Abstract
The present study aimed to assess the interaction between familial and cultural factors on child problem behaviors of learning disabled children aged between 7 and 14 within the perspective of ecocultural theory (N = 90). Mothers completed the Family Support Scale, Portrait Values Questionnaire, and Social-Demographical Form, and teachers completed Specific Learning Disability Symptom Checklist and Child Behavior Problem Scale. Pearson correlation, regression analysis, and Hayes’ moderation analysis were conducted. After testing the relationship between support resources and problem behaviors, support analyses were repeated after including learning disability (LD) level. Finally, the effect of cultural values on perceived support resources and child problem behaviors after controlling child LD level was investigated. The findings of this study revealed that ideas about child development depend on cultural and individual factors, and a culturally sensitive understanding of child behaviors can guide researchers in developing more effective intervention programs, particularly for children with LD.
Learning disability (LD) is one of the nonnormative life events that children are diagnosed usually after starting primary school. LD is a biologically originated, neurodevelopmental disorder that demonstrates cognitive abnormalities, impairments in verbal and nonverbal information processing of brain, and/or disruption in processing abilities of individuals manifested by significant difficulties in the acquisition and use of listening, speaking, reading, writing, reasoning, and/or mathematical abilities (Diagnostic and Statistical Manual of Mental Disorders [DSM-5]; American Psychiatric Association, 2013; Goldstein, 2011; Heiman & Berger, 2008; Raghavan & Patel, 2005). LD affects 5% to 15% of school-age children, and 4% of adults in the world (APA, 2013; Goldstein, 2011). Due to the considerable amount of children who are affected by LD, assessing these children and their families becomes imperative to increase both the understanding of their development and the quality of services offered to them. Although there are subtypes of LD that are defined in the DSM-5, LD was included as a composite score in the analyses, including difficulties in reading, writing, mathematics, nonverbal, and verbal language in the present study.
Children with LD face different developmental outcomes due to both biological (e.g., neurocognitive and adaptive deficits; Antshel & Joseph, 2006) and environmental factors (e.g., going to special education center; Lyytinen et al., 2001). A review of the literature shows that many researchers examined the influences of contextual factors (e.g., familial and school) on children with LD including socioemotional level (Al-Yagon, 2012), academic functioning (Al-Yagon & Mikulincer, 2004), and close relations (Bonifacci, Storti, Tobia, & Suardi, 2016). Although there have been many studies examining the effect of contextual factors on LD (Heiman & Berger, 2008; Hubert, 2006; Sandy, Kgole, & Mavundla, 2013), there are virtually limited studies that combine the influence of different factors such as familial and cultural factors. As a result, the present study aimed to integrate the effects of familial and cultural factors on problem behaviors of children with LD within the framework of ecocultural theory.
Ecocultural Theory
The previous studies investigating disabled children and their families were criticized for including mostly univariate variables, distal measurements of family characteristics, and being pathology-oriented (Bernheimer, Gallimore, & Weisner, 1990). However, the new social and ecological views trivialized old approaches, conceptualized the disability as a multidimensional issue, and guided the development of comprehensive applications for disabled children and their families (Bernheimer et al., 1990). Ecocultural theory is one of these new approaches that integrates family ecology, members, and culture into one ground (Worthman, 2010). Ecocultural theory (Weisner, 1997, 2002a, 2002b) assumes that familial factors (e.g., family specific support resources) and cultural factors (e.g., values) organize and shape family activities, and influence child developmental outcomes such as child daily living activities, communication skills, and developmental status of disabled children (Bernheimer et al., 1990; Weisner, 2002b). For example, it includes the family-constructed meaning of the families’ circumstances (e.g., the goals and values influencing family perception of child disability), and also their responses to these circumstances. Based on this theory, Nihira, Weisner, and Bernheimer (1994) formed 12 ecocultural factors (e.g., integration into nondisabled networks, multiple services use, and availability) via home interviews of disabled children’s families, predicting 30% to 60% variance of the child outcomes such as child socialization skills (Nihira et al., 1994). Given the importance of these ecocultural factors on child outcomes, the present study covers seven of these factors (e.g., socioeconomic status, multiple service usage, variety and amount of formal and instrumental help, use of information from professionals, supplemental help for family, help available within family, connectedness of family) using the family support resources scale (Kaner, 2003, 2010).
To clarify, socioeconomic status was described as income in Nihira and colleagues’ (1994) study and studied as financial support in the present study. Likewise, multiple service usage (accessibility and utilization of services for disabled child), variety and amount of formal and instrumental help (usage of variety of supports form professionals, programs or partner), and use of information from professionals (effort of information-seeking for child prognosis and well-being) were studied as informational support. Moreover, supplemental help for family (additional help to family in caring child related activities such as relatives or grandparents) and help available within family (availability of help to caring of disabled child within the family such as husband or sibling) were studied as caregiving support. Besides, connectedness of family was described as the quality of relationship between parents and father participation and help in disabled child care, studied as intimate relations support. In addition to these resources, emotional support was also added to the present study. This support captures the availability and satisfaction of emotional support (e.g., sharing one’s anxiety, feelings, happiness with someone) taken from close relatives, friends, and so on. In all, these specific support resources were thought to be helpful for families to adapt to the disability, shape family activities and parenting practices, and provide a protective context for the family and disabled children (Morrison & Cosden, 1997; Nihira et al., 1994; Sandy et al., 2013).
Child Problem Behaviors
Child adjustment is defined as adaptation to particular contexts (e.g., school environment) and related experiences (Rothbart & Bates, 2008). It includes both positive behaviors such as hope and effort and adaptive difficulties such as problem behaviors. In the LD literature, internalizing and externalizing problems are mostly studied problems (Al-Yagon & Mikulincer, 2004; Batum, 2007; Michaels & Lewandowski, 1990), because they are both prevalent (Greenham, 1999; Nelson & Harwood, 2011), and adversely affect the performance of children on cognitive and academic tasks (Nelson & Harwood, 2011). Thus, child problem behaviors were taken as the outcome variable. Besides, although problem behaviors seem to be a more prominently studied outcome in LD literature, there appears to be a lack of research integrating the effects of specific support resources factors on problem behaviors. Although there were limited studies, informational support (Greenspan & Widener, 2004; Rothman & Cosden, 1995), financial support (Wagner & Blackorby, 1996), intimate relations, and emotional and caregiving support (Bristol, Gallagher, & Schopler, 1988; Sandy et al., 2013) were found to be critical for families and increase child well-being and adjustment.
Furthermore, in Turkey, studies investigated mainly the importance of social support (Ozsoy, Ozkahraman, & Callı, 2006), the relation between perceived social support and the mother’s, child’s (Akıncı, 2011), or family’s well-being (Atalay, 2013) or stress (Kaner, 2004). It was seen that the studies conducted in Turkey about disability mainly overlooked the differential effects of specific support resources on child problem behaviors. Thus, different from previous studies, the effects of specific support resources were investigated separately as indicated in ecocultural theory.
Values
More broadly, ecocultural theory assumes that the actions of people (Bernheimer et al., 1990), such as parenting practices and families’ daily activities, are influenced by both cultural and individual values, and values are the product of the culture (Weisner, 1998, 2002a, 2009). According to Schwartz (1999), “cultural values represent the implicitly and explicitly shared abstract ideas about what is good, right and desirable in a society.” In Schwartz’s (1999) values theory, 10 basic needs were defined. They have dynamic, both similar and contrast relation between each other, and they constitute a circular model (Schwartz, 2006; Schwartz & Bardi, 2001). These values are situated in four poles on the circular model. According to studies, values in the first pole that is related to individualistic values are defined as “openness to change” and values in the opposite pole that is associated with collectivistic values are defined as “conservation” (Demirutku, 2007; Schwartz, 1990), and these two poles were used in the present study.
Values and Support
Values are assumed to influence the family activities, resources, and routines of the family in ecocultural theory (Weisner, 2002b). The literature was mainly based on how values shape the individual’s tendency or chance to get the support. While some researches argued positive relation between collectivism and a better quality of social support from the others (Triandis, Leung, Villareal, & Clack, 1985), others argued negative relation between them (Feather, Woodyatt, & McKee, 2012; Goodwin, Costa, & Adonu, 2004). To our knowledge, how values shape the support resources of the family was not studied before for children with LD. However, researches stressed a general tendency of people perceiving more social support in ethnic minority families (Harry, 2002) and when they internalized collectivistic values (Miletic, 1986). As a result, the values may play a key factor on how these specific groups’ families perceive social support resources.
Values and Parenting
Link of the values with different cultural interests also canalized the researchers to study the relation of values with family or parenting practices and with child outcomes. In studying parenting practices differences and culture, the dimension of individualism–collectivism is one of the cultural parameters that have been used to determine the tendency of societies or/and individuals (Kagitcibasi, 1996, 2009). To begin with, collectivistic cultures is mostly related with child dependency, using more parental control, in which child is mostly perceived as the main source of old age security, and authoritarian parenting was the most observed parenting style (Demirutku, 2007; Kagitcibasi, 1996, 2009; Rudy & Grusec, 2001). However, in individualistic cultures, authoritative parenting is the most observed parenting style (Tudge, Hogan, Snezhkova, Kulakova, & Etz, 2000). This type of parenting was accepted as including firm control, encouraging autonomy, and willingness to reason and negotiate in a more effective familial environment (Garcia & Gracia, 2009). On the contrary, while some studies compare the cultures as individualistic and collectivistic (Schwartz, 1990; Triandis, 1993), some researchers argue that both values could coexist in some cultures and/or individuals such as in Turkey (Green, Deschamps, & Paez, 2005; Kagitcibasi, 1996, 2009; Tamis-LeMonda et al., 2008). In these societies/individuals, there is a combination or coexistence of individual and group/family loyalties, because both values are accepted as basic needs of individuals (Kagitcibasi, 1996, 2009). It was argued that instead of separation of values, internalizing both values leads to balance of parental control and warmth in parenting, in turn positively influencing the child (Kagitcibasi, 1996, 2009). However, all this literature was based on normally developing children and their families.
Although there were very limited number of studies in LD, some studies found that especially mothers of LD children were high on conformity values (Miletic, 1986) and were more rigid (Heiman & Berger, 2008). Failure of the child was seen as the result of an unsuccessful parent–child relationship instead of the child’s own duty in collectivistic cultures such as China (Tews & Merali, 2008). On the other hand, mothers who internalized individualistic values were found to believe more in the importance of early development in infancy instead of relying on cultural beliefs, which is mostly seen in traditional cultures (Harry, 2002). According to Weisner (2002a), all these findings demonstrate that it was not accurate to generalize social differences because child rearing values, parent–child relations, and child development are effected by broader contextual variables such as the family, the neighborhood, the socioeconomic system, and values as stressed in ecocultural approaches. Because coexistence of individual and group/family loyalties is more prominent in Turkey (Kagitcibasi, 1996, 2009), we used Schwartz’s values to see the effect of coexistence of values on specific support resources and problem behaviors of children with LD.
The Present Study
Taken together, the research presented here was driven by the motivation to incorporate familial and cultural factors in studying problem behaviors of children with LD within the perspective of ecocultural theory. Moreover, different from Nihira and colleagues’ (1994) study, child disability level was studied as the control variable.
In this study, the relations were examined in four main steps: (a) the relation between support resources and child problem behaviors, (b) the relation between support resources and child problem behaviors after controlling for disability level, (c) the effect of coexistence of values on perceived support resources after controlling child disability level, (d) the influence of coexistence of values on child problem behaviors after controlling child disability level.
The following hypotheses were tested within the present study:
Method
Participants
LD children
The sample was composed of 128 mothers of LD children attending special education centers and the children’s teachers. However, after missing scales and outliers were excluded, 90 cases reported by both mothers and teachers were included in the study. There were 42 LD girls and 48 LD boys. The mean age of children was 9.31 (SD = 1.79). Children’s mean age of taking the LD diagnosis was 7.93 (SD = 1.48), and the mean of duration of children going to special education was 14.29 months (SD = 16.06). All children also continue to their formal education. The data were collected from different counties of Ankara using the convenience sampling method. All 90 children had been only diagnosed with LD through previous psychoeducational evaluations conducted by Guidance and Research Center of Turkey. All children receiving the LD diagnosis manifested an average IQ level and showed markedly lower achievements or rating in standardized measures of reading, writing, and/or mathematics.
The mothers
Mothers who were married, and had only one biological disabled child were included in the study. The mean age of the mothers was 37.85 (SD = 7.30), and the mean of mothers’ duration of marriage was 16.29 years (SD = 7.03). Moreover, 86.7% of the mothers were not employed (n = 78), while 13.3% of them were employed (n = 12). Besides, of this group of mothers, 11.1% were illiterate (n = 10), 54.4% were graduated from primary school (n = 49), 23.3% were graduated from secondary school (n = 21), 6.7% were graduated from high school (n = 6), 3.3% were graduated from open school (n = 3), and 1.1% graduated from university (n = 1). Regarding perceived income, 41.1% of mothers were from low income (n = 37), 55.6% were from middle income group (n = 50), and 3.3% were from high income (n = 3).
Measurements
Mothers completed two self-report measures and social-demographic information form, and teachers completed only one measure.
Mother-Related Measurements Completed by Mothers
Social-demographic information form
Mother’s age, education level, number of children, working status, perceived family income, duration of marriage, diagnosis of the child, and child’s age, gender, age of getting the diagnosis, duration of special education (in months), other diagnosis (to exclude the child if there were any other diagnosis, for example, attention problems) were included as the demographical questions in the form.
Family support scale
Family support scale was developed by Kaner (2003) to determine social support resources of parents with disabled children. The scale measures support resources of parents or caregivers of intellectually disabled, deaf and blind, psychically disabled children and children having speech disorders between ages 1 and 18. There are five dimensions of the scale: emotional support (α = .91, “When I need to talk, I believe that there is someone who will listen to me”), informational support (α = .86, “There is someone who will give me information about characteristics of my child”), caregiving support (α = .90, “There is someone who will help me in taking care of my child”), intimate relations support (α = .85, “There is someone with whom I can share my most important decisions in life”), and financial support (α = .91, “There is someone who can help me when I have financial difficulties”). The Cronbach’s alpha coefficients for the present study were found for .92; .89; .84; .87; .81, respectively, and total scale’s alpha is .95. Scale was composed of 31 items, and parents choose one of the four options for each item (4 = always, 3 = sometimes, 2 = rarely, and 1 = never), whether they have anybody who supports them. The mothers can get an overall score ranging from 31 to 124. In the revised version of the scale, there is also a question that asks whether “they are satisfied with that specific support resource” (Kaner, 2010). The participants need to answer as 4 = very satisfied, 3 = satisfied, 2 = a little satisfied, and 1 = not satisfied. The mothers can get an overall score ranging from 31 to 124. Due to high correlation between both having support and satisfaction from support (r = .92, p < .00), the aggregated scores of them was used in the analyses.
Portrait values questionnaire (PVQ)
PVQ was developed by Schwartz et al. (2001) to assess individual’s values. This scale includes 40 short verbal portraits of hypothetical individuals, describing one’s goals, or wishes that point implicitly the importance of the given value. The participants should rate the level to which the portrayed person on each item was similar to themselves on a 6-point scale (6 = “very much like me” to 1 = “not like me at all”). It includes four poles, but two poles were included in the present study: The values in the openness to change pole, stimulation (“He looks for adventures and likes to take risks. He wants to have an exciting life”) and self-direction (“It is important to him to make his own decisions about what he does. He likes to be free to plan and to choose his activities for himself”); and the values in the conservation pole, security (“It is important to him to live in secure surroundings. He avoids anything that might endanger his safety”), tradition (“Religious belief is important to him. He tries hard to do what his religion requires”), and conformity (“It is important to him always to behave properly. He wants to avoid doing anything people would say is wrong”), were used. Demirutku (2004) adapted the scale to the Turkish population and found that Cronbach’s alpha coefficients ranged between .63 (tradition) to .75 (conservation). In the present study, the two poles were used as subscales, and alpha coefficients were found as .53 for openness to change pole and .78 for conservation pole.
Child Related Measurements Completed by Teachers
Specific Learning Disability Symptom Checklist (Teacher Form)
The checklist was developed by Erden (2016) to assess reading, writing, learning, attention, and math abilities of the learning disabled children (Özkök & Erden, 2011; Sarıpınar & Erden, 2010). It includes 78 items and four subscales as understanding and using verbal and nonverbal language (e.g., “She or he reads some paragraphs repeatedly in order to understand”), sustaining attention (e.g., “She or he gets bored quickly during the lesson”), failing in reading and writing (e.g., “She or he likes writing”), and coping with mathematical concepts (e.g., “She or he knows multiplication table by heart”). Teachers completed this checklist demonstrating the level of the frequency of the difficulties the child faces in his or her abilities for each domain on a 5-point scale (5 = very extremely to 1 = not at all). The Cronbach’s alpha coefficients were found as .90 and higher for the subscales. For the present study, total score of the checklist was included in the analysis, and Cronbach’s alpha coefficients were found as .96 for total scale.
Child behavior problem scale (Teacher Form)
The scale was developed by Kaner (2007) to detect problem behaviors of disabled children and the level of discomfort of the teacher related to the problem behavior. The scale includes 45 items and four subscales as conduct problems, immature and internalizing problems, hyperactivity and externalizing problems, and attention problems. The Cronbach’s alpha coefficients were found .75 and above for the subscales. The teacher was asked to rate the student behaviors in the list (e.g., internalizing problem “Introverted,” externalizing problem “Restlessness, hyperactivity”) on a 4-point scale (3 = “I feel extreme discomfort” to 0 = “The child doesn’t perform the behavior”). In the present study, only immature and internalizing problems, and hyperactivity and externalizing problems subscales were used. The Cronbach’s alpha coefficients were found .86 for immature and internalizing problems, and .87 for hyperactivity and externalizing problems.
Procedure
First, the ethical permission was taken from Hacettepe University Ethic Committee. Then, the data were collected in the 2015–2016 fall semester. The children who continue to special education centers were selected, and the participants were contacted by the help of three special education centers. The aim and goals of the study were discussed with mothers via phone, and a meeting was arranged. The families were interviewed by the researcher in their homes or in the special education center according to mothers’ preferences. Besides, teacher forms with an informed consent were delivered to teachers through mothers. The criterion for teacher evaluations was to know the child at least for 3 months. The teacher forms were collected from the mothers in the second short visit.
Data Analysis
All data were analyzed in IBM SPSS23. To see relationships between family supports, values, child LD level, and child problem behaviors, Pearson correlation analyses were conducted. Next, regression analyses were conducted for prediction of problem behaviors, and then to see slope of interaction, a moderation analysis was conducted that was recently introduced by Hayes (2013).
Results
Descriptive Analysis of Study Variables
Descriptive results for mothers’ and teachers’ reports were summarized in Table 1.
The Mean and Standard Deviation of Study Variables.
Preliminary Analyses
Before starting analysis, the ipsatisation method was used to correct individual differences in scale use. Each person’s responses were centered by subtracting their rating of values from their own mean score in order to convert absolute scores into scores that indicate the relative importance of each value in the value system (Schwartz, 1992, 1996). Values have effects on cognition, emotion, and behavior through an arrangement among multiple values that are simultaneously relevant to action, so ipsatising is both desirable and preferable in analysis (Schwartz, 1996; Tetlock, 1986). Moreover, Fischer (2004) stresses the congruity of ipsatising value scores for the Schwartz value theory instruments. That is why, in the present study, it was chosen to conduct the analyses with ipsative scores. For this sake, the study variables were also centered on their means.
Correlations between variables
Bivariate correlations were examined (see Table 2). More markedly, income level was positively correlated with specific support resources. Moreover, duration of education and marriage negatively correlated with child LD level and internalizing problems. Support resources were not correlated neither conservation values nor openness to change values. LD level was negatively correlated with support resources except financial support (r = −.19, p > .05). Unexpectedly, only informational support was negatively correlated with internalizing problems, and emotional support, caregiving, and financial social support were negatively correlated with externalizing problems. Furthermore, conservation values were not correlated with both internalizing and externalizing problems. However, openness to change values was marginally correlated with only externalizing problems. While child LD level was positively correlated with child internalizing problems, child LD level was not correlated with child externalizing problems. Finally, child’s gender, age of getting the diagnosis, number of children, and age of the mother were not correlated with any study variable, and that is why they were excluded from the analyses.
Bivariate Correlations Between the Study Variables (N = 90).
Duration of special education was measured in months.
p < .10. *p < .05. **p < .01.
Hierarchical Regression for Perceived Social Support Resources and Internalizing Problems
Results for internalizing problems without controlling for LD level
Hierarchical regression analyses were conducted for each support resource and internalizing problems separately (H1). First, a two-step model was conducted for internalizing problems of children with LD. In Step 1, internalizing problems were regressed on child age, income level, mothers’ education level, and duration of marriage. However, only duration of marriage was a significant predictor, so others were excluded from the model. In Step 2, specific support resources were added separately. Only perceived informational support significantly predicted internalizing problems. That is why only the results for informational support were presented. As seen in Table 3, both the duration of marriage (adjusted R2 = .06) and perceived informational support negatively predicted internalizing problems (adjusted R2 = .09).
Summary of Hierarchical Regression Analysis for Perceived Informational Support Predicting Internalizing Problems (Not Controlling For Learning Disability Level).
p < .05.
Results for internalizing problems after controlling for LD level
LD level was added to the models and analyses were repeated after controlling child LD level (H2). As seen in Table 4, duration of marriage and LD level did make significant contribution to the explained variance in the equation (adjusted R2 = .09). While the effect of duration of marriage negatively and marginally predicted, LD level significantly and positively predicted internalizing problems. When perceived informational support was added, it did marginally account for additional variance in predicting overall internalizing problems (adjusted R2 = .12). Therefore, mothers’ perceived informational support become a marginal and negative predictor of internalizing problems after adding LD to the model.
Summary of Hierarchical Regression Analysis for Perceived Informational Support Predicting Internalizing Problems (Controlling For Learning Disability Level).
p < .10. *p < .05. **p < .01.
Hierarchical Regression for Perceived Social Support Resources and Externalizing Problems
Results for externalizing problems without controlling for LD level
Two-step model was conducted for each support resource and externalizing problem separately (H1). In Step 1, externalizing problems were regressed on child age, income level, mothers’ education level, and duration of education (in months). However, only duration of education (in months) was a significant predictor, so others were excluded from the model. In Step 2, specific support resources were added separately. Perceived emotional, caregiving, and financial support were found significant. That is why only the results for emotional, caregiving, and financial supports were presented, respectively.
To begin with, for emotional support, as seen in Table 5, while duration of education positively and marginally predicted (adjusted R2 = .03), perceived emotional support negatively predicted externalizing problems (adjusted R2 = .07).
Summary of Hierarchical Regression Analysis for Perceived Emotional Support Resources Predicting Externalizing Problems (Not Controlling for Learning Disability Level).
p = .06. *p < .05.
Moreover, the two-step model for perceived caregiving support was significant. For this analysis, in Step 1, only duration of education (in months) was included in the model. The variable did make marginally significant contribution to the explained variance in the equation (R2 = .03), and the duration of education (in months) marginally and positively predicted externalizing problems. When perceived caregiving support was added, it accounted for additional variance in predicting overall externalizing problems (adjusted R2 = .07). The effect of caregiving was significant and negative on externalizing problems (β = −.22, p < .05).
Next, for perceived financial support, at Step 1, only duration of education (in months) was added to the model. The variable did make marginally significant contribution to the explained variance in the equation (R2 = .03), and the effect of duration of education (in months) marginally and positively predicted externalizing problems. When perceived financial support was added, it accounted for additional, but marginal variance in predicting overall externalizing problems (adjusted R2 = .05). Perceived financial support marginally and negatively predicted externalizing problems (β = −.18, p = .09).
Results for externalizing problems after controlling for LD level
LD was added to the models in the first step to control its effect, and analyses were repeated after controlling LD level of the child (H2). When LD level was included in the model, for caregiving and financial support, only disability level explained the variance for externalizing problems. For emotional support, however, both emotional and LD level become nonsignificant in the model. Therefore, an interaction effect between perceived emotional support and LD level was suspected. To investigate a possible interaction effect further, a third step of interaction was added to the model for emotional support (see Table 6). The model was found to be marginally significant (adjusted R2 = .09). When the interaction of both variables was added, it marginally and positively predicted externalizing problems. Furthermore, to see how this interaction has taken place, a test of simple slope recently introduced by Hayes (2013) was run. The result indicated that perceived emotional support tended to be a negative predictor of externalizing problems for children showing mild (−1 SD) difficulties in LD (−.7193, SE = .14, p = .01). Moreover, the relation was negative for children showing average difficulties in LD (.0009, SE = .09, p < .05). In contrast, although for mild and average scores, it seemed to be a negative predictor, for children showing severe (+1 SD) difficulties in LD score, the relation was not significant (−.7212, SE = .11, p = .73; see Figure 1). When the child shows severe difficulties, the effect of perceived emotional support tends to disappear.
Summary of Hierarchical Regression Analysis for Perceived Emotional Support Resources Predicting Externalizing Problems (Controlling for Learning Disability Level).
Interaction of learning disability and perceived emotional support.
p = .06. *p < .05. **p < .01.

The interaction between child’s LD level, mother’s perceived emotional support, and externalizing problems.
Hierarchical Regression for Values
First, a three-step model was conducted for specific support resources (H3). In Step 1, supports were regressed on income level, mothers’ education level, and LD level (only these variables were correlated with the outcome; see Table 2). In Step 2, conservation and openness to change values were added separately in models. Finally, the interaction of both values was added in Step 3. However, conservation values, openness to change values, and the interaction of values (internalizing both values) did not predict any perceived support resources after controlling LD level.
Second, the relationship between values and problem behaviors of learning disabled children was inspected (H4). Therefore, three-step hierarchical regression analyses were conducted for internalizing problems. In Step 1, only LD level was included since it was a significant predictor. In Step 2, openness to change and conservation values were added. Finally, the interaction of both values was added in Step 3. As seen in Table 7, LD level was marginally positive and significant in predicting internalizing problems (adjusted R2 = .07). When openness to change and conservation values were included in hierarchical regression, they did not explain the variance of internalizing problems (adjusted R2 = .05). Besides, the interaction marginally and negatively predicted internalizing problems (adjusted R2 = .07). To examine this interaction further, a test of simple slope recently introduced by Hayes (2013) was run. The results demonstrated that it tended to be a positive relation between internalizing problems and openness to change values for mothers who have low (−1 SD) conservation values (.6016, SE = .12, p = .28). The relation seemed to become negative for people with average score in conservation values (–.0329, SE = –.02, p = .98), however this relation is not significant. Moreover, the results demonstrated that it tended to be a negative relation between internalizing problems and openness to change values when mothers have high (+1 SD) score in conservation values (−.6674, SE = –.12, p = .32; see Figure 2).
Summary of Hierarchical Regression Analysis for Mothers’ Values Predicting Internalizing Problems.
Interaction between openness to change and conservation values.
p = .07. *p < .05. **p < .01.

The interaction between mothers’ values and internalizing problems.
Besides, the three-step hierarchical regression analysis was also carried for externalizing problems as had been done for internalizing problems. However, there was no relationship found between mother’s values and externalizing problems.
Discussion
The aim of the present study was to examine the relationship among mothers’ perceived social support resources, values, and child problem behaviors. The findings were discussed considering each independent variable separately (perceived social support resources and mothers’ values). After discussing main findings, limitations, and suggestions, implications of the study were presented.
Support Resources and Internalizing Problems
The results indicated that only perceived informational support negatively predicted internalizing problems. Moreover, informational support was marginally and negatively predicted after controlling for LD level. In literature, it was mainly argued that information and support taken from professionals were generally inadequate in these families (Sandy et al., 2013; Tews & Merali, 2008), and especially argued to be important in dealing with behavioral, educational, and emotional needs of the children (Greenspan & Widener, 2004; Hodapp, Dykens, & Masino, 1997). To our knowledge, there were no studies showing a direct relation between perceived informational support and internalizing problems. However, to answer why this support was only related with internalizing problems imposes a critical point. According to Hooven, Gottman, and Katz’s (1995) study conducted with normally developing children and their families, maternal coaching (e.g., being aware of child emotion, help the child to learn the ability to deal with their emotions in a supportive manner) was related with teachers’ reports of low levels of internalizing problems, but not related with externalizing behaviors. Gottman, Katz, and Hooven (1996) stated that maternal coaching does not simply promote children’s expressions of emotions, but rather encourages children to control their negative feelings and emotions (e.g., anxiety), to calm down and focus their attention (Eisenberg et al., 2001). Although this study was conducted with normally developing children, the same could be also applied to children with LD. Perceived informational support could help parents how to effectively deal with disability and to understand child emotions related to LD. This may result in guiding the child about regulating their negative emotions and learning to express their feelings, which in turn result in lower internalizing problems. Moreover, this support seems to be additionally critical because it was still marginally significant even after controlling for LD level. This support should be taken into account in studies and interventions.
Support Resources and Externalizing Problems
Perceived emotional and caregiving support significantly, and financial support marginally predicted externalizing problems, but none of them predicted externalizing problems after controlling LD level. Nevertheless, only the interaction of perceived emotional support with LD level marginally predicted externalizing problems. In the literature, it was generally stressed that emotional, financial, and caregiving support resources were crucial for children with LD (Brookins, 1993; Sandy et al., 2013; Wagner & Blackorby, 1996). However, although the relation between specific support resources and child development was studied in developmental delayed children (Nihira et al., 1994), it was not studied in detail in LD literature.
First, perceived emotional support negatively predicted child externalizing behaviors in this study. It was stressed that caregivers of children with LD needed emotional support (Sandy et al., 2013), and cohesion and support in the family structure influenced child problem behaviors (Morrison & Cosden, 1997). On other hand, to our knowledge, there were no studies indicating a direct relationship between mothers’ perceived emotional support and externalizing problems of children with LD. This finding could be explained through an indirect path. Expressing emotions is likely to be related with parent’s positive expressivity with the child in the family context (Bristol et al., 1988; Eisenberg et al., 2001). In other words, as caregivers of learning disabled children reported that they had feelings of anger, frustration, and helplessness (Sandy et al., 2013), expressing negative feelings to their friends and family members or receiving emotional support may lead parents to calm down or help them to regulate their negative feelings (Belsky, 1984). Because these mothers have the opportunity to share their negative emotions with their friends or relatives, they will be better in coping with the stress associated with the disability. In turn, these mothers may create a warm environment for their disabled children. For example, in normally developing children, researchers found that parent’s emotional support (e.g., warmth) negatively predicted children externalizing problems, but not internalizing problems in the family context (Eisenberg et al., 2001; Mccarty, Zimmerman, Digiuseppe, & Christakis, 2005). Furthermore, when LD level was added to the model, an interaction between perceived emotional support and LD level was detected. The path of the relation between LD level and emotional support showed that when the disability level was increasing, perceived emotional support effect on externalizing problems seemed to disappear. This finding indicated that even support is effective in mild and moderate level of disabled children, and when the situation becomes more complicated, emotional support for the mother may not be sufficient in dealing with externalizing problems of children.
Second, the findings for perceived caregiving support and financial support were in line with literature (Dunst, Leet, & Trivette, 1988; Nihira et al., 1994; Waggoner & Wilgosh, 1990). Similar to emotional support, parents who take help in child care may have more time for their work-related activities, in turn decreasing their sense of isolation. Moreover, financial support would create the chance for the child in attaining services such as additional educational help or help in satisfying daily needs of the child. With the decrease in financial or child care related stress, these mothers may focus more on the disability and cope better with child externalizing problems. Although these findings were consistent with literature, these studies did not control disability level.
Furthermore, contrary to literature (Bristol et al., 1988), intimate relations support did not predict child problem behaviors in the present study. To explain this finding, it can be stated that both parents were exposed to the same situation within the family. The role of intimate relations is not rejected, but it is agreed with Dunst et al.’s (1988) proposal that external support resources may be more critical for them in coping with disability.
In sum, when the LD level was included in the analysis, the predictive power of the supports on child problem behaviors both decreased or became nonsignificant. There are some points that should be underlined. First, although support might be a protective factor for families and child problem behaviors, disability level seems to be also a prominent predictor of problem behaviors in these children (Rothman & Cosden, 1995). Second, as Lyon (1996) claimed, focusing on individual needs of students may be a more effective approach to meet the wide range of needs, choices, and abilities of students who participate in special education when they have severe symptoms. Finally, as not analyzing the severity level was emphasized as a limitation (Heiman & Berger, 2008), our findings show the important role of LD level especially when the relation among familial or contextual factors and child outcomes are investigated.
Values and Support Resources
In contrast to Hypothesis 3 and literature (Feather et al., 2012; Goodwin et al., 2004; Triandis et al., 1985), values did not predict perceived support resources. There could be several reasons. First, the relatively small sample size might have led to attenuation in the sizes of correlation coefficients may be one of the factors. Second, children with LD are a specific group. Therefore, the factors that determine these families’ support resources can be related to other factors. For example, poverty and economic problems were seen as crucial factors in attaining services (Brookins, 1993; Sandy et al., 2013). As a result, this assumption may be replicated in future studies.
Values and Internalizing Problems
It was found that neither conservation nor openness to change values lead to internalizing problems. However, children showed least internalizing problems whose mothers have high scores in both values as hypothesized, after LD level was controlled. The present finding about the relation between values and internalizing problems is also in line with the earlier works with normally developing children (Kagitcibasi, 2009; Rudy & Grusec, 2001, 2006), indicating both promotion of autonomy and relatedness in psychological/emotional interdependent families.
Heiman and Berger (2008) argued that parents of children with LD may be more rigid, less open to change, and have difficulties in adapting to increased demands related to LD such as lower academic achievement. This may lead to dysfunction of family and negatively influencing the child (Morrison & Cosden, 1997). As found in this study, mothers who have internalized both values may provide a supportive environment. Children can express their feelings more easily in the family context where mothers have internalized openness to change values, and support their child’s autonomy (Kagitcibasi, 2009). In addition, internalization of conservation (relatedness) values, that is maintaining close ties, will provide a supportive environment for the child and help family members cope with increased demands of the disability. In all, this environment is speculated as an optimal environment for a child with an LD as it promotes more flexibility in the case of changes and healthy growth.
Values and Externalizing Problems
Contrary to internalizing problems, there was no relationship found between values and externalizing problem behaviors. In the literature, externalizing problems were mostly observed when attention deficit hyperactivity disorder (ADHD) was comorbid with LD (Batum & Oktem, 2011; Willcutt & Pennington, 2000). However, ADHD could be a confounding variable in predicting externalizing problems because the observed behaviors in ADHD are also mostly measured in externalizing problems scales. Therefore, in these children, externalizing problems were likely to be observed more than internalizing problems. As comorbid cases were excluded from the analyses, externalizing problems might have been observed less than internalizing problems in the present study. Thus, comorbidity of other diagnoses, child severity, or other unknown variables rather than values may explain or could be an underlying factor for externalizing problems.
Limitations and Suggestions
To begin with, the sample size of the present study was small, and a larger sample size would increase the power to test more complex models. Second, because behavioral problems information was taken from teachers via mothers, it resulted in a smaller sample size of teacher’s reports than mothers’ reports. Third, by its definition, values are socially desirable characteristics. Moreover, as the items of the scale are abstract in nature, it might make it harder to interpret the scale for the mothers in the present study. Future research would benefit from including longitudinal component in a more diverse sample (e.g., intellectual disability, autism) while exploring different demographic variables (e.g., marital status, birth order, etc.) in a larger sample.
Implication
The main contribution of the present study is to provide empirical evidence to the relationship between contextual factors and child problem behaviors of children with LD. To our knowledge, there are no studies investigating the relationship between coexisting values and specific perceived social support resources for children with LD, especially in Turkey. This study partially supported this claim of the theory even after the child LD level was controlled. Especially, findings for values indicated that the importance of a culturally sensitive understanding of concepts can help researches develop more effective prevention and intervention programs. Moreover, intervention programs should include additional educational, behavioral, and psychosocial help especially for children showing more severe symptoms because the quality and effectiveness of interventions such as being sensitive to child disability level is a critical point for child development (Lyon, 1996). Besides, future research may benefit including LD level when examining relationship between predictor and outcome variables. Finally, assessing the relations with multiple informants enabled us to take different perspectives into account, and helped to capture a more comprehensive picture and to eliminate a possible single source bias.
Conclusion
To our best knowledge, this present study is one of the first studies that investigate the relationships among cultural factors (values), familial factors (specific supports), and problem behaviors of children with LD. Moreover, these findings add to our understanding of the role of support in outcomes of children with LD by (a) determining the importance of including different support resources, (b) providing partial evidence for the effect of coexistence of values, (c) demonstrating the importance of including child LD level in analysis, and (d) suggesting the crucial effect of cultural and individual factors on child development.
Footnotes
Acknowledgements
We are grateful to the families who participated in this project.
Authors’ Note
This article was the product of the student’s master’s thesis by supervision of PhD, Berna Aytac.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
