007	.016	.013	.013	.013	.828*	.694^†	.569	.633^†	.648^†
Behavioral control squared¹	.003*	.003*	.004*	.004**	.004**	−.029
Psychological control		.046	.045	.054	.045		.696	.628	.970	.681
Physical control		.088	.091	.085	.085		−.698	−.498	−.822	−.841
Maternal warmth		−.055*	−.050^†	−.047^†	−.045		.125	.658	.379	.411
Behavioral control × maternal warmth²			−.002					−.312**
Psychological control × maternal warmth²				−.025^†					−.835**
Physical control × maternal warmth²					−.028					−.896
R ²	.05	.12	.12	.14	.13	.10	.11	.17	.15	.13

Notes. ^† p < .10; * p < .05; ** p < .01.

1. In order to reduce multicollinearity, the square term was computed after it was normalized to have a mean of zero.

2. In order to reduce multicollinearity, the interaction term was computed after both variables were normalized to have a mean of zero.

High and low levels of behavioral control exercised by the mothers predicted high levels of externalizing behaviors in children. The effect size and the significance of the quadratic term of behavioral control did not change when other dimensions of control and maternal warmth were entered in the model (Table 3, Panel 1, Model 2). These results are displayed in Figure 1. The figure indicates that there was an optimum level of behavioral control (11–14 attempts to behavioral control per minute) that was empirically associated with the lowest levels of externalizing behaviors. This level falls just under the mean level for this sample.

Figure 1.

The curvilinear association of behavioral control with the observational measure of child externalizing behaviors (DPICS).

The regression model for maternally reported behavior problems (ECBI) indicated that the curvilinear term for behavioral control did not have a significant coefficient (Table 3, Panel 2, Model 1). The linear term for behavioral control had an effect size that was identical to the total effect size of behavioral control on observed child externalizing behaviors (0.21). Psychological control, physical control, and maternal warmth did not significantly predict ECBI. Maternally reported behavior problems were also weakly predicted by SES (Table 3, Panel 2, Model 2).

Moderating effects of maternal warmth

Coefficients of interaction terms of each one of the three measures of control (behavioral, psychological, and physical) with maternal warmth were estimated on the two measures of externalizing behaviors. The observational measure of externalizing behaviors (DPICS) was not significantly associated with these interaction terms (Table 3, Panel 1, Models 3, 4, and 5). However, for the maternally reported externalizing behaviors, interaction terms of warmth with behavioral and psychological control were significant (Table 3, Panel 2, Models 3 and 4). The coefficient of the interaction of warmth with physical control was not significant (Table 3, Panel 2, Model 5).

To facilitate the interpretation of these interactions, Figures 2 and 3 display the predicted ECBI scores at varying levels of warmth and behavioral control and warmth and psychological control, respectively. The results showed that among the mothers who expressed low levels of warmth, the higher the levels of behavioral and psychological control, the higher were the externalizing behaviors of the children (see Figures 2 and 3). The predicted differences in mean ECBI scores of children significantly differed only if maternal warmth was low, indicating a detriment for children whose mothers used high levels of behavioral control. The predicted mean ECBI scores did not differ significantly for low and high levels of psychological control; however, the same trend was observed. Behavioral or psychological control was not significantly associated with substantial differences in child externalizing behaviors when the mothers displayed warmth at the mean levels of this sample, or at a level higher than the sample mean.

Figure 2.

ECBI externalizing behaviors predicted by the interaction of maternal warmth with behavioral control.

Figure 3.

ECBI externalizing behaviors predicted by the interaction of maternal warmth with psychological control.

Discussion

This study had three aims. We investigated (i) the patterns of association of different types of maternal control (behavioral, psychological and physical control) with externalizing behaviors of 3-year-old children in Turkey; (ii) whether the association of maternal control with externalizing behaviors depended on the level of warmth of the mother; (iii) the association of maternal control with both externalizing behaviors in micro level interactions (as indicated by an observational measure), and in generalized patterns of behavior (as indicated by maternal reports), using two measures that quantified these two types of externalizing outcomes.

The results indicated that mothers of 3-year-old children in Turkey used high levels of behavioral control, but the relative frequencies of use of three types of control was consistent with the literature on parental control in early childhood years (Aunola & Nurmi, 2005; Rubin & Mills, 1990). The uses of three types of control were positively associated which indicates that the mothers did not have a strong preference for one type of control over the others. In contrast to this finding, studies conducted with Anglo-American adolescent samples showed that mothers had such a preference (Barber et al., 1994; Shek, 2007). However, Aunola & Nurmi (2005) found that mothers of 5–6-year-old children used behavioral and psychological control simultaneously, similar to our findings. It appears that parents of young children exercise any available strategy of control, probably because young children need a lot of direction for behavioral regulation.

The overall levels of behavioral, psychological, and physical control were positively associated with externalizing behaviors of the children, regardless of the particular measure of child behaviors. This finding lends support to the conceptual model based on the hypotheses of modeling of power-assertive strategies (Bandura, 1977), high levels of parental control-inhibiting self-regulation (Gurland & Grolnick, 2005; Rubin & Mills, 1990), and the hypothesis of coercive interactions (Belsky, Pasco Fearon, & Bell, 2007; Patterson, 1982). Our study does not allow us to specify the exact mechanism of association, but some of the findings may be used to further speculate on this issue.

The studies of parent–child relationships suggest three possible reasons why parent behaviors may be associated with child behaviors. First, there is a possibility that disruptive and challenging child behaviors elicit high levels of parental control although previous studies remain inconclusive regarding this possibility. Nevertheless, there is ample evidence that positive, attentive, sensitive, and compliant child behaviors elicit positive parenting practices (Bell & Chapman, 1986; Verhoeven et al., 2010). Second, studies focusing on problem behaviors suggest that negative parenting practices lead to problem behaviors in children (Aunola & Nurmi, 2005; Gershoff, 2002). Third, there are theoretical and empirical evidence for the bidirectionality in parent–child relationships (Harach & Kuczynski, 2005; Patterson, 1982). In this study, we posited a parent-to-child direction of influence and focused on the processes that may underlie that directional association. We conducted some sensitivity analyses on this issue using additional data obtained when the current sample of children were 4 years old. Those results supported the hypothesis that earlier maternal behaviors predicted later child behaviors, but not the opposite (results available upon request). This study is of a correlational nature. Further experimental and longitudinal studies are needed to investigate the relative strengths of each direction of this influence.

There was a U-shaped association of behavioral control with externalizing behaviors. This finding contributes to previous studies showing that, when the mothers use an optimal level of behavioral control, children display low levels of behavior problems because having reasonable and developmentally appropriate limits for the children allow them to learn self-regulation (Barber et al., 1994). Nevertheless, the empirically-established “optimal” level of behavioral control in other studies was below the mean level for this sample, suggesting a context-dependent optimum.

Maternal warmth had a moderating role on the association of behavioral and psychological control with externalizing problems only for the maternally-reported child behavior problems that measured general patterns of behavior. Three reasons may have accounted for the discrepancy in the findings for observational and maternal report measures of externalizing behaviors. First, the limited dispersion of the observed externalizing behaviors may have resulted in a lack of power to detect interaction effects. Second, maternally-reported externalizing behaviors reflect the maternal perceptions of those behaviors. Those mothers who perceived their children’s behaviors as moderate might be more likely to display warmth than others, resulting in a spurious association. Furthermore, temperamental disposition of the child may influence both parenting behaviors and the maternal reports of child behaviors. This likelihood was investigated with a sensitivity analysis, by including a measure of reactive temperament in the analyses. The findings on maternal control were robust. Although we do not eliminate the possibility of a statistical issue (dispersion) and a measurement issue (maternal reports), we focus the remainder of this discussion on the substantive reasons for the differences in the findings for micro level situational displays of externalizing behavior versus generalized patterns of externalizing behaviors.

The observational method yielded a measure of the situational externalizing behaviors of the children during the course of an interaction with their mothers. Three-year-old children may not have been able to perceive and respond to both warmth and control at the same time, during the course of an interaction, because of their limited regulatory capacity. The negative emotional state because of uninvited maternal control may have prevailed in generating their behaviors. In contrast, maternally-reported generalized patterns of behavior may be outcomes of modeling, perception and interpretation of maternal behaviors that are constructed over many interactions. There is theoretical reason to believe that the level of maternal warmth can moderate these processes that operate on a different time scale than instantaneous responses. On this time scale, maternal warmth may attenuate generalized levels of externalizing behaviors in the presence of high levels of control (Belsky et al., 2007; Patterson, 1982).

Children’s perception and interpretation of parental control may be specific to each culture (Putnick et al., 2012). The perception of control as parental rejection has negative outcomes (Putnick et al., 2012; Rohner et al., 2005; Stern et al., 2007), whereas the normativeness and legitimateness of parental control may buffer its negative effects (Lansford et al., 2010). In the current study, because children were 3 years of age, it was impossible to study the effects of child perception and interpretation. Nevertheless, the findings about the moderating role of maternal warmth suggest that behavioral and psychological control are not always perceived negatively.

Physically coercive and power-assertive control appears to be associated with negative behavioral outcomes regardless of the context of warmth (Strassberg et al., 1994). The reasons may be that (i) physical control is perceived as parental rejection regardless of other parental behaviors (Rohner et al., 2005), (ii) it provides a strong behavioral model of aggression (Strassberg et al., 1994), and (iii) may induce a fear of retribution that impedes the ability to learn from mistakes, resulting in a low level of ability to self-regulate in any cultural context (Joussement et al., 2008). Some studies have suggested that there may be beneficial effects of physical punishment when used carefully as a disciplinary method (Larzelere & Kuhn, 2005). Since our structured task did not call for a high level of control of the child’s behavior, the use of physical control during this task may be indicative of an indiscriminant use of physical control.

One of the limitations of the current study is that it could not address how the effects of parental control may change as the child gets older. A child may seek interactions that do not involve control, and may interpret control more negatively as s/he gets older. For preschool ages, the ultimate aim of the parent may be to control the child with whatever strategy available. Parental reasoning may also change as the child gets older and achieves self-regulation. The sample size of the current study is large when compared with similar observational studies (Timmer, Borrego, & Urquiza, 2002; Werba, Eyberg, Boggs, & Algina 2006). Nevertheless, even larger sample sizes than the current one may be desirable for the statistical power that will allow the analyses of complex patterns of moderation.

The results of the investigation of the association of maternal warmth with maternal control and their joint effects on children underscores the importance of studying diverse maternal behaviors together. For the subsample of children whose mothers had a low level of warmth, similar to those documented in Western European and Anglo-American cultures, a high level of parental control was associated with a high level of externalizing behaviors, similar to previous studies. However, for the subsample of children whose mothers had levels of warmth substantially higher than those documented in Western European and Anglo-American cultures, the pattern of association was reversed, pointing to a typical restriction of range problem with some of the previous studies.

The results pertaining to the protective role of maternal warmth from the detrimental effects of behavioral and psychological control has an important policy implication for cross-culturally adopted interventions for reducing externalizing behaviors in children. It is possible that in contexts where parental control is normatively high, programs that focus on achieving moderate declines of control simultaneously with strong increases in expressions of warmth may be successful.

Footnotes

Funding

The research reported in this manuscript has been funded through The Scientific and Technological Research Council of Turkey (TUBITAK) [Project number 106K347].

References

Achenbach

T. M.

Edelbrock

Howell

(1987). Empirically based assessment of the behavioral/emotional problems of 2- and 3- year-old children. Journal of Abnormal Child Psychology, 15, 629–650.

Akcinar

Baydar

(2011, July). Parental Control and Its Behavioral Consequences for Preschool Children in Turkey and in USA. Paper presented at the International Association for Cross-Cultural Psychology (IACCP) Regional Conference, Istanbul.

Arslan

(2010). Adaptation of Dyadic Parent Child Interaction Coding System (Unpublished master’s thesis). Koc University, Istanbul, Turkey.

Aunola

Nurmi

J. E.

(2005). The role of parenting styles in children’s problem behavior. Child Development, 76, 1144–1159.

Bandura

(1977). Social learning theory. Englewood Cliffs, NJ: Prentice-Hall.

Barber

B. K.

(1996). Parental psychological control: Revisiting a neglected construct. Child Development, 67, 3296–3319.

Barber

B. K.

Olsen

J. E.

Shagle

S. C.

(1994). Associations between parental psychological and behavioral control and youth internalized and externalized behaviors. Child Development, 65, 1120–1136.

Barber

B. K.

Stolz

H. E.

Olsen

J. A.

(2005). Parental support, psychological control, and behavioral control: Assessing relevance across time, culture, and method. Monographs for the Society of Research in Child Development, 70(4), serial no. 282. Boston, MA: Blackwell Publishing.

Batum

Yagmurlu

(2007). What counts in externalizing behaviors? The contributions of emotion and behavior regulation. Current Psychology: A Journal For Diverse Perspectives On Diverse Psychological Issues, 25(4), 272–294.

10.

Baumrind

Larzelere

R. E.

Owens

E. B.

(2010). Effects of preschool parents’ power assertive patterns and practices on adolescent development. Parenting, 10, 157–201.

11.

Baydar

Akcinar

Arslan

. (2007). Dyadic parent-child interaction coding system. Unpublished manuscript.

12.

Baydar

Kuntay

Goksen

Yagmurlu

Cemalcilar

. (2007). Eyberg Child Behavior Inventory-TR. Unpublished manuscript.

13.

Baydar

Reid

M. J.

Webster-Stratton

(2003). The role of mental health factors and program engagement in the effectiveness of a preventive parenting program for Head Start mothers. Child development, 74(5), 1433–1453.

14.

Bell

R. Q.

Chapman

(1986). Child effects in studies using experimental or brief longitudinal approaches to socialization. Developmental Psychology, 22(5), 595–603.

15.

Belsley

D. A.

Kuh

Welsch

R. E.

(1980). Regression diagnostics: Identifying influential data and sources of collinearity. New York, NY: John Wiley and Sons.

16.

Belsky

Pasco Fearon

R. M.

Bell

(2007). Parenting, attention and externalizing problems: Testing mediation longitudinally, repeatedly and reciprocally. Journal of Child Psychology and Psychiatry, 48, 1233–1242.

17.

Chen

Liu

(2000). Parental warmth, control, and indulgence and their relations to adjustment in Chinese children: A longitudinal study. Journal of Family Psychology, 14, 401–419.

18.

Deater-Deckard

Dodge

K. A.

(1997). Externalizing behavior problems and discipline revisited: Nonlinear effects of culture, context, and gender. Psychological Inquiry, 8, 161–175.

19.

Dekovic

Pels

Model

(Eds.). (2006). Unity and diversity in childrearing: Family life in a multicultural society. Lewiston, NY: Edwin Mellen Press.

20.

Dwairy

Achoui

. (2010). Parental control: A second cross-cultural research on parenting, culture, and psychological adjustment of children. Journal of Child and Family Studies, 19(1), 16–22.

21.

Erkman

Rohner

R. P.

(2006). Children’s perceptions of corporal punishment, parental acceptance, and psychological adjustment in a Turkish metropolis. Cross-Cultural Research, 40, 250–267.

22.

Eyberg

Robinson

E. A.

(1983). Conduct problem behavior: Standardization of a behavioral scale with adolescents. Journal of Clinical Child Psychology, 12, 347–354.

23.

Gartstein

M. A.

Bridgett

D. J.

Dishion

T. J.

Kaufman

N. K.

(2009). Depressed mood and parental report of child behavior problems: Another look at the depression–distortion hypothesis. Journal of Applied Developmental Psychology, 30, 149–160.

24.

Gershoff

E. T.

(2002). Parental corporal punishment and associated child behaviors and experiences: A meta-analytic and theoretical review. Psychological Bulletin, 128, 539–579.

25.

Gurland

S. T.

Grolnick

W. S.

(2005). Perceived threat, controlling parenting, and children’s achievement orientations. Motivation & Emotion, 29, 103–121.

26.

Harach

L. D.

Kuczynski

(2005). Construction and maintenance of parent-child relationships: Bidirectional contributions from the perspective of parents. Infant and Child Development, 14, 327–343.

27.

Hasebe

Nucci

M. S.

(2004). Parental control of the personal domain and adolescent symptoms of psychopathology: A cross-national study in the United States and Japan. Child Development, 75, 815–828.

28.

Hofstede

G. J.

Minkov

(2010). Cultures and organizations. Software of the Mind (3rd edn). New York, NY: McGraw-Hill.

29.

Hughes

Blom

Rohner

R. P.

Britner

P. A.

(2005). Bridging parental acceptance-rejection theory and attachment theory in the preschool strange situation. Ethos, 33, 378–401.

30.

Joussemet

Vitaro

Barker

E. D.

Cote

Nagin

D. S.

Zoccolillo

Tremblay

R. E.

(2008). Controlling parenting and physical aggression during elementary school. Child Development, 79, 411–425.

31.

Kagitcibasi

(1990). Family and socialization in cross-cultural perspective: A model of change. In Berman

(Ed.), Cross-cultural perspectives: Nebraska symposium on motivation, 37, 1989 (pp. 135–200). Lincoln: University of Nebraska Press.

32.

Kagitcibasi

(1996). Family and human development across cultures: A view from the other side. Mahwah, NJ: Erlbaum.

33.

Kagitcibasi

Ataca

(2005). Value of children and family change: A three decade portrait from Turkey. Applied Psychology: International Review, 54, 317–337.

34.

Kim

Rohner

R. P.

(2002). Parental warmth, control, and involvement in schooling predicting academic achievement among Korean American adolescents. Journal of Cross-Cultural Psychology, 33, 127–140.

35.

Kircaali-Iftar

(2005). How do Turkish mothers discipline children? An analysis from a behavioural perspective. Child: Care, Health & Development, 31, 193–201.

36.

Lansford

J. E.

Malone

P. S.

Dodge

K. A.

Chang

Chaudhary

Tapanya

Deater-Deckard

. (2010). Children’s perceptions of maternal hostility as a mediator of the link between discipline and children’s adjustment in four countries. International Journal of Behavioral Development, 34, 452–461.

37.

Larzelere

R. E.

Kuhn

B. R.

(2005). Comparing child outcomes of physical punishment and alternative disciplinary tactics: A meta-analysis. Clinical Child & Family Psychology Review, 8(1), 1–37.

38.

Mills

R. S. L.

Rubin

K. H.

(1998). Are behavioural and psychological control both differentially associated with childhood aggression and social withdrawal? Canadian Journal of Behavioural Science. 30, 132–136.

39.

Parpal

Maccoby

E. E.

(1985). Maternal responsiveness and subsequent child compliance. Child Development, 56, 1326–1334.

40.

Patterson

G. R.

(1982). Coercive family process. Eugene, OR: Castalia.

41.

Pomerantz

E. M.

Wang

(2009). The role of parental control in children’s development in Western and East Asian countries. Current Directions in Psychological Science, 18, 285–289.

42.

Putnick

D. L.

Bornstein

M. H.

Lansford

J. E.

Chang

Deater-Deckard

Di Giunta

Bombi

A. S

. (2012). Agreement in mother and father acceptance-rejection, warmth, and hostility/rejection/neglect of children across nine countries. Cross-Cultural Research, 46, 191–223.

43.

Robson

(2002). Real world research: A resource for social scientists and practitioner-researchers. Oxford, UK: Blackwell.

44.

Robinson

E. A.

Eyberg

S. M.

(1981). The dyadic parent-child interaction coding system - standardization and validation. Journal of Consulting and Clinical Psychology, 49, 245–250.

45.

Rohner

R. P.

Khaleque

Cournoyer

D. E.

(2005). Parental acceptance rejection theory, methods, evidence, and implications. Ethos, 33, 299–334.

46.

Rubin

K. H.

Mills

R. S. L.

(1990). Maternal beliefs about adaptive and maladaptive social behaviors in normal, aggressive, and withdrawn preschoolers. Journal of Abnormal Child Psychology, 18, 419–435.

47.

Rudy

Grusec

J. E.

(2006). Authoritarian parenting in individualistic and collectivistic groups: Associations with maternal emotion and cognition and children’s self-esteem. Journal of Family Psychology, 20, 68–78.

48.

Schaffer

H. R.

Crook

C. K.

(1980). Child compliance and maternal control techniques. Developmental Psychology, 16, 54–61.

49.

Shek

D. T.

(2007). Perceived parental behavioral control and psychological control in Chinese adolescents in Hong Kong: A replication. Adolescence, 42, 569–74.

50.

Stafford

Bayer

C. L.

(1993). Interaction between parents and children. Newbury Park, CA: SAGE.

51.

Stern

S. B.

Rohner

R. P.

Sacks-Stern

B. J.

(2007). Relations of self-reported maternal acceptance and behavioral control with acting-out aggressive behavior in children diagnosed with ADHD. Psychological Reports, 101, 675–684.

52.

Strassberg

Dodge

Petit

G. S.

Bates

J. E.

(1994). Spanking in the home and children’s subsequent aggression toward kindergarten peers. Development and Psychopathology, 6, 445–461.

53.

Timmer

S. G.

Borrego

Jr Urquiza

A. J.

(2002). Antecedents of coercive interactions in physically abusive mother-child dyads. Journal of Interpersonal Violence, 17(8), 836–853.

54.

Verhoeven

Junger

van Aken

Deković

van Aken

M. A. G.

(2010). Parenting and children’s externalizing behavior: Bidirectionality during toddlerhood. Journal of Applied Developmental Psychology, 31(1), 93–105.

55.

Werba

B. E.

Eyberg

S. M.

Boggs

S. R.

Algina

(2006). Predicting outcome in parent-child interaction therapy-Success and attrition. Behavior Modification, 30, 618–646.

56.

World Bank (2010 ). GNI per capita, PPP (current international $). Retrieved from http://data.worldbank.org/indicator/NY.GNP.PCAP.PP.CD?%25page=6

57.

Robinson

C. C.

Yang

Hart

C. H.

Olsen

S. F.

Porter

C. L.

. (2002). Similarities and differences in mothers’ parenting of preschoolers in China and the United States. International Journal of Behavioral Development, 26, 481–491.