The informal and professional supports available for families of children with mental retardation in Korea and the United States were examined and the effects of culture and social support on maternal stress considered. Thirty-eight American and 40 Korean mothers participated in the home-visit interview. A pilot study identified 10 types of informal and 11 types of professional support for the two national groups. Results showed that American mothers received more informal and professional support in almost all domains of social support; Korean mothers experienced more stress. Possible explanations for these differences were explored.
Editor in charge: Steven J. Taylor
Families of children have mental retardation often face challenges because of the special demands placed upon them. The families typically make a strenuous effort to organize the family environment and to reach out to obtain resources that will meet the needs of their child. Successful utilization of resources has been important for the families in coping with the challenges and everyday demands in caring for a child with mental retardation (Beresford, 1994; Hentinen & Kyngas, 1998; Rimmerman & Duvdevani, 1996; Thyen, Kuhlthau, & Perrin, 1999). In particular, social support influences the successful adaptation of such families (Boyce, Behl, Mortensen, & Ackers, 1991; Dunst, Trivette, & Cross, 1986; Greenberg, Seltzer, Krauss, & Kim, 1997; Llewellyn, McConnell, & Bye, 1998; McConachie, 1994; Trivette & Dunst, 1992).
Although the role of social support has been widely documented in mental retardation family research, very little information is available concerning the role of culture in determining the experiences and the effect of social support (Stoneman, 1997). Different cultural backgrounds may lead families to look for different types of support as a reflection of their attitudes and the availability of the support system. In the present study I documented the comparative experiences of social support for families of children with mental retardation from two distinctively different countries—Korea and the United States. I focused on specific types of informal and formal support available to families of children with mental retardation in each country, their utilization of the resources, and the perceived quality of the support. In addition, mothers' reports were used to examine the effects of the particular nation and social support on mothers' well-being, controlling for child characteristics.
There are a few comparative studies conducted between Korea/Asia and the United States on families of children with mental retardation. Oh (1984), in her investigation of Korean families of such children, examined family functioning and its relationship to social support. When compared with normative data on randomly selected American mothers of typically developing children (Norbeck, Lindsey, & Carroero, 1983), Korean mothers of children with mental retardation had a smaller social support network. Frequency of contact was significantly lower than the United States norm. The social support network of the families comprised primarily kin groups and a very limited number of formal and professional supports.
In their cross-cultural study, Nihira, Tomiyasu, and Oshio (1987) compared Japanese and American families of children with mental retardation. They found that the relation between cognitive opportunities at home and the child's social competence appeared similar in the two cultures, but the relation between the affective aspects of the home environment and the child's psychosocial adjustment differed. The Japanese families had higher levels of conflict, emotion, and psychosocial problems. They and their children with mental retardation experienced more stress than did their United States counterparts.
Specification of similarities and differences between cultures/countries would add important information on the universality and uniqueness of the experiences of these affected families. In this study my goal was to provide additional information on this issue by examining how social support that is shaped by cultural context influences the psychological well-being of these mothers.
In describing the differences between Korean and Western cultures, researchers have often used the model of collectivist versus individualist orientation (Cha, 1994; Han & Ahn, 1994; Triandis, 1989). Korean culture is said to be collectivist, emphasizing family values, meeting expectations of families through achievement, obedience, and conformity to social rules. Western culture is more individualistic, focusing on meeting the needs of individuals through achieving personal rather than family goals and bringing honor to the individual rather than to the family (Cha, 1994; Han & Ahn, 1994).
Because of different cultural orientation, families in both countries also vary in their help-seeking behavior. In a collectivist culture such as Korea, when members of a family face problems, they tend to depend on the family for problem-solving instead of looking for outside sources of help (Hurh & Kim, 1988). The same principle applies to Korean mothers of children with mental retardation. That is, they are more likely to rely on family members to help them take care of their child with mental retardation, whereas American mothers would be more likely to depend on professionals.
The regular education system in Korea consists of 6 years of elementary school, 3 years of middle school, 3 years of high school, and 4 years of college. The special education schooling system follows the same pattern but also includes kindergarten and 1 to 3 years of specialized vocational training courses. According to the Special Education Promotion Act of 1977, as many as 18 years of free mandatory special education is guaranteed to all students with disabilities from kindergarten to specialized vocational training (Korea Research, 1997). Special schools are further classified according to type of disability (e.g., schools for individuals who are blind, hearing impaired, physically handicapped, mentally retarded, or emotionally disturbed). In 1997, there were 114 special schools nationwide in Korea (Korea Research, 1997). Special classes (e.g., all-day, half-day, and supported learning) are also offered at regular elementary, middle, and high schools to provide integrated education to students with disabilities. The students who receive these services have various disabilities, including visual impairment, hearing impairment, physical handicap, learning disability, emotional instability, and speech impairment (Korea Research, 1997). Since the Special Education Promotion Act was enacted in 1977 and revised three times (1978, 1987, 1997), the special education system in Korea improved a great deal. More special schools have been built and special classes expanded so that more students with disabilities receive the kind of education they need.
Education for children with disabilities in Korea became mandatory under this Act; however, children with disabilities still receive a substandard education (Korea Research, 1997). According to the findings of the Korean research group, an estimated 2.4% of all children in the 5- to 17-year age range needed special education. The special school and class enrollment ratio was very low: Only 42% of all entitled students were enrolled in special education. Among those enrolled, 54% had severe disabilities and attended special schools. The children in special classes at regular schools represented about 35% of all children with mild disabilities. Korean researchers also noted that those children who were not enrolled in special education programs were either accommodated in welfare facilities or attended regular classes at regular schools, where their individual capabilities and difficulties were not taken into consideration. In addition, only about half of graduates from elementary school level special schools or special classes graduated from middle and high school. There were far fewer special classes or schools established at the middle school than were available at the elementary school level (Korea Research, 1997).
In 1977, a total of 307 mothers who had a child with mental retardation were recruited from elementary schools in two metropolitan areas in the United States and Korea; 78 (38 American and 40 Korean mothers) participated and completed the interview.
The participating mothers ranged in age from 31 to 50 years (M = 39). Most American mothers were married (87%); 5 were divorced, widowed, separated, or living with a partner without marriage (13%). All but one of the Korean mothers were married (98%); the other was divorced (2%). High school completion was the highest education for 42% of American mothers; 53% had completed at least 2 years of college. High school graduation was the highest educational level for 40% of Korean mothers; 30% had at least a 2-year college education. Overall, the American mothers had significantly more education than did Korean mothers, t(76) = 2.77, p < .05, but there was no significant difference in education between the American and Korean fathers.
The combined mean number of children in the family was 2.6 (range = 1 to 5). When the number of children in the family ranged from 1 to 3, there was no difference in birth order in children with mental retardation. However, in families whose number of children ranged from 4 to 5, the child with mental retardation was either third or fourth in birth order. The average age of the children with mental retardation was 9.7 years (range = 4 to 14). There was no statistically significant age difference between the two groups of children. On the short version of the Adaptive Behavior Scale—ABS (Nihira, Foster, Shellhaas, & Leland, 1974), which was used to assess the level of adaptive functioning of children, there was no significant difference between Korean and American children. Likewise, there was no statistically significant difference in level of maladaptive functioning as measured by the Maladaptive Behavior Scale from the Developmental Disabilities Profile (Brown et al., 1986).
A measure of social support for families of children with mental retardation
A social support scale was developed by adapting Norbeck's Social Support Questionnaire (Norbeck, Linsey, & Carrieri, 1981; Stewart, Hirth, Klassen, Markrides, & Wolf, 1997) for its conceptual base and format. The open-ended format of the scale was considered relevant in a cross-cultural context because certain types of support may exist in one culture, but not the other. When the respondents are provided with an open-ended format in which they list all types of support, the scale is able to incorporate any type of support available to the mothers. Social support consists of two properties based on the scale: structural and functional support. The structure of social support is defined as the number of types of support and the duration of relationships and the frequency of contact with people who provide support. There were also three aspects of functional support: affect, affirmation, and practical aid (Norbeck et al., 1981).
I conducted a pilot study to examine the relevance of the scale in the two countries and to develop a scale that was appropriate in both cultural contexts. Twenty mothers from the two countries were asked to list those who were available to help in raising their child. They were then asked to rate each type of support on the domains of structural and functional support. Throughout the pilot study, social support was divided into two types of support, informal and professional. Informal support was defined as help or service that was not paid for; when the help was based on fee or operated through funded services for people with disabilities, it was categorized as professional support. Based on this categorization, 10 types of informal support and 11 types of professional support were identified (Table 1).
Several domains of social support were identified from the scale as important to mothers: number of sources, frequency of contact, emotional support, informational support, and financial support. These domains were considered for informal and professional support (financial support was considered only for informal support). The domain for duration of relationship with support providers was not included in this study because there was high turnover among professionals. The affirmation domain was dropped as not relevant for Korean mothers with traditional upbringing: For this domain mothers were asked whether others agreed with their opinions, and many Korean mothers had puzzled looks, implying that they wondered whether they could have ever had opinions. Informational support was added because it emerged as important during open-ended questions throughout the pilot study. Altogether, the number of sources of support, frequency of contact, emotional and informational support, and financial aid (only for informal support) were considered in the main study. New questions were developed for each domain of social support to reflect the mothers' experiences with support in caring for their child. The new scale was named the Social Support Scale for Families of Children With Disabilities (hereafter called the Social Support Scale).
Participants in the main study checked the availability of each type of support in the list prepared during the pilot study. Although a certain type of support may exist (e.g., spouse or mother-in-law), the mothers may not consider it as available in helping them care for the child with mental retardation. For each type of support that was endorsed as available, the participants were also asked to rate frequency of contact with the person (or group) who provided the support (from 1 [daily] to 5 [once or twice a year]), emotional support (1 [none] to 4 [a great deal]), informational support (1 [none] to 4 [a great deal]), and financial aid (1 [none] to 4 [a great deal]). Composite scores for each domain were generated by adding the scores in each domain.
To determine reliability of the Social Support Scale, I used factor analysis for each type of support and examined the patterns of the extracted factors. The scale showed reliable factor patterns on emotional, informational, financial, frequency of contact, and number of supporting people across different types of support. However, the functional and structural properties of support were not separated; these two constructs were highly correlated. Thus, they were not considered separately in further analyses.
Validity of the scale was checked by using the Multidimensional Scale of Perceived Social Support—hereafter called Scale of Perceived Social Support (Cecil, Stanley, Carrion, & Swann, 1995; Zimet, Dahlem, Zimet, & Farley, 1988). Reportedly, this scale has adequate internal reliability (.88) and stability over time (.95). The construct and convergent validity of this scale has also been demonstrated (Cecil et al., 1995; Stanley, Beck, & Zebb, 1998; Zimet et al., 1988). The Scale of Perceived Social Support was correlated significantly with all domains of informal support of the Social Support Scale, r =.42 to .59, p < .05, except the financial support for Americans, and with all domains for the Korean group, r =.39 to .50, p < .05. However, neither correlated with any domains of professional support, except informational support for the Americans, nor with any of domains of professional support for Koreans. The Scale of Perceived Social Support itself consists of items on perceived helpfulness of informal support. By highly correlating with the scores in almost all domains of informal support, the developing scale demonstrated the concurrent validity.
A measure of adaptive and maladaptive functioning of the child
To measure adaptive functioning of the children in the study, I administered a shortened version of the AAMD Adaptive Behavior Scale—ABS (Godfrey et al., 1986; Nihira et al., 1974). This scale yields three factors of social competence: Personal Self-Sufficiency, Community Self-Sufficiency, and Personal-Social Responsibility. The reliability of the scale was .89 for the American group and .91 for the Korean group.
The Maladaptive Behavior Scale from the Developmental Disabilities Profile (Brown et al, 1986; Jacobson, 1998), which lists 13 inappropriate and aggressive behaviors, was used to assess maladaptive functioning of the children. The internal consistency of the scale was .72 for the American group and .60 for the Korean group. The criterion-related validity of the scale was demonstrated in other studies, yielding significant differences in mean score across settings, with problem behaviors decreasing from less to more restrictive settings (Brown et al., 1986).
A measure of maternal stress
Friedrich's Short-Form of the Questionnaire on Resources and Stress was used to measure the mothers' experienced stress (Friedrich, Greenberg, & Crnic, 1983; Glidden & Floyd, 1997). For this study, the 30 items were used that comprise the components of the parent and family problems and pessimism. The scale correlated significantly with the Beck Depression Inventory (Friedrich et al., 1983). Overall reliability of the scale in this study was .92 for Americans and .89 for Koreans.
Translation of the Instruments
Because of political, social, economic, and cultural differences between Korea and the United States, the samples from both cultures were expected to differ on many dimensions, possibly putting measurement equivalence at risk. The translation procedures were designed to obtain content and semantic equivalence for all five scales. Content equivalence is established when the content of each item of the instrument is relevant to the phenomena of each culture being studied (Liang & Bogat, 1994). All the scales were developed in the United States except for the Social Support Scale, which was developed in both countries. Reliability and validity establish the relevance of the scales to American culture. Three experts (one Korean and two Korean Americans) rated each item on each scale except the Social Support Scale as relevant, irrelevant, or questionably relevant. They judged all of the items as relevant to Korean culture, so all questions were retained in the measures.
Semantic equivalence of measurements was established following the back-translation technique, suggested by Brislin (1970) and Liang and Bogat (1994). First, a bilingual translator translated each measure into Korean. Then, another bilingual person back-translated the instruments from Korean to English. Finally, a third bilingual person rated each item on a 3-point scale, where 1 was exactly the same meaning in both versions, 2 was almost the same meaning in both versions, and 3 was different meaning in each version. Items that were rated as different were considered for rewording of the items. Reworded items were examined by repeating the same back-translation technique. With one exception, all items in the Adaptive Behavior Scale were retained as having the same meaning to both cultures, with the exception of the item concerning the child's level of skill in using a knife, fork, and spoon. It was dropped because Korean children learn to use chopsticks instead, and the implements seemed to demand different levels of skill.
To control variables that may affect outcomes, I used the following inclusion criteria: First, the mothers had to be able to understand the interview content and follow instructions. Second, as an attempt to control race and ethnicity variables, only European Americans without a recent immigration history were selected among American mothers. Third, the children were elementary-school age. Finally, the children were within the range of mild to severe mental retardation. Mothers were excluded if they and/or their children did not meet all these criteria.
The mothers in the Chicago sample were recruited from both city and suburban areas. In the city of Chicago, the author obtained a Chicago public schools data book (Hess, 1994), identified 75 public schools that had at least 30% European American children, and contacted administrators of these schools to locate European American families of children with mental retardation. Initially, school principals were contacted for research approval. Out of 75 schools contacted, 40 did not have European American children with mental retardation, 5 schools did not respond to three follow-up phone calls, and 14 schools declined to participate in the research study; 16 schools participated. The number of mothers contacted ranged from 1 to 20 per school. Of the 103 mothers contacted, 22 (21%) participated in the interview, and 18 (17%) met all the criteria for inclusion in the study.
In the Chicago suburban areas, a directory listing special education service administrators was obtained from the Illinois State Board of Education, and directors of district offices of special education were contacted by mail to obtain approval for research. The general rule was to contact district offices of special education nearest to the city first, reaching out to farther suburban areas until enough potential participants were identified and contacted. Altogether, nine school districts were contacted; seven of them agreed to participate. The number of mothers contacted through district offices ranged from 7 to 42. Of 135 mothers contacted, 27 (20%) participated in the interview, and 20 (15%) met the criteria for inclusion in the study.
Unlike the United States, there is no urban and suburban distinction based on economic status in metropolitan cities in Korea. Communities that stand out based on economic status tend to be scattered throughout each city and district. Pusan is one of the metropolitan cities in South Korea with about 250 elementary schools, with an average of 32 classes per school and 39 students per class (Bureau of Statistics, 1996). Almost all the schools were public schools, and special education services were mandatory in the public schools.
Public schools in three of five districts in the city were contacted. Out of 43 regular schools contacted, 15 had no participants who met the criteria, 7 declined to participate, and 1 did not respond to follow-up calls. From the 20 schools that participated, 25 mothers were contacted. Of these, 21 (84%) participated in the interview and 19 (76%) met the study criteria.
The children from special classes in regular schools had learning disabilities or mild mental retardation; therefore, an attempt was made to recruit more children within the range of mild and moderate mental retardation from the two special education schools in the city exclusively for children with mental retardation. The school principals were called initially for research approval. Out of 44 mothers contacted through the two schools, 22 mothers (50%) participated in the interview; 21 met the inclusion criteria (48%).
A total of 78 mothers (38 American, 40 Korean) met the study criteria and completed the interview. There was a discrepancy between the participation rate (16% for Americans and 58% for Koreans) due to differences in contact procedures. In the United States, most initial contacts with mothers were made by mail through teachers. In Korea, most initial contacts were made by phone by teachers. Teachers who called mothers elicited a much higher consent rate. The author, who is bilingual in Korean and English, conducted the home-visit interviews in both countries.
Types and Availability of Informal and Professional Support
When the availability of informal support was compared, there was no significant difference between the two groups, with the exception of friends. American mothers had more friends available to help them care for their children than did Korean mothers, χ2 = (1, N = 78) = 14.2, protected p < .001. When the availability of professional support was compared, American mothers reported more school therapists, χ2 = (1, N = 78) = 51.2, protected p < .001, school personnel, χ2 = (1, N = 78) = 12.0, protected p < .001, doctors, χ2 = (1, N = 78) = 24.9, protected p < .001, and recreations programs, χ2 = (1, N = 78) = 24.0, protected p < .001, available for their children than did Korean mothers. Table 1 depicts the types and availability of informal and professional support for the American and Korean samples and their statistical comparisons. The two samples yielded comparable lists of types of informal and professional support.
Contrary to my expectation that the Korean mothers would have more informal support, no difference was found between the two groups of mothers. As expected, the American mothers generally reported more sources of professional support available to them than did the Korean mothers. School therapists and personnel and pediatricians specializing in disabilities were taken for granted by many American mothers, but rarely by Korean mothers. In Korea, many professional services such as speech therapy are not offered at the regular school. Thus, they were not always readily available to many children and the families.
The two groups of mothers used some services at a comparable level, but with different purposes. When American mothers utilized private therapists, they were mostly speech therapists. Korean mothers, however, were more likely to hire tutors who worked with their children on educational materials. Although the current use of family support programs was comparable between the two groups, many American mothers reported having used these programs to secure needed information and resources when their children were very young, whereas Korean mothers rarely recognized those programs at all. Although there were nationally and locally organized parental support programs in both countries, fewer of the Korean mothers were aware of collective efforts to make changes in special education systems.
The proportion of availability of informal support within each group was examined. These percentages show the proportion of each support by national group, not the comparison between the groups. These are different statistics from those reported earlier.
For the American mothers, the largest proportion of support was from the child's sibling (18%), followed by husband (17%), parents (11%), mother's sibling (11%), friend (9%), mother of child with special needs (9%), relative (9%), and parents-in-law and neighbor (6%). In the case of Korean mothers, the largest proportion of support was from the husband (22%), followed by the child's sibling (20%), mother of child with special needs (16%), mother's sibling (13%), parents (9%), and parents-in-laws (8%). Again, whereas the proportion of informal support was relatively more evenly distributed for American mothers, Korean mothers relied heavily on a network of smaller size, such as their close family members and other mothers of children with special needs.
The proportion of availability of professional support within each national group was examined. The American mothers received the largest proportion of support from special education teachers (21%), followed by school therapists (16%), doctors (15%), recreation programs (14%), school personnel (9%), private therapists (9%), and babysitters (6%). On the other hand, Korean mothers relied on special education teachers for the most support (39%), followed by private therapists (18%), regular education teachers (16%), and doctors (7%). Many more school-based services and other services for children with disabilities were available for the American mothers, whereas Korean mothers relied heavily on educators at school and their private resources.
Comparisons of Informal and Professional Support Between the Two Nations
I expected that Korean mothers would depend on informal support to a greater extent than would American mothers. Informal support was examined in terms of five domains: number of sources of support, frequency of contact, emotional support, informational support, and financial support. Hotelling's T2 test was conducted on the two groups of mothers to measure the differences between them on the five domains of informal support. Mother's education was included in the analysis as a covariate given that American mothers had a higher level of education than did Korean mothers, t(76) = 2.77, p < .05. Significant differences were found between the two groups of mothers on the Social Support Scale. Specifically, American mothers had more informal support than did Korean mothers overall, F(4, 72) = 20.53, p < .05. A series of univariate analysis of variance revealed that there were statistically significant differences between the two groups on number of sources of support, emotional support, and informational support, F(1, 75) = 4.38, p < .05; F(1, 75) = 4.62, p < .05; and F(1, 75) = 11.64, p < .001, respectively. The American mothers had more individuals who provided informal support than Korean mothers had, and they drew more emotional and informational support from them. No statistically significant differences were found between the groups with respect to the frequency of contact support and financial support. Although Korean mothers maintained equivalent frequency of contact with the people who provided support, American mothers elicited much more emotional and informational support from their contacts.
I predicted that American mothers would depend on professional support to a greater extent than would Korean mothers. Professional support was examined in terms of four domains of social support: number of sources of support, frequency of contact, emotional support, and informational support. Hotelling's T2 test, with mother's education as a covariate, was conducted on the two groups of mothers to measure the differences between them on the four domains of professional support. American mothers also had more professional support than did Korean mothers, F(4, 72) = 10.15, p < .001. Univariate analyses of variance revealed that there were significant differences between the two groups on number of sources of support, frequency of contact with support, emotional support, and informational support, F(1, 75) = 23.78, p < .05; F(1, 75) = 12.62, p < .05; F(1, 75) = 26.04, p < .001; and F(1, 75) = 33.42, p < .001, respectively. The American mothers had more professionals who provided support than did Korean mothers, and they maintained more contacts with them. They also drew more emotional and informational support from the professionals than Korean mothers did.
Relationships Among Social Support, Child and Mother Characteristics, and Stress
When both groups of mothers were compared on the stress measure (Questionnaire on Resources and Stress-SF), Korean mothers reported more stress than did American mothers, t(76) = 4.2, p < .001. On the perceived social support measure (Scale of Perceived Social Support), Korean mothers also perceived less social support than did American mothers, t(76) = 6.4, p < .001.
The variables that were assumed to be associated with certain domains of informal and professional support, such as stress, perceived social support, and child (adaptive and maladaptive functioning) and mother (education) characteristics, were examined within each group. Tables 2 and 3 present means and standard deviations (SDs) for the variables for the American and Korean groups. When the stress variable was considered in relation to social support, stress experienced was not related to any domain of social support but was significantly related to perceived social support for both groups. On the other hand, the perceived support was correlated with all domains of informal and professional support for both groups.
Correlations between the domains of informal support and those of professional support were also examined for each group (Tables 2 and 3). For American mothers, informational support gained from informal support was significantly correlated with the three domains of professional support, frequency of contact, emotional support, and informational support. The mothers who received informational support from informal resources tended to contact professionals more often and draw more emotional and informational support from them. However, the remaining informal support variables were not related with any domains of professional support. For Korean mothers, the informational domain from professional support was correlated with all domains of informal support. Those mothers who received informational support from professionals were more likely to have informal resources and frequent contact with them and to draw emotional and informational support. However, the remaining informal support domains were not related to any domains of professional support.
For American mothers, child's level of adaptive and maladaptive functioning was significantly related to mother's experience of stress. Also, mothers with children who had severe maladaptive behavior and less adaptive functioning perceived less social support. However, these relationships were not found among Korean mothers.
For American mothers, those with higher education were more likely to receive professional support in all support domains. They also had more informational support from informal sources. Similarly, Korean mothers with higher education had more support from professionals in all domains of support, had more people who provided informal support, and received more emotional and informational support from these people.
Predictors of Maternal Stress
Factors related to maternal stress were considered together to predict the level of stress experienced by mothers. These factors were nationality, the child's level of adaptive and maladaptive functioning, informal and professional support, and perceived social support as the predicting variables and stress as the dependent variable. The variables in each domain of both informal and professional support were highly correlated with one another; thus, composite scores were created for informal and professional support. The scores for each domain were standardized, and factor analyses confirmed that the domain variables were clearly loaded on the two professional and informal support factors, respectively. These standardized scores were added to make the composite scores for informal and professional support.
Path analyses were conducted using multiple regression given that certain variables were assumed to take on mediating roles among the independent variables, thus showing hypothetical causal relationships. For example, although the level of maternal stress was not related to any specific domain of informal and professional support, the level of stress was significantly related to perceived social support. On the other hand, perceived support was correlated with all domains of informal and professional support for both groups. The specific ratings on each domain might not be sufficient to explain the stressful experience, but overall perceived quality of support could mediate between stressful experiences and different components of social support.
The independent variables were entered in the order of nationality in the first block, child's level of adaptive and maladaptive functioning in the second block, informal and professional support in the third block, and perceived social support in the fourth block, with stress as the dependent variable.
To determine whether the model of maternal stress differed between the two groups of mothers, I tested interactions of nationality with each of the other independent variables by adding them to the regression equations. No significant interactions were found, so nationality is treated as a main effect in the common model for the two countries. The results of the final path model are shown in Figure 1. This path model presents only those direct paths that showed a .05 level (one-tailed test) of statistical significance. An examination of Figure 1 suggests that child's level of adaptive functioning had a direct impact on maternal stress but also affected it indirectly through its effects on informal support and perceived social support. Mothers of children with less adaptive functioning experienced more stress than did the other mothers, due to lack of informal support and lower perceived quality of general support. Child's maladaptive behavior also had direct impact on maternal stress, but it also had an indirect effect through perceived social support. Specifically, mothers of children with more maladaptive behavior were more stressed than were the other mothers because they had lower perceived quality of support.
When the nationality of the mothers was considered, Korean mothers were more stressed due to lack of informal support and less perceived quality of support. Informal support indirectly related to maternal stress via perceived social support. The results indicate that mothers' perception that they receive support is important. Korean mothers relied on professional support much less than did American mothers, but, overall, professional support was not related to maternal stress.
The hypothesis that Korean mothers, coming from a collectivist society, would rely on informal support to a greater extent than did American mothers was not confirmed. This result may reflect several factors. Korea is a dynamically changing society, becoming more mobile and modernized; thus, the informal support network may not be as available as it used to be to these mothers. In fact, many mothers in the study reported geographic segregation from their parents or relatives. In addition, during interviews some mothers did not even want to talk about their child with their own mother or mother-in-law for fear that any discussion of the affected child might worry them. This is in contrast to the traditional cultural practice in which married women rely on their own extended family network in caring for their children. In some cases, relationships with in-laws deteriorated because the daughters-in-law carried a child with disabilities, which was perceived to bring shame to the family. In addition, because of negative messages they received from those in their informal network, including relatives, neighbors, and friends, Korean mothers were not open about their child's disability with those outside of their families. As a consequence, the mothers perceived them as less helpful than expected.
The Korean mothers seem to be doubly trapped because of (a) the society's negative attitudes toward disabilities that prevent them from being open about their children and sharing the experiences with people around them and (b) the loss of social support network associated with the society's mobility. As previously documented in other cross-cultural studies, the negative attitudes and stigma the society carries toward people with disabilities seem to contribute to the lower level of support the mothers receive in Korea (Lee & Kang, 1997; Moon, 1994; Oh, 1984). The social stigma discouraged these families from being open about their affected children and from relying on informal sources of support.
The hypothesis that American mothers depend on professionals more than do Korean mothers was confirmed. The two cultures were comparable in terms of the types of professional support because all professional services in the United States also existed in Korea. However, many professionals were far less available in Korea because professional services are still in the early stages of development. The higher level of support in all domains of professional support rated by American mothers suggests that the children and their families in the United States received a higher quality of services.
A path analyses was used to examine the impact of social support on the psychological well-being of the mothers. The path analyses results were generally consistent with the previous findings regarding the influence of child characteristics on the experienced stress of the mothers. Mothers of children with less adaptive functioning and more behavior problems experienced more stress (Cameron & Orr, 1989; Floyd & Gallagher, 1997; Suarez & Baker, 1997; Walker et al., 1992). The path analyses additionally revealed that although the adaptive functioning level of children was related to receiving informal support (those informal support providers could probably handle children with higher functioning better), the child's maladaptive level of functioning was related to the perceived support rather than sources of support. Regardless of sources of support, these mothers needed to feel that they and their child were cared for.
The role of informal support, professional support, and perceived support were examined for their mediating effects to predict maternal stress in these path analyses. The analyses confirmed the role of social support as a mediating factor in maternal experiences of stress (Dunst, Trivette, & Cross, 1986; Quittner, Glueckauf, & Jackson 1990; Suarez & Baker, 1997), implying that when social support occurs, mothers can experience a lower level of stress. The positive effects of social support on mothers' well-being found here confirms the results of previous research (Dunst, Trivette, & Cross, 1986; Dunst, Trivette, & Jodry, 1996), but also demonstrate the mediating influences of perceived social support between sources of support and well-being of mothers. This proves that people who provide support to mothers of children with mental retardation need to be aware of how the mothers perceive their support and that their perceptions may eventually have an impact on their stressful experiences. The importance of perceived quality of social support also further confirmed that perceived satisfaction with support is one of the most important aspects of social support in determining the psychological outcome (Barrera, 1981; Barrera & Ainlay, 1983; Stanley, Beck, & Zebb, 1998). Korean mothers perceived less social support than did American mothers, and this perceived lack of social support affected Korean mothers, who reported higher levels of stress than did their American counterparts. The mothers' perception that they are cared for and helped would then be very important in relation to their experience of stress, along with the informal sources of support that, in turn, had a direct influence on perceived support.
These comparative findings provide practical implications for those who work with mothers of children with mental retardation in Korea or culturally similar societies. Although certain child and maternal characteristics affect perceived stress, other factors may also negatively impact Korean mothers, such as traditional beliefs about disability and negative social experiences. Although changing beliefs and attitudes could be a difficult and challenging task, assertiveness training could help mothers in altering their attitudes. Considering the traditional upbringing of many Korean women, lack of assertiveness has been a barrier for them in seeking resources, information, advice, and services for their children. Training them to be assertive could be the most important influence in assisting these Korean mothers to voice their children's rights for education and services and to be more forceful in working with professionals as partners rather than recipients of services. Assertiveness training may have a long-ranging impact on changing traditional beliefs of these Korean mothers.
Although Korean mothers experience severe social tensions, and most of them have to carry the burden of attending to their children with mental retardation as sole caregivers, service workers could try to engage the help of all potential supporters within the family and neighborhood rather than demanding more of already overly stressed mothers (McConkey, 1995). Education, training, and involvement of siblings, fathers, and other relatives could provide some solutions that would not only relieve the mothers of the caregiving burden but also help to prevent the isolation of the children, their mothers, and their family in the community. The improved relationships and understanding about the child's condition would also contribute to developing positive attitudes toward children with mental retardation.
In this study, certain aspects of support were noted as especially important, such as informational and perceived quality of support. In particular, informational support was found to be one important factor among the domains of social support. Considering that provision of information was related to, and possibly could impinge on, other aspects of support, education, training, and dissemination of information continue to play very important roles in helping these families. When parents have information, they will have access to more services and to more information and will be able to work with professionals more assertively.
Informal support was related to perceived social support but not professional support. The scale that assessed the perceived quality of social support consisted of questions addressed mostly to informal sources of support, which might have contributed to a lack of association between professional support and perceived support. Because perceived social support emphasizes shared feelings and the perception that others will be present when help is needed, it will be important for professionals to enable the families to feel that they are being helped and cared for, probably by taking the time to let them talk about their concerns or by showing empathy for their experiences. Encouraging them to meet and talk with other parents could also prove valuable in helping them to gain perceived social support.
The present study may be limited by sampling methods and geographic locations. I depended on mothers who volunteered to participate in the interview, so results may be representative only of mothers who were willing to share their experiences regarding their children with mental retardation. Geographic limitations were unavoidable in selecting samples from each country. Mothers were recruited from major metropolitan areas in each nation. In Chicago, a major effort was made to include families from city and suburban schools. Mothers living in suburban areas or close to suburban areas, however, were more willing to participate than were those from the inner city. Therefore, it appears that middle- and upper-middle-class families were overrepresented in this study. Also, the American sample contained European American mothers without recent immigration history, two-parent families, and those who worked mostly part-time or who were stay-at-home mothers. Although the dramatic differences between two countries may justify the conclusions, caution should be taken in drawing generalizations from this study. Because the path analyses were based on correlation coefficients that described two related variables in a cross-sectional study, caution should be exercised in drawing conclusions on causal relationships regarding these mothers' experiences,
The present study is one of the few systematic investigations in which the different experiences of families of children with mental retardation was examined cross-culturally. I attempted to systematically measure the efficacy of social support without ignoring the qualitative sources of support unique to each culture. Although certain findings confirmed previous research findings, such as the effect of child variables and the mediating effect of perceived quality of support on the well-being of mothers, I found that Korean mothers experienced more stress, mainly because of the lack of informal support and perceived quality of support.
Although further cross-cultural study could provide a more in-depth understanding of the issues of families of children with mental retardation in different cultural contexts, in this study I attempted to illustrate two groups of mothers from two countries where culture makes distinct differences in terms of family functioning and child-rearing practices. An extended comparative study on other ethnic families would further enhance the understanding of the role of social support, perceived support, and experienced stress as affected by culture and different social support systems.
NOTE: This study was part of the author's dissertation submitted to the University of Illinois at Chicago in partial fulfillment of requirements for the doctoral degree. The manuscript was written during a postdoctoral fellowship funded by the National Institute on Disability and Rehabilitation Research. The author thanks Gershon Berkson, Kathleen Crittenden, Christopher Keys, Robin Miller, Tong-He Koh, Leonard Newman, Suzanne Feetham, and Judy Curry for their invaluable comments. The author also thanks the school personnel and the families who participated in the study.
Jin Y. Shin, PhD, Assistant Professor, New College, 130 Hofstra University, Hempstead, NY 11549 ( Jin.Y.Shin@Hofstra.edu). The author is also affiliated with the New York State Institute for Basic Research in Developmental Disabilities