The basis for an examination of the relationship between family well-being and its antecedents is found in diverse fields (Haines, Hurlbert & Beggs, 1996; House, Umberson & Lambert, 1988; Kahneman, Diener, & Schwarz, 1999; Wellman & Wortley, 1990). Personal characteristics, including level of human capital, affect family well-being directly and indirectly through their effects on social support. Teachman, Paasch and Carver (1997) suggested that both financial and human capital must be accompanied by social relationships to enhance well-being. Better general health and lower depression were associated with higher levels of support (Israel, Farquhar, Scholz, James, & Parker, 2002; Robertson, Elder, Skinner, & Conger, 1991). In their study of recovery from setbacks, Parks-Yancy, DiTomaso, and Post (2007) found that having access to social and financial capital resources reduced the socio-economic impact of the setbacks and was of greater importance than the setback itself in determining outcomes for both men and women.

Pearlin (1985, p. 44) defined a network as "the outer boundaries of supports upon which an individual can draw." Characteristics of the network refer to the size of the network and the types of relationships among members, the characteristics of members, and the resources embedded within the network (Lin & Dumin, 1986; Wellman & Wortley, 1990). Network characteristics appear to be stronger determinants of social support than personal characteristics (Wellman, 1982). Some systems of support may be used for all exigencies, while other networks are specialized for the type of problem for which support is needed (Pearlin, 1985). Wellman and Wortley (1990) found that parent-child ties and ties that are strong (i.e., are intimate, voluntary and mutual) provide a broad range of support. Networks with a larger proportion of kin, given kinship norms of mutual assistance and the proximity of kin, are expected to provide greater support to members (Ahluwalia, Dodds, & Baligh, 1998; Hao, 1996; Miller & Darlington, 2002; Wellman & Wortley, 1990). Edin and Lein (1997) found that for poor mothers, their own mothers were their biggest source of support.

For low-income families, these social networks often respond to economic hardship by increasing the flow of social support to meet immediate needs (Lin & Dumin, 1986; Stack, 1974). These social networks are thought to serve a "coping" function, meeting immediate, day-to-day needs, providing things that they do not have the money to buy (Briggs, 1998), and providing a sense of satisfaction not attributed to support from non-kin (Wan, Jaccard, & Ramey, 1996). The importance of kin in the network is increased through household extension, an adaptive response to economic hardship that involves sharing a residence with at least one person who is not part of the nuclear family. Household extension may improve social support and family well-being through resource pooling (Cohen 2002).

Where people live affects who they know and with whom they interact and, therefore, the resources to which they have access (Hogan, Eggebeen, & Clogg, 1993; Lee, Netzer, & Coward, 1994; Tigges, Browne & Green, 1998). Beggs, Haines and Hurlbert (1996a) argue that personal networks in rural areas, in contrast to urban ones, are composed of stronger ties. Members of rural networks tend to exchange more than one type of resource and are tied to each other through multiple roles.

The impact of social support on family well-being may be manifested in a number of ways, some concrete and objective and others intangible and subjective. Much social support by friends and family is offered in times of need. Tangible support may provide food, clothing, shelter, transportation, and childcare – the goods and services needed for everyday survival of the family. Levitan and Feldman (1991) suggest that inter-household exchanges can enable a family to avoid crises. Social and economic aspects of well-being are often interwoven in social networks (Levitan & Feldman, 1991; Nelson & Smith, 1999). Research suggests that social networks provide people with an underlying sense of security and enhanced well-being beyond the tangible help provided (Kossek, et al., 2003; Pearlin, 1985). Rural low-income mothers often depend on informal support, most of which comes from family members, to help them manage in difficult circumstances. This support may also enhance their general well-being.

For this study, the researcher used 187 cases for which there were data from all three waves of interviews of the multi-state project, NC1011, Rural Families Speak. The participants resided in rural counties in 13 states. They were recruited through programs that serve low-income families, including the Food Stamp Program, the Supplemental Nutrition Program for Women, Infants, and Children (WIC), welfare-to-work programs, and Cooperative Extension programs. To be eligible for participation, families had to have annual household incomes at or below 200% of the federal poverty line and at least one child aged 12 years or younger. Families were selected purposively to represent the types of families with children who would be affected by welfare reform in their counties: they would use similar services, live in similar locations, and/or compete in the local job market with welfare recipients and leavers.

The project design involved collecting rich qualitative data that included the meanings participants attribute to their circumstances in addition to several quantitative instruments. The qualitative data were collected from the mothers in the households by trained interviewers, using a semi-structured interview protocol. In-person interviews in three waves were conducted beginning in the year 2000 in the participants’ homes or in neutral places such as a private room in the community library. Quantitative data involved instruments assessing risk for depression, food security, health, and social support. Interviews lasted from one and one-half to two and one-half hours and were tape-recorded. Quantitative data were coded by a team at one university using participants’ responses to instruments and additional variables from the transcripts. Qualitative data were used to create childhood risk factors and family support variables from Wave 1. As with the quantitative data, the transcripts were coded by a single team of trained project personnel using agreed upon protocols.

Family well-being was defined as the ability of the participants and their children to meet their basic needs and have an overall sense of happiness and security. The four indicators of family well-being were the participant’s risk for clinical depression, the family’s food security, participant’s perception of the adequacy of the family’s income, and her overall satisfaction with her life. All four measures of family well-being were rescaled into percentage indexes, giving each of the outcome variables equal weight in the cluster analysis.

The participant’s risk of clinical depression was measured using a widely accepted instrument, the Centers for Disease Control’s CES-D (Radloff, 1977). A CES-D score of 16 or above signals that one is at risk for depression, with higher scores indicating greater risk of being depressed. Food security is the ability to acquire nutritionally adequate and safe food in a socially acceptable way (Anderson, 1990; Hamilton, Cook, Thompson, Buron, Frongillo, & Olson; 1997). Food security was indicated by the number of items checked in the eighteen-item U.S. Household Food Security Survey Module (Olson, Anderson, Kiss, Lawrence & Seiling, 2004). Food insecurity rises as the number of items checked increases. For use in the cluster analysis, a standard food security index was calculated by multiplying the food security score by 100 and dividing by 18 (number of items).

The other two measures were based on the respondent’s replies to direct interview questions. Respondents were asked how satisfied they were with their life overall, and responses were recorded on a five point scale. For use in the cluster analysis, the responses to the interview question were multiplied by 20 to yield a "standard satisfaction score." Respondents also were asked how adequate they thought their income was, and responses were recorded on a five point scale from 1 meaning "not at all adequate" to 5 meaning "very adequate." The wording of the perceived income adequacy question did not consider respondents’ ability to pay for their needs. For use in the cluster analysis, the income adequacy response was recoded in the same manner as satisfaction.

The clusters were compared using both continuous and categorical variables. Personal attributes included human capital, family structure, and family background. Measures of human capital included respondents’ age, education, physical and mental health, and frequency of successfully managing their resources. Age of respondents was coded in actual years, but education was coded in categories based upon level of education completed. For management of resources, respondents were asked how often they were able to complete four tasks: pay their bills, stick to a budget, stretch groceries to the end of the month, and prepare a well-balanced meal. Physical health was measured by summing a set of sixteen chronic health problems that can interfere with daily life (Sturm and Wells, 2001) in Waves 1 and 2 and using summary scales from the SF-36 Health Survey in Wave 3. In addition to the CES-D score as an outcome measure for risk of clinical depression in Waves 1, 2, and 3, mental health was also assessed using summary scales in the SF-36 Health Survey (Ware, 2008) in Wave 3. The summary scales in the SF-36 are based on 100 points, and higher scores indicate better health.

Family functioning was measured using the Family APGAR scale (Smilkstein, 1978). It measures five parameters of family functioning: Adaptability, Partnership, Growth, Affection, and Resolve. The family members referred to in this instrument are the people within the household, unless one is living alone. Proximity to mother and father was estimated from the qualitative data. Emotional support was considered a proxy for quality of the relationship with mothers and fathers. They were positive, non-instrumental contributions by a parent. These acts could be as vague as "being there when needed" or as specific as "can always be relied on to give me good advice about disciplining the children."

The informal support participants received from their parents was assessed in detail, but there was only an overall indication (Yes/No) of whether any kind of support was given by siblings, other relatives, and/or friends or neighbors. Participants were asked how many persons they could call on for help if needed. Classification of social support types involved detailed specification of instrumental and emotional support from mothers and fathers and to mothers and fathers. Instrumental support was mainly of two kinds: material resources, defined as money and objects and services, such as babysitting or car repair, defined as acts performed by a person in the network for the benefit of the respondent for which there was a paid alternative.

Other variables tested for differences among the clusters were participants’ and partners’ educational attainment, feeling of safety in their living environment, amount of stress felt by the participant, her reported ability to cope with that stress, childhood risk factors that included growing up in a single parent home, experience of abuse or neglect, frequent moving, divorce, substance abuse, leaving home before age 18. Although income and wages were compared for all three years, only Wave 3 income differed among the clusters as did income increasing from W1 to W3.

The richness of the data presented both challenges and opportunities for analysis. Having three waves of data meant that here were many variables of interest and multiple measures for the well-being factors. The selection of K-means cluster analysis would allow the researcher to identify groups of respondents with relatively homogeneous outcomes (or well-being). The four rescaled or "standardized" indicators of family well-being were used in the cluster analysis. The well-being measures were standardized so the differences in scale among the measures would not drive the clustering. The standardization method chosen (see above for more detailed explanation) rescaled the variables to a base of 100 rather than using standard deviations to rescale the variables. In the cluster analysis procedure the number of clusters was set at four to provide opportunities for distinguishing among the groups while having clusters that were large enough to have a significant number of cases.

The four clusters converged on the 12^th iteration. They highlighted some nonlinearities among the well-being measures of interest. Once identified, the clusters were tested for differences in health characteristics; family functioning, proximity, and relationships; informal network size and support; and education, employment, income, household composition, childhood risk factors, level of stress, and feelings of safety using ANOVA and Chi-square.

This study included four measures of family well-being. The clearest and largest difference among the clusters was mental health in the form of the risk for depression as measured with the CES-D. The researcher chose the most salient characteristic of each cluster to label them. Four clusters were specified, and the cluster centers differed significantly by all four measures of well-being for all three years (See Table 1).

The mothers in Cluster A had the lowest scores on all four of the well-being measures. Their means were considerably lower for risk of depression. They were greatly challenged by their physical and mental health and by the stress they experienced (see Table 2). The typical participant had four chronic health conditions in T1 and T2, and almost half of them reported having had an injury or serious illness in the past year for both waves as well. When asked specifically whether they suffered from depression or anxiety, a large majority responded that they did. About half also reported that their health or that of the family members interfered with their ability or their partner’s ability to be employed and health status had an impact on their finances. Cluster A participants had the lowest level of family function and few exchanged emotional support with their mothers or fathers. Almost half of Cluster A participants had fewer than three people to whom they could go if they needed help. Fewer than 1 in 5 participants received instrumental support from family members other than their parents. There was no significant difference among the clusters with respect to number of instrumental support resources they

Table 2: Means and Proportions of Participants’ Characteristics Differing Among the Clusters*

got from their parents. They were not satisfied with their level of support. Over half of the Cluster A participants had completed education beyond high school, the largest proportion among the clusters; however, their partners had the fewest with post-secondary education. The Cluster A group had the fewest who felt safe in their residential environment; they had the most stress and the least ability to cope with it. At T1 these participants had the largest number of childhood risk factors, and almost 40% reported that they had experienced physical, emotional, and/or sexual abuse within the last three years. Fewer than 20% of them were employed in T3; they had the lowest average household income ($18,274); and the fewest who had experienced an income increase since T1.

The most notable feature of the Cluster B participants was that they had higher levels of well-being as they moved from T1 to T3. This improved well-being was reflected in the other factors that were tested for the clusters (see Table 2). They had fewer chronic conditions and fewer of them reported depression/anxiety in T2, although the percent who reported having had an injury or serious illness in T2 was larger than that of T1. Cluster B participants had the highest Family APGAR score and had the largest proportion who received emotional support from both their mothers and their fathers. Over 40% of the Cluster B group had six or more person on whom they could call if they needed help. They did not receive instrumental support from family members other than their parents, but over half of them received support from neighbors and/or friends. They reported a high level of satisfaction with their informal support. Most of Cluster B participants were high school graduates without additional education. They felt safe in their residential environment, had the next to the lowest level of stress and the highest ability to cope with their stress. They had experienced the fewest childhood risk factors and had the smallest percent that recently had experienced abuse at T1. Over half of them were employed, and about 90% of them had had an increase in income between T1 and T3. Their mean income was second largest of the four clusters.

The participants in Cluster C had a moderate level of well-being for Income Adequacy, Risk for Depression and Overall Satisfaction. They had a high level of Food Security. Their health was relatively good, with only about one-fourth of them reporting an illness or injury in T1 and T2 (see Table 2). They reported about the same number of chronic conditions as mothers in Cluster B (half of those for Cluster A). A smaller proportion than those in Clusters A and B reported that health affected their employment (20%) or their financial condition (40%). Over half of Cluster C participants lived with their mothers or within five miles of their mothers. They gave emotional support to their mothers and fathers in the largest proportions, but they got less emotional support from their mothers and fathers than did those in Clusters B and D. It is assumed that they received instrumental support from their mothers, but the results were not significantly different than the other clusters. This group had the largest percent who got support from other family members and friends and/or neighbors, and they were moderately satisfied with their level of support. The Cluster D participants were mixed in terms of their education: over half had completed post-secondary education, but over one-fifth had not finished high school. More than half of their partners had not completed high school. They felt safe in their residential environments. Cluster C participants had a fairly high mean stress level, but they reported they could cope with their stress fairly well. They had a low number of childhood risk factors, but about 15% of them had experienced abuse in the three years prior to T1. Two-thirds of the Cluster C participants (the largest percent) were employed in T3; 80% had an increase in annual income from T1 to T3. Their annual income in T3, $25,012, was lower than those for Clusters B and D, but higher than Cluster A.

The Cluster D participants had the highest level of well-being on all four measures for all waves. It is interesting to note how much smaller was the percentage of participants in that cluster who reported health problems (see Table 2). They had the fewest chronic conditions in both T1 and T2 and the smallest percentage who reported having had in injury or serious illness at both T1 and T2. They also had the smallest proportion who reported that their health affected their ability to work for pay or their financial circumstances. The Cluster D group had the highest level of family functioning. About 40% of them lived with or within five miles of their mothers, and they were second in percentage who received emotional support from their mothers and their fathers. They did not give emotional support to mothers or fathers, however. They had the largest support networks, with more than half of them reporting six or more people on whom they could call if they needed help. About one-fifth to one-third of the Cluster D mothers received instrumental support from their siblings, other family members, or friends/neighbors. They expressed the highest level of support of the four clusters. Half of the mothers had completed post high school education, and their partners were the best educated of the clusters. They felt safe in their residential environments. Cluster D participants had the lowest average stress score and had the next to the highest score with respect to ability to cope with their stress. They had a fairly low score for childhood risk factors, and the next to lowest proportion who had experienced abuse prior to T1. About two-thirds of these mothers were employed at T3. They had the highest annual household income, with a mean of $29,264, and 75% of them had had an increase in their income between T1 and T3.

Cluster A had the lowest level of health in all categories. The scores were especially disparate for the impact of physical and emotional health on their ability to perform everyday roles. The clusters were consistent in their relative health scores across these scales as well as with their level of well-being. The graphs below illustrate the comparisons among the four clusters on the physical and mental health summary scales. It is notable that Cluster A is much lower on most of the scales, while the three other clusters are relatively close to each other.

One’s ability to manage his or her resources depends on a number of factors, including skill (human capital), amount of income and expenses, alertness and ability to focus (which may be diminished with depression or other mental illness), and level of informal support (in-.kind income). Almost 90% of mothers in Cluster D were always able to pay their bills, whereas, only about 20% of those in Cluster A were able to do so (see graphs below). On the other hand, the clusters were closer together for the two skills that related to food security: stretching groceries to the end of the month and preparing a well-balanced meal. Mothers in Cluster A used informal strategies to obtain resources had a higher percentage than did mothers in the other clusters for selling or pawning possessions, bartering or trading services, and using a soup kitchen or food bank. This was probably due to their lower income, inability to work, and smaller support networks. It may also be due to a lower level of management skills. Depression and other health issues may reduce their ability to perform tasks of daily living.

This study involved a purposive sample, so the findings cannot be generalized beyond this group of low-income families. However, there are some interesting results from these rich data that can shed light on the well-being of low-income rural families. The lack of correspondence between the socio-demographic characteristics (income, employment, age, marital status, and family size and composition) and the well-being measures found in this analysis is somewhat surprising.

The most powerful finding is that health, both mental and physical, is strongly associated with the overall sense of well-being for these rural low-income mothers, an impact that can extend years beyond the onset of the health condition (Pearlin, Schieman, Fazio, & Meersman, 2005). Cluster A participants had the lowest scores on all of the outcome variables for all three Waves. In addition, only their level of food security improved, while the other variables got worse over time, making the spread between their scores and those of the other groups even larger in Wave 3. As a group, they had significantly worse health and experienced a much greater impact of both mental and physical health on their ability to perform everyday functions. Their ability to manage their resources was significantly lower as well.

Another interpretation is that the measuring stick against which these mothers compare their current situation may not be other similar families, as it is for objective measures, but it may be their own well-being at an earlier time. Mothers in Clusters B and D experienced improvement in their well-being between T1 and T3, whereas, Cluster A participants had losses in income, loss of jobs, ongoing health problems, and high stress.

Key factors associated with well-being of Clusters B and D were the high level of functioning of their immediate families and the emotional and social support that was available to them. They had the largest networks and the highest satisfaction with their support. Turner (2006) found that employment and support from family and friends reduced the depressive symptoms among rural single mothers. She also found that lifetime traumas and adversity, recent life events, general life strain and parenting strains all contributed to depressive symptoms, but financial strain and strain related to rural residence did not.

It is clear that there are low-income rural mothers who have chronic health conditions, often a combination of physical and mental health problems, that interfere with their ability to work and to perform everyday tasks. Many rural communities have inadequate health care facilities for both preventive and chronic conditions. Many of these communities have no mental health care providers or none that will take Medicaid patients. When health conditions interfere with one’s ability to hold a job, employer-provided health insurance is not available. Furthermore, many families do not qualify for Medicaid, even if providers will accept that form of payment. Health problems can not only interfere with a mother’s ability to work and manage the family’s resources, they can also reduce her ability to care for herself and her family. For those mothers who have little or no family support, the situation can become worse over time.

Ahluwalia, I. B., Dodds, J. M., and Baligh, M. (1998). Social support and coping behaviors of low-income families experiencing food insufficiency in North Carolina. Health Education and Behavior, 25 (5), 599-612.

Beggs, J. J., Haines, V. A. & Hulbert, J. S. (1996a). Revisiting the rural-urban contrast: Personal networks in nonmetropolitan and metropolitan settings. Rural Sociology, 61(2), 306-325.

Beggs, J. J., Haines, V. A. & Hurlbert, J. S. (1996b). Situational contingencies surrounding the receipt of informal support. Social Forces, 75 (1), 201-222.

Briggs, X. (1998). Brown kids in white suburbs: Housing mobility and the many faces of social capital. Housing Policy Debate, 9 (1), 177-221.

Cohen, P.N. (2002). Extended households at work: living arrangements and inequality in single mothers’ employment. Sociological Forum, 17 (3), 445-63.

Danzinger, S., Kalil, A. & Anderson, N. (2000). Human capital, health and mental health of welfare recipients: co-occurrence and correlates. Journal of Social Issues, 21, 671-92.

Duncan, C. M. (1999). Worlds Apart: Why Poverty Persists in Rural America. New Haven, CT: Yale University Press.

Edin, K. & Lein, L. 1997. Work, welfare and single mothers’ economic survival strategies. American Sociological Review, 62(2), 253-66.

Haines, V. A., Hurlbert, J. S. &Beggs, J. J. (1996). Exploring the determinants of support provision: Provider characteristics, personal networks, community contexts and support following life events. Journal of Health and Social Behavior, 37(3), 252-264.

Hamilton, W. L., Cook, J. T., Thompson, W.W., Buron, L. F., Frongillo, E. A. Jr., Olson, C. M., (1997). Household Food Security in the United States in 1995. Summary Report of the Food Security Measurement Project. Washington, DC: Food and Consumer Service, U.S. Department of Agriculture.

Hao, L. (1994). Kin Support, Welfare, and Out-of-Wedlock Mothers. New York: Garland Publishing, Inc.

Hao, L. (1996). Family structure, private transfers, and the economic well-being of families with children. Social Forces, 75(1), 269-292.

Hogan, D., Eggebeen, D, & Clogg, C. (1993). The structure of intergenerational exchanges in American families. American Journal of Sociology, 98 (6), 1428-58.

Israel, B.S., Farquhar, S.A., Scholz, A.J., James, S.A., & Parker, E.A. (2002). The relationship between social support, stress, and health among women on Detroit’s east side. Health, Education, and Behavior, 29 (3), 342-360.

Kahn, R. L. & Juster, T. (2002). Well-being: Concepts and measures. Journal of Social Issues, 58 (4), 627-644.

Kahneman, D., Diener, E. & Schwarz, N. (Eds.) Well-Being: The Foundations Of Hedonic Psychology (pp. 61-84). New York: Russell Sage Foundation.

Lee, G. R., Netzer, J. K., & Coward, R. T. (1994). Filial responsibility expectations and patterns of intergenerational assistance. Journal of Marriage and the Family, 56, (3), 559-565.

Levitan, L. & Feldman, S. (1991) For love or money: Nonmonetary economic arrangements among rural households in central New York. Research in Rural Sociology and Development, 5, 149-172.

Lichter, D. T., & Jensen, L. (2000). Rural America in transition: Poverty and welfare at the turn of the welfare at the turn of the 21^st century. Paper presented at the meeting of a Research Conference on Poverty, Welfare and Food Assistance.

Lin, N. & Dumin, M. 1986. Access to occupations through social ties. Social Networks 8, 365-385.

Lin, N. & Ensel, W. (1989). Life stress and health: stressors and resources. American Sociological Review, 54(3), 382-99.

Miller, R. J. & Darlington, Y. (2002). Who Supports? The providers of social support to dual-parent families caring for young children. Journal of Community Psychology, 30(5), 461-473.

Nelson, M. K. & Smith, J. (1999). Working Hard and Making Do: Surviving in Small Town America. Berkeley, CA: University of California Press.

Olson, C. M., Anderson, K., Kiss, E., Lawrence, F. C., Seiling, S. B. (2004). Factors protecting against and contributing to food insecurity among rural families: a mixed-methods analysis. Family Economics and Nutrition Review, 16(1), 12-20.

Parks-Yancy, R., DiTomaso, N. & Post, C. (2007) The mitigating effects of social and financial capital resources on hardships. Journal of Family and Economic Issues, 28, 429-448.

Pearlin, L. I. (1985). The sociological study of stress. Journal of Health and Social Behavior, 30, 241-256.

Pearlin, L. I., Schieman, S., Fazio, E. M. & Meersman, S. C. (2005). Stress, health and the life course: Some conceptual perspectives. Journal of Health and Social Behavior, 46, 205-219.

Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1(3), 385-301.

Rogers, C. C. & Dagata, E. (2000) Child poverty in nonmetro areas in the 1990's. Rural America, 15, 28-36.

Robertson, E. B., Elder Jr, G. H., Skinner, M., & Conger, R. D. (1991). The costs and benefits of social support in families. Journal of Marriage and the Family, 53, 403-416.

Schwarz, N. & Strack, F. (1999). Reports of subjective well-being: Judgmental processes and their methodological implications. In D. Kahneman, E. Diener, & N. Schwarz (Eds.) Well-Being: The Foundations Of Hedonic Psychology (pp. 61-84). New York: Russell Sage Foundation.

Seccombe, K. (1999). So you think I drive a Cadillac. Nedham Heights, MA: Allyn & Bacon.

Smilkstein, G. (1978) The Family APGAR: A proposal for family function test and its use by physicians. Journal of Family Practice, 6(6), 1231-1239.

Stack, C. (1974). All Our Kin: Strategies For Survival In The Black Community. New York: Harper and Row.

Stoloff, J.A., Glanville, J. L. & Bienenstock, E. J. (1999). Women’s participation in the labor force: the role of social networks. Social Networks, 21(1), 91-108.

Sturm, R. & Wells, K.B. 2001. Does obesity contribute as much to morbidity as poverty or smoking? Public Health, 115, 229-235.

Teachman, J. D., Paasch, K. Carver, K. (1996). Social capital and dropping out of school early. Journal of Marriage and the Family, 58, 773-783.

Tigges, L.M., Browne, I. & Green, G. P. (1998). Social isolation of the urban poor: Race, class and neighborhood effects on social resources. Sociological Quarterly, 39(1), 53-77.

Turner, H. A. (2006) Stress, social resources, and depression among never-married and divorced rural women. Rural Sociology, 71(3), 479-504.

Wan, C. K., Jaccard, J., & Ramey, S. L. (1996). The relationship between social support and life satisfaction as a function of family structure. Journal of Marriage and the Family, 58(2), 502-513.

Ware, J.E., Jr. (2008) SF-36 health survey update. Downloaded on 4-10-2008 from http://www.sf-36.org/tools.SF36.shtml

Weber, B. A., Duncan, G. J., & Whitener, L. A. (Eds.) (2002). Rural dimensions of welfare reform. Kalamazoo, MI: W. E. Upjohn Institute for Employment Research.

Wellman, B. (1982). Studying personal communities (pp. 61-80). Beverly Hills, CA: Sage Publications.

Wellman, B. & Wortley, S. (1990). Different strokes from different folks: Community ties and social support. American Journal of Sociology, 96(3), 558-588.