Family structure effects on parenting stress and practices in the African American family.

The predominant approach to African-American parenting research focuses on disadvantages associated with single parenthood to the exclusion of other issues. The current research suggests that this does not represent the diversity in family structure configurations among African-American families, nor does it give voice to the parenting resilience of single mothers. We argue that rather than marital status or family configuration, more attention needs to be given to the inadequacy of resources for this population.

In the current study, we examined the parenting of infants by African-American mothers and found that mothers' marital status and family configuration did not affect parenting stress or practices. This suggests, then, that single mothers parent as well as their married, partnered, and multigenerational counterparts. It seems that the economic status and parenting perceptions of mothers contributed more to parenting stress than did marital status or family structure. Our study, then, challenges the accepted wisdom in our political and popular culture that has insisted upon the centrality of the nuclear family to all aspects of familial and even national health. Instead, we have shown that a true commitment to strong families and healthy children begins with a focus on the debilitating effects of poverty in the African-American community.

Key words: African-American, infant parenting, parenting stress

**********

Parenting an infant is a stressful occupation, even under ideal circumstances. Infants require constant attention and, particularly in the first few months of life, must have their every need met constantly (see Combs-Orme, Wilson, Cain, Page & Kirby, 2003, for a discussion of the essentials of parenting specific to infants). Thus it is clear that the full-time job of parenting an infant ideally involves more than one parent. Yet the reality is that many children grow up in homes that do not have two parents present. This is especially true in the African-American community, in which nearly 50% of children are born outside of marriage (or live at some time in a home without one parent or the other) (Andersen, 2000). This study seeks to identify the family structure circumstances that make parenting more or less stressful and successful, and thus seeks to contribute to knowledge that might support parents and families in their efforts to provide the best care possible for every child.

Current research on parenting practices within the African-American community is limited due to an absence of longitudinal research; a severe lack of attention to intragroup variability; a disregard for the inherent diversity in the African-American community; and a minimization of the staggering effects of economic deprivation, racism and social stratification on processes and functioning in the African-American home (Garcia-Coll, Lamberty, Jenkins, McAdoo, Crnic, Wasik & Garcia, 1996). When race is the focal point of research on parenting, between-group differences are paramount in analysis (i.e. European-American vs. African-American), and researchers typically employ a cultural-equivalent theoretical framework (employing theoretical shifts from Allen, 1978). This framework touts European-American, middle-class values and practices as the ideal or norm, and compares African-American values and practices to that norm. African-Americans that are more acculturated and exhibit the values and behaviors of the normed group are depicted as legitimate and are highlighted. However, much of the "race-comparative" research encourages the documentation of unfavorable outcomes of African-American children and families (McLoyd, 1990), and conclusions often concentrate on how African-American children are abnormal, deficient, or incompetent when compared to the middle-class European-American mainstream (Barbarin, 1993; Garcia-Coll, et al., 1996; McLoyd, 1990; Myers, Rana & Harris 1979; Washington & McLoyd, 1982). Thus, the current state of literature on African-American children and their families has as its core the explanation of developmental deviations in comparison to European-American, middle-class norms. Critics of this approach assert that race-comparative research often blames African-American parents for not transmitting the "right" educational, moral, and ethical values to their children, while ignoring situational, contextual, and systemic factors (McLoyd & Randolph, 1984; Spencer, 1990).

Franklin and Boyd-Franklin (1985) contend that traditional African values and beliefs have been transmitted from generation to generation and continue to influence African-American parenting. One of the most important of these traditions is that parents traditionally view childrearing as a communal task to be shared by all members of the community (Franklin & Boyd-Franklin, 1985; Garcia-Coll, Meyer & Brillon, 1995; McAdoo, 1978). Thus, historically African-American parents have relied on extended family networks (i.e. relatives, neighbors, fictive kin, and church members) to share in child care (Wilson, 1989). Research that examines communal childrearing practices in the African-American home asserts that this tradition may mitigate some of the negative effects of single parenthood (Jackson, 1993; Young, 1970).

The traditional Western model of the family includes two married parents and their children (the model nuclear family). However, not all families can or want to model themselves after this norm. Baca Zinn and Eitzen (1999) report that only 10% of U.S. families fit the ideal of a two-parent family in which the male works and the female stays home to care for the children. Thus, in order to properly study parenting, we must ask who exactly provides love and support to children. And, for African-American families, marital status and extended kin networks are significant variables in the study of parenting.

Encouraged by a review of studies on African-American mothers (Murray, Bynum, Brody, Willert & Stephens, 2001) the current study employs a risk and resilience approach to the study of African-American parenting practices from a cultural-variant theoretical perspective (Allen, 1978). The cultural-variant perspective views African-American families as different, yet functional in their own right. The perspective is encapsulated by bell hooks' infamous charge to bring marginalized groups, "from margin to center" (hooks, 1984), and this is the basis of an Afrocentric and risk and resilience approach. To place African-Americans in the center of their reality, we will offer a critical examination of intragroup variability in infant parenting practices.

The research on co-caregiving influences on parenting practices in the African-American home is sparse. In particular, there is very little research on parenting practices in intact, two-generation African-American families. Much of the research focuses on single African-American mothers and young African-American mothers parenting with the assistance of their mothers. Moreover, more diverse family structures (i.e. co-residence with unmarried partners) are often not examined due to the use of marital status as a proxy for family structure (Murray et al, 2001).

The literature reveals that race-homogeneous studies find that maternal grandmothers are typically the co-caregivers with single, African-American mothers (Brody, Flor & Neubaum, 1998; Hunter, 1997; Pearson, Hunter, Ensminger & Kellam, 1990). However, Hunter, Pearson, Ialongo, and Kellam (1998) found that biological fathers are more frequently nominated for co-caregiving in a race-comparative sample. Hunter, et al. (1998) suggests that family proximity and residence in the rural South are primary considerations for the election of maternal grandmothers as the co-caregivers in single-mother, African-American families. Maternal age is also associated with co-resident co-caregiving with grandmothers. The youngest mothers are more inclined to parent with the assistance of their mothers due in part to reduced options. However, Chase-Lansdale, Books-Gunn and Zamsky (1994) found that the age at first birth was not a significant predictor of maternal parenting quality, nor were grandmothers more effective parents than co-residing adolescent mothers.

The relationship between the co-caregiver and the African-American mother appears to have an effect on the quality of parenting provided by the mother. Barbarin and Soler (1993) found that, with the exception of mother-grandmother combinations, the beneficial effect of living in a two-adult household holds true only for biological parents. Tolson and Wilson (1990) found that two-generation African-American families tend to be more organized than multigenerational families. Several studies (Chase-Lansdale et al., 1994; Jackson, Gyamfi, Brooks-Gunn & Blake, 1998; Wakschlag, Chase-Lansdale & Brooks-Gunn, 1996) have established that maternal co-residence with grandmothers was negatively associated with parenting, whereas maternal maturity and autonomy was associated with competent parenting. Jackson (1998) and Jackson and her colleagues (1998) speculate that the instrumental support provided by a co-residing grandmother can be a source of distress for the mother because of differing views on parenting practices and boundary issues. And, Brody, Flor and Neubaum (1998) found that co-caregiver conflict was associated with less mother-child involvement.

In related research, mothers were significantly more likely to parent appropriately when the mothers and grandmothers did not live together (Wakschlag et al., 1996). This may be due to modeling effects. That is, modeling appropriate parenting practices "from a distance" is thought to be more effective because multigenerational relationships are more individuated when mothers and grandmothers live separately (Wakschlag et al., 1996, 2141). Modeling is thought to be less effective in co-residing conditions because grandmothers' participation in the parenting role may contribute to tensions between mothers and grandmothers (due to differing views on parenting practices) and mothers and children (due to blurred boundaries between the parental sphere and the child sphere). In addition, there may be a "selection effect" (Wakschlag et al., 1996, p. 2141) in that more mature (i.e. older, more autonomous) mothers may live separately from their mothers, be more open to learning from their mothers, and practice more appropriate parenting. Clearly, the age of the mother must be taken into account in examining the effects of co-residence on parenting. Inversely, positive co-caregiving behaviors practiced by mothers and grandmothers in the African-American home, such as the provision of instrumental support and practical assistance, have been found to be positively associated with high levels of parental control, the use of physical punishment, and affectionate behaviors toward children (Brody, Flor & Neubaum, 1998). Wilson, Kohn, Curry-El and Hinton (1995) also found that having more adults in the home significantly influenced maternal perceptions of punishment behavior. That is, the more adults in the home, the more favorable maternal perceptions about punishment (spanking and yelling).

Intragroup studies on maternal co-caregiving in the African-American family suggest that maternal grandmothers are most often selected as co-caregiving partners. We investigated co-caregiving partners in the present study and extended the scope of co-caregivng to include unmarried partners, married partners, and maternal grandmothers. We also compared parenting with a co-caregiver to parenting alone.

Contrary to popular perceptions about the positive effects of multigenerational households on young mothers' parenting quality, maternal co-residence with grandmothers appears to be negatively associated with positive parenting. However, this relationship may be mediated by the quality of the co-caregiving relationship and by maternal maturity. Inversely, maternal autonomy and psychological maturity appear to be associated with competent parenting. We examined the quality of mother--grandmother relationships, maternal age, and co-residence on parenting stress and practices in the current study. Moreover, previous research has found that instrumental support and practical assistance by grandmothers is associated with high levels of parental control and the use of physical punishment, and that the number of adults in a home influence perceptions of punishment. Thus, we examined maternal attitudes toward the use of corporal punishment in relation to family structure in the present study.

Methods

This study was part of a larger longitudinal study of maternal parenting practices with newborns (The Volunteer Infant Parent Study--VIPS) (Combs-Orme, Cain, & Wilson, 2004; Combs-Orme et al., 2003). A total of 246 European-American and African-American mothers were recruited from the population of delivering mothers at a University-affiliated, publicly-funded hospital in a mid-size southeastern city between February and November, 1999. Approximately 3200 babies are delivered annually at this urban hospital surrounded by suburban areas and remote mountainous counties without delivery facilities. Interviews were conducted in mothers' hospital rooms within 36 hours after birth. The sub-sample of predominantly urban, African-American mothers is used in this study (n = 103, 42%).

We provided mothers with $10 gift certificates for a large discount department store chain for participation. Due to the sensitive nature of some of the questions in the recruitment interview, mothers were only interviewed in private. The recruitment interview took approximately 30 minutes to complete.

An opportunity sample was used because circumstances did not permit probability sampling, as the hospital was not willing to provide an enumeration of its delivering patients, and resources would not permit an interviewer to be on the Unit at all times.

The recruitment interview included an Informed Consent Form, the Adult-Adolescent Parenting Inventory-2 (AAPI-2) (Bavolek, 1984), the Parental Bonding Inventory (PBI) (Parker, Tupling & Brown, 1979), and extensive tracking data to facilitate follow-up.

We maintained early and continuous tracking of respondents between recruitment and follow-up using personal letters and with the help of collaterals and resources such as internet websites, local utility companies and public housing boards. In total, 93% (n = 96) of the African-American mothers were interviewed at follow-up between August, 1999, and July, 2000.

Mothers were given gift certificates valued at between $10 and $60 for a large discount department store chain to participate in the second interview, with mothers who were hard to schedule being encouraged to participate with increased incentives. Of the follow-up interviews, 97% (n = 93) were conducted in mothers' homes, the homes of family or friends, or at various community locations including coffee shops, offices or jail, and 3% (n = 3) were conducted over the telephone when this was the only way mothers would participate.

There were no differences in maternal age, education, marital status, employment, total family income or previous parenting between African-American mothers who stayed in the sample for follow-up (n = 96) and those who did not participate at follow-up (n = 7) ([X.sup.2](6) = 7.59, p = .270).

The follow-up interview included measures of: parenting stress, behavioral and emotional adjustment, parenting practices and the home environment.

Extensive demographic data were gathered at recruitment and follow-up, including household composition or family structure. In eight cases mothers reported living with great-grandmothers, foster-grandmothers, or aunts and were coded as living with "grandmothers." Five mothers reported living with grandmothers and unmarried partners and were coded as living with "grandmothers." We also recorded mothers' date of birth, previous parenting experience (binary), and maternal education. Data were collected at follow-up about changes since delivery in household composition or family structure (see above for coding), maternal employment (binary), and family income.

Family structure, as described above, is used as the main independent variable in the study, with possible confounders of the relationship between co-caregiving and parenting entered as control variables. Control variables include: maternal age (due to the presumed influence of maturity on parenting practices); maternal employment (due to the possible influence of work- and time management-related stress on parenting); prior parenting (due to the presumed learning curve associated with parenting); and total family income (due to the relationship between adequacy of resources and parenting). It should be noted that infant co-caregiving is inferred based on co-residence.

Research has indicated correlations among the nurturing individuals receive in childhood, relationship quality, and subsequent parenting (Brody, Flor & Neubaum, 1998; George & Solomon, 1999; Solomon & George, 1996). The quality of the co-caregiving relationship was estimated using two measures of mother-grandmother relationship. The Parental Bonding Instrument (PBI) (Parker et al., 1979), administered at delivery, is a 25-item instrument that measures an adult's perceptions of his/her primary caregivers on dimensions of caring and protection. Of the 83 mothers in the current study who reported the identities of their primary caregivers, 95% reported maternal caregivers.

The care subscale of the PBI has 12 items allowing for a maximum score of 36 (higher scores indicate greater care), and the protection subscale has 13 items allowing for a maximum score of 39 (higher scores indicate greater overprotection and control). The overall measure and the subscales (care and protection) show acceptable published test-retest, split-half, and inter-rater reliability and concurrent validity (Parker et al., 1979).

At follow-up, mothers also rated their relationships with their own mothers using a three-point scale (worse, average or better when compared to others) on the Young Adult Self Report (YASR) (Achenbach, 1997), which measures adaptive functioning in social relationships. Achenbach (1997) reports acceptable one-week test-retest reliability.

According to Abidin (1995), parenting stress is the tension parents feel in fulfilling their parenting functions. This tension or stress may be associated with mothers' self-perceived competence in the parenting role, feelings of social isolation and emotional closeness to their infants, physical health, feelings of restriction within the parenting roles, and depression. Also associated with parenting stress are infant qualities that make it difficult for parents to fulfill their parenting roles, such as hyperactivity or demandingness. Parenting stress is important in the study of parenting because stress can lead to child abuse.

The Parenting Stress Index-Short Form (PSI-SF) (Abidin, 1995), a 36-item, standardized instrument used to measure stress related to parenting and parent-child interactions, is widely used in investigations of parenting stress and intervention research. The measure was standardized for use with parents of children from I month to 12 years of age.

The PSI-SF has three subscales (parental distress, difficult child, and parent-child dysfunctional interaction), each with a range from 12 to 60, with higher scores indicating greater parenting stress. For comparison, raw scores are converted to percentile scores, and scores above the 90th percentile represent clinical levels of stress that should be referred for professional assistance.

Published coefficient alphas for the PSI-SF subscales vary from .70 to .84. Construct validity is supported by theoretically meaningful correlations between the PSI-SF scores and constructs such as child adjustment. In addition, studies show higher (more stressed) PSI-SF scores among neglectful, drug-addicted, maladjusted, and abusive parents (Abidin, 1995). For low-income African-American mothers, Hutcheson and Black (1996) found the PSI-SF to to have acceptable levels of internal consistency and stability over six months and high concurrent validity.

The Home Observation for Measurement of the Environment (HOME) Inventory for Families of Infants and Toddlers (Caldwell & Bradley, 1984), the most widely used measure for observational data on the quality of the home environment, is a naturalistic observational technique used to measure parenting practices. Scoring for the HOME is based on a minimum of an hour of observation, along with information gleaned during the interview.

Four subscales in the HOME that measure various aspects of parenting behaviors were used as dependent variables: emotional and verbal responsivity, acceptance, provision of appropriate play materials, and parental involvement. Scores are based on dicotomous responses (present or not present) for each question, and summated scores are compared to norms, with respondents in the lowest quartile considered "at risk" for poor child development related to poor parenting.

Published internal consistency estimates summarized by Bradley (1994) are consistently over .80 for total scores, with subscale coefficients from .30 to .80. Inter-rater reliability has been consistently reported to be .80 or greater. A review of the concurrent and predictive validity of the HOME showed significant relationships to children's intellectual level and cognitive development (Benasich & Brooks-Gunn, 1996). Although Berlin, Brooks-Gunn, Spiker, and Zaslow (1995) contend that the HOME Learning Materials Subscale may overlook some ways in which poor mothers provide general learning experiences without economic resources, Bradley, Caldwell, Rock, Barnard, Gray, Hammond, Mitchell, Siegel, Ramey, Gottfried, and Johnson (1989) and Bradley, Mundfrom, Whiteside, Casey, and Barrett (1994) assert that the HOME is valid for use with economically disadvantaged and African-American families.

Maternal attitudes toward the use of corporal punishment are measured using the Strong Belief in the Use and Value of Corporal Punishment Subscale of the Adult-Adolescent Parenting Inventory (AAPI-2) (note that maternal attitudes toward the use of corporal punishment are measured, not the actual use of corporal punishment). The AAPI-2 is a validated and reliable 40-item inventory designed to measure parenting attitudes (Bavolek & Keene, 1999). Bavolek (1984) reports that sampling considerations in the establishment of the AAPI norms included "geographic region, urban and rural settings, ethnic group, sex, socioeconomic status and age" (p. 45). The Corporal Punishment subscale has eleven items. Age-specific (adolescent and adult) scores are compared to norms (sten scores). Low sten scores (1 to 4) indicate a risk for practicing known abusive parenting practices (i.e. hitting, intimidation, pain and belittlement); high sten scores (7 to 10) indicate parenting attitudes that reflect a nurturing, non-abusive parenting philosophy (i.e. the use of alternative strategies to corporal punishment); and mid-range sten scores (4 to 7) represent the parenting attitudes of the general population.

Results

The objective of these analyses was to identify how family structure/co-caregiving affects maternal stress, parenting, and attitudes toward the use of corporal punishment. In order to investigate these research questions, multiple linear regression was used due to the continuous nature of the dependent variables. To test the effect of co-caregiving, the control variables were entered into the regression equation first; family structure was entered into the regression equation after entry of the control variables. Two-tailed tests (Alpha [less than or equal to] .05) were used because results in either direction were of importance and there is inadequate prior research for definitive predictions about the directions of the relationships.

Few scale items had missing data; items with missing data had only a small percentage missing data; and few respondents had any missing data. Missing item values were imputed with expectation maximization using non-missing values for the other items in the particular scale (Acock, 1997) (SPSS Version 10).

Coefficient alpha was computed for each measure. All measures used in the study had good to excellent internal reliability, ranging from .64 to .89. Statistical power is the probability that the null hypothesis will be correctly rejected, and it is the complement of a Type II error. The sample sizes used in the regression analyses reported here are adequate to detect medium to large effect sizes, but not small effect sizes (Cohen, 1988).

Table 1 shows that of the 96 mothers and infants in the study, 36% were living alone, 23.9% were living with grandmothers, 25% with unmarried partners, and 14.5% with married partners. Family structure was significantly related at the .000 level to both age of mother and total family income. The youngest mothers were living with grandmothers and the oldest with married partners. Those living alone were the poorest: nearly three-quarters had incomes under $5,000. Next disadvantaged were those living with grandmothers, followed by those with unmarried partners. Married mothers had the greatest economic resources.

Table 1 also reveals family structure was significantly related at the .005 level to both previous parenting and maternal education, but not related to maternal employment. Over three-quarters of married mothers reported prior parenting experience (85.7%), whereas only about one-quarter of mothers living with grandmothers did. Mothers living alone, and mothers living with unmarried partners, were about as likely to have parented prior to the index children as not.

Only mothers living with unmarried partners were more likely to be employed (54.2%). Between 57.1% and 69.6% of the other mothers were unemployed. Of the mothers who were working (42%), 30% were doing so full-time.

The majority of all mothers, regardless of family structure, had at least a high school education (range from 87% of mothers living with their mothers, to 100% of mothers living with unmarried partners and married partners). Over half of mothers living with married partners had some college (57.1%).

Tables 2 and 3 show the results for the multiple regression analyses. Co-caregiving did not affect maternal stress after controlling for maternal age, education, employment, total family income, previous parenting, maternal care, maternal overprotection, and maternal relationship quality with her caregiver. However, as high as 16% of our sample reported parental distress levels in the clinical range, and 17% of the variance in parental distress could be explained using maternal demographic variables. In particular, bivariate correlations reveal that parental distress significantly increased as total family income decreased. Distress also significantly increased for mothers who reported that as children they received less optimal care from their primary caregivers.

Co-caregiving also was not related to maternal parenting after controlling for the possible cofounders. However, 16% of the mothers exhibited "at risk" levels of maternal acceptance and involvement, potentially placing their infants at risk developmentally.

In addition, co-caregiving did not affect maternal attitudes toward the use of corporal punishment after taking account of the control variables. However, the mean score for the corporal punishment subscale was 34.29 (SD = 5.46), indicating that these mothers (67%) may be at risk for abuse (i.e. hitting, intimidation, pain and belittlement). Maternal demographic variables did not affect maternal attitudes toward the use of corporal punishment.

Discussion

The current study reveals that with respect to parenting stress and practices, the multigenerational and two-parent family is not necessarily an improvement over single motherhood. Poverty and the quality of the parenting experienced by mothers influenced parenting more than marital status or family structure. Moreover, the relative lack of severe parenting dysfunction among the poor and single-parent mothers in our study suggests great hidden assets, resilience, and strengths in the African-American family. Our study, then, has important implications for the way we think about single motherhood in general and African-American parenting in particular on levels ranging from the clinical to the political. In other words, our study challenges the accepted wisdom in our political and popular culture that has insisted upon the centrality of the nuclear family to all aspects of familial and even national health. Instead, we have shown that a true commitment to strong families and healthy children begins with a focus on the debilitating effects of poverty in the African-American community.

A brief review of our sample reveals that the majority of African-American mothers were parenting their infants with the help of co-caregivers, however, the largest single proportion of mothers were parenting their infants alone. Mothers parenting alone were the poorest; however, they experienced comparable levels of parenting stress, and were as responsive, accepting, and involved with their infants as mothers parenting with the help of married partners, unmarried partners and their infants' grandmothers. In addition, single African-American mothers had comparable learning materials in their homes as mothers parenting with the help of co-caregivers, and promoted the use of corporal punishment in a comparable manner. These findings suggest overarching parenting resilience among poor and single mothers.

While marital status and family structure did not affect parenting stress, 16% of our sample reported parental distress levels that warranted referrals for professional assistance. Lower incomes and mothers' reports that they themselves were not nurtured adequately increased parenting role stress. Not surprisingly, the financial burdens of parenthood contributed to parenting role stress regardless of family structure. Of interest is the finding that mothers' perceptions of the parenting they received significantly affected their distress, indicating a potential pathway for the transmission of parenting practices across generations.

Moreover, while our study found that marital status and family structure did not affect maternal parenting practices, 16% of our mothers exhibited "at risk" levels of maternal acceptance and involvement, potentially placing their infants at risk developmentally. Again, the vast majority of mothers (upwards of 84%) scored in the normative range for acceptance and involvement, indicating that the majority of mothers are faring well in their parenting capacities.

The mothers who appear to be significantly less accepting and involved with their infants may be exhibiting culturally acceptable and traditional parenting practices. Maternal acceptance appears to be closely related to discipline style. Bradley (1998a,b,c) contends that firm, hands-on discipline in the African-American culture is traditional, functional, and appropriate. However, this sentiment is widely contested. Maternal involvement, measures how consistently the mother talks to the infant, provides toys and structural play for the infant, and keeps the infant in visual range. Young (1970) found that African-American children are not commonly encouraged to converse with adults but are encouraged to listen and follow orders and interact with other children. The adaptive strengths suggested in African-American families (i.e. extended family support and shared family roles) (Franklin & Boyd-Franklin, 1985; Garcia-Coll et al., 1995; McAdoo, 1978; Young 1970) would also suggest that African-American mothers' involvement with children is not as developmentally salient as it would be under European standards due to additional caregivers' involvement. These culturally acceptable practices could conceivably decrease the level of involvement between the infants and mothers, but whether they are detrimental to the infant developmentally should be determined empirically. Even if low maternal involvement is detrimental, involvement by other family members may act to protect African-American infants from negative effects.

Similarly, while marital status and family structure did not affect maternal attitudes toward the use of corporal punishment, 67% of mothers reported agreement with very strict, rigid, and authoritarian discipline practices (i.e. hitting, intimidation, pain and belittlement). This finding urges a review of a point of contention in the African-American parenting literature regarding the use of harsh, physical discipline. Bradley (1998a,b,c) takes one side of the debate. She summarizes "seminal studies" on African-American parenting to conclude that African-American parents embrace a firm, hands-on approach to child discipline that is functional, appropriate and administered by caring supportive parents. She discusses this firm discipline approach as an unconscious influence transmitted from slavery where African-American parents maintained harsh controls over children to protect them from suffering and death inflicted by the slave owners. And, she extends this model to a contemporary context, arguing that firm discipline is needed in today's society to prepare African-American children to live, work and function in a racist society. Raymond, Jones and Cooke (1998) take the opposing position and find Bradley's sentiments regarding the functional and appropriate use of corporal punishment unfounded. They state that although "slave-parenting" techniques may have been transmitted to this generation, they are no longer functional or adaptive and may be transmitting underachievement and a violent propensity.

The problem with this debate is that African-American parenting practices and child outcome studies have not been explored sufficiently to make any conclusions. Our study suggests that the majority of African-American mothers embrace very strict, rigid, and authoritarian attitudes toward discipline. However, actual practices and outcomes for children were not measured. This is an important area that requires further exploration.

The aforementioned results should be understood in light of study limitations. Probability sampling was not achieved, possibly threatening external validity. Additionally, the sample sizes used to test co-caregiving and parenting stress and practices were sufficient only to detect large effect sizes. Thus it is conceivable that co-caregiving does affect parenting stress and practices in practically significant ways that could not be detected in the current study.

Table 1
Demographic Characteristics

                                               Percent   Percent
                                    Percent    Mother,   Mother,
                         Percent    Mother,    Baby &    Baby &
                         Mother     Baby &    Unmarried  Married
                         & Baby   Grandmother  Partner   Partner
                        (n = 35)   (n = 23)   (n = 24)  (n = 14)

Characteristic
Maternal Age *
14-18 years               17.1       65.2       20.8       7.1
19-23 years               54.3       21.7       37.5      28.6
24-28 years               20.0        8.7       25.0      35.7
29-33 years                2.9        4.3       12.5      21.4
34-37 years                5.7        0.0        4.2       7.1
Income **
<5,000                    71.4       39.1       29.2       7.1
5,000-19,999              28.6       34.8       37.5      50.0
20,000-34,999              0.0       21.7       33.3      14.3
35,000-49,999              0.0        0.0        0.0      14.3
50,000+                    0.0        4.3        0.0      14.3
Previous Parenting ***
Yes                       54.3       26.1       54.2      85.7
No                        45.7       73.9       45.8      14.3
Maternal Employment ****
Not employed              60.0       69.6       45.8      57.1
Part-time                 11.4       13.0       12.5       7.1
Full-time                 28.6       17.4       41.7      35.7
Maternal Education *****
8th grade or less          0.0        8.7        0.0       0.0
GED                       11.4        4.3        0.0       0.0
High School               54.3       69.6       58.3      35.7
Voc. Training             22.9        8.7       25.0       7.1
College                   11.4        8.7       16.7      57.1

Note. The percentage of missing data ranged from 0 to 2%.

* F = 7.8, p = .000

** F = 11.5, p = .000

*** [chi square] = 12.7, p = .005

**** [chi square] = 3.9, p = .69

***** [chi square] = 13.0, p = .005

Table 2
Co-Caregiving Effects on Maternal Stress (N = 92)

                                          Total Stress

    Variables                        B          t        Beta

Age                                 -.65      -1.35      -.25
Education                           1.40        .62       .10
Employment                         -1.67       -.45      -.06
Income                             -1.10       -.52      -.08
Previous Parenting                  1.88       1.12       .17
Care                                -.28      -1.35      -.15
Overprotection                       .06        .29       .03
Relationship Quality               -1.10       -.76      -.08

Step 1                       [R.sup.2] = .13 ([R.sup.2] adj = .04)
                                       F (8,83) = 1.51,
                                         p = .165

Married Partner                    -9.34      -1.70       *
Unmarried Partner                  -5.21      -1.33       *
Grandmother                        -6.35      -1.45       *

Step 2                             [R.sup.2] change = .04,
                                   F change (3,80) = 1.27,
                                        p = .289

Overall Model                [R.sup.2] = .17 ([R.sup.2] adj = .05)
                                       F (3,80) = 1.46,
                                         p = .164

                                       Parental Distress

    Variables                        B          t        Beta

Age                                  .03        .11       .02
Education                          -1.01       -.07      -.11
Employment                          1.28        .54       .07
Income                             -1.72      -1.25      -.19
Previous Parenting                   .21        .20       .03
Care                                -.22      -1.65      -.19
Overprotection                       .14       1.03       .11
Relationship Quality                -.80       -.86      -.09

Step 1                       [R.sup.2] = .17 ([R.sup.2] adj = .09)
                                       F (8,83) = 2.07,
                                          p = .048

Married Partner                    -2.02       -.57       *
Unmarried Partner                  -2.21       -.87       *
Grandmother                        -1.70       -.60       *

Step 2                             [R.sup.2] change = .01,
                                   F change (3,80) = .28,
                                          p = .838

Overall Model                [R.sup.2] = .18 ([R.sup.2] adj = .06)
                                       F (3,80) = 1.54,
                                          p = .133

                                    Dysfuctional Interaction

    Variables                        B          t        Beta

Age                                 -.39      -1.85      -.34
Education                           1.25       1.26       .20
Employment                         -1.29       -.80      -.11
Income                               .27        .29       .05
Previous Parenting                  1.08       1.46       .22
Care                                -.02       -.24      -.03
Overprotection                      -.03       -.29      -.03
Relationship Quality                -.15       -.23      -.03

Step 1                        [R.sup.2] = .07 ([R.sup.2] adj = -.02)
                                        F (8,83) = .81,
                                           p = .599

Married Partner                    -5.85      -2.42       *
Unmarried Partner                  -3.20      -1.85       *
Grandmother                        -4.07      -2.11       *

Step 2                             [R.sup.2] change = .08,
                                   F change (3,80) = 2.59,
                                           p = .058

Overall Model                [R.sup.2] = .15 ([R.sup.2] adj = .04)
                                       F (3,80) = 1.33,
                                         p = .225

                                       Difficult Child

    Variables                        B          t        Beta

Age                                 -.38      -1.63      -.30
Education                            .61        .55       .09
Employment                          -.51       -.28      -.04
Income                              -.64       -.61      -.09
Previous Parenting                  1.01       1.23       .18
Care                                -.19      -1.82      -.21
Overprotection                       .07        .71       .08
Relationship Quality                -.58       -.82      -.09

Step 1                       [R.sup.2] = .14 ([R.sup.2] adj = .06)
                                       F (8,83) = 1.70,
                                          p = .111

Married Partner                    -2.66       -.99       *
Unmarried Partner                   -.48       -.25       *
Grandmother                        -1.25       -.59       *

Step 2                             [R.sup.2] change = .01,
                                   F change (3,80) = .37,
                                            p = .777

Overall Model                [R.sup.2] = .15 ([R.sup.2] adj = .04)
                                       F (3,80) = 1.31,
                                         p = .236

* Standardized Beta Coefficients are not used with dummy coded
variables

Table 3
Co-Caregiving Effects on Maternal Parenting (N= 81)

                                          Responsivity

Variables                            B          t        Beta

Age                                  .01        .12       .03
Education                            .19        .60       .10
Employment                           .11        .22       .03
Income                               .04        .13       .02
Previous Parenting                   .03        .11       .02
Care                                -.01       -.20      -.03
Overprotection                       .00        .14       .02
Relationship Quality                -.24      -1.22      -.15

Step 1                       [R.sup.2] = .05 ([R.sup.2] adj = -.06)
                                       F = (8,72) = .43,
                                           p = .899

Married Partner                      .24        .32       *
Unmarried Partner                    .65       1.19       *
Grandmother                          .65       1.14       *

Step 2                             [R.sup.2]change = .03,
                                   F change (3,69) = .65,
                                           p = .588

Overall Model                [R.sup.2] = .07 ([R.sup.2] adj = -.08)
                                       F = (3,69) = .48,
                                           p = .907

                                           Acceptance

Variables                            B          t        Beta

Age                                 -.05       -.74       .17
Education                            .34       1.13       .19
Employment                           .18        .38       .06
Income                               .13        .48       .08
Previous Parenting                   .24       1.03       .19
Care                                -.03      -1.03      -.14
Overprotection                       .03       1.07       .13
Relationship Quality                -.04       -.21      -.03

Step 1                       [R.sup.2] = .05 [R.sup.2] adj = -.05)
                                       F = (8,72) = .49,
                                           p = .859

Married Partner                     -.48       -.69       *
Unmarried Partner                    .50        .94       *
Grandmother                          .64       1.16       *

Step 2                             [R.sup.2]change = .04,
                                   F change (3,69) = 1.10,
                                            p = .356

Overall Model                  [R.sup.2] = .10 ([R.sup.2] adj = -.05)
                                         F = (3,69) = .66,
                                             p = .773

                                           Involvement

Variables                            B          t        Beta

Age                                 -.01       -.09      -.02
Education                            .18        .61       .10
Employment                          -.28       -.57      -.08
Income                               .62       2.32       .35
Previous Parenting                  -.28      -1.16      -.20
Care                                 .00       -.12      -.02
Overprotection                       .01        .34       .04
Relationship Quality                -.21      -1.09      -.12

Step 1                       [R.sup.2] = .16 ([R.sup.2] adj = -.06)
                                       F = (8,72) = 1.67,
                                           p = .121

Married Partner                     -.38       -.53       *
Unmarried Partner                    .65       1.22       *
Grandmother                         -.48       -.86       *

Step 2                             [R.sup.2]change = .05,
                                   F change (3,69) = 1.45,
                                           p = .235

Overall Model                [R.sup.2] = .21 ([R.sup.2] adj = .08)
                                       F = (3,69) = 1.63,
                                           p = .109

                                        Learning Materials

Variables                            B          t        Beta

Age                                  .01        .19       .04
Education                            .34       1.40       .18
Employment                          -.32       -.61      -.09
Income                               .43       1.51       .24
Previous Parenting                  -.12       -.49      -.09
Care                                 .00        .04       .01
Overprotection                       .02        .63       .08
Relationship Quality                -.01       -.07      -.01

Step 1                       [R.sup.2] = .11 ([R.sup.2] adj = .01)
                                       F = (8,72) = 1.09,
                                            p = .384

Married Partner                      .07        .09       *
Unmarried Partner                    .19        .34       *
Grandmother                         -.01       -.01       *

Step 2                             [R.sup.2]change = .00,
                                   F change (3,69) = .05,
                                           p = .985

Overall Model                [R.sup.2] = .11 ([R.sup.2] adj = -.03)
                                       F = (3,69) = .77,
                                           p = .667

* Standardized Beta Coefficients are not used with dummy coded
variables

References

Abidin, R. R. (1995). Parenting Stress Index: Professional Manual (3ed.). Odessa, FL: Psychological Assessment Resources, Inc.

Achenbach, T. M. (1997). Manual for the Young Adult Self-Report and Young Adult Behavior Checklist. Department of Psychiatry, University of Vermont.

Acock, A. C. (1997). Working with missing values. Family Science Review, 10(1), 76-102.

Allen, W. R. (1978). The search for applicable theories in black family life. Journal of Marriage and the Family, 40, 117-129.

Andersen, L. M. (2000). Thinking About Women: Sociological Perspectives on Sex and Gender (5th ed.). Boston, MA: Allyn and Bacon.

Baca Zinn, M. & Eitzen, D. S. (1999). Diversity in American Families. New York: Addison Wesley Longman.

Barbarin, A. O. (1993). Coping and resilience: Exploring the inner lives of african american children. Journal of Black Psychology, 19(4), 478-492.

Barbarin, A. O. & Soler, E. R. (1993). Behavioral, emotional, and academic adjustment in a national probability sample of african american children: Effects of age, gender, and family structure. Journal of Black Psychology, 19(4), 423-446.

Bavolek, S. J. (1984). Handbook for the AAPI Adult-Adolescent Parenting Inventory. Park City, UT: Family Development Resources, Inc.

Bavolek, S. J. & Keene, R. G. (1999). Adult-Adolescent Parenting Inventory (AAPI-2) Administration and Development Handbook. Asheville, NC: Family Development Resources Inc.

Benasich, A. A. & Brooks-Gunn, J. (1996). Maternal attitudes and knowledge of child-rearing: Associations with family and child outcomes. Child Development, 67(3), 1186-1205.

Berlin, L. J., Brooks-Gunn, J., Spiker, D., Zaslow, M J. (1995). Examining observational measures of emotional support and cognitive stimulation in black and white mothers of preschoolers. Journal of Family Issues, 16(5), 664-686.

Bradley, R. C. (1998a). Child discipline in African American families: Scholars must question the generalizability of poor research [9pp]. Family Journal [Online serial], 6(4). Available: ProQuest Direct [1999, Sept. 18].

Bradley, R. C. (1998b). Child rearing in african american families: A study of the disciplinary practices of african american parents. Journal of Multicultural Counseling and Development, 26, 273-281.

Bradley, R. C. (1998c). Cultural interpretations of child discipline: Voices of African American scholars [46pp]. Family Journal [Online serial], 6(4). Available: ProQuest Direct [1999, Sept. 9].

Bradley, R. H. (1994). The Home Inventory: Review and reflections. Advances in Child Development, 25, 241-288.

Bradley, R. H., Caldwell, B. M., Rock, S. L, Barnard, K, E., Gray, C., Hammond, M. A., Mitchell, S., Siegel, L., Ramey, C. T., Gottfried, A. W., & Johnson, D. L. (1989). Home environment and cognitive development in the first 3 years of life: A collaborative study involving six sites and three ethnic groups in North America. Developmental Psychology, 25(2), 217-235.

Bradley, R. H., Mundfrom, D. J., Whiteside, L., Casey, P. H., Barrett, K. (1994). A factor analytic study of the infant-toddler and early childhood versions of the HOME inventory administered to white, black, and Hispanic American parents of children born preterm. Child Development, 65, 880-888.

Brody, H. G., Flor, L. D. & Neubaum, E. (1998). Coparenting processes and child competence among rural african-american families. In M. Lewis & C. Feiring (Eds.), Families, Risk, and Competence (pp. 227-244). Mahwah, NJ: Lawrence Erlbaum Associates.

Caldwell, B. & Bradley, R. H. (1984). Home Observation for Measurement of the Environment. Little Rock: Center for Research on Teaching and Learning, University of Arkansas.

Chase-Lansdale, L. P., Brooks-Gunn, J. & Zamsky, E. S. (1994). Young african-american multigenerational families in poverty: Quality of mothering and grandmothering. Child Development, 65, 373-393.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.

Cohen, J. & Cohen, P. (1983). Applied Multiple Regression Correlation Analysis for the Behavioral Sciences (2nd ed., pp. 172-176). Hillsdale, NJ: Lawrence Erlbaum Association Publications.

Combs-Orme, T., Cain, D.S., Wilson, E. E. (2004). Do maternal concerns at delivery predict parenting stress during infancy? Child Abuse and Neglect, 28(4), 377-392.

Combs-Orme, T., Wilson, E. E., Cain, D. S., Page, T., & Kirby, L. D. (2003). Context-based parenting in infancy: Background and conceptual issues. Child and Adolescent Social Work Journal, 20(6), 437-472.

Davis, D. W. (1997). The direction of race of interviewer effects among African-Americans: donning the black mask. American Journal of Political Science, 41(1), 309-323.

Franklin, A. J., & Boyd-Franklin, N. (1985). A psychoeducational perspective on black parenting. In H. P. McAdoo & J. L. McAdoo (Eds.), Black Children: Social, Educational & Parental Environments. Beverly Hills, CA: Sage Publications.

Garcia-Coll, C., Lamberty, G., Jenkins, R., McAdoo, P. H., Crnic, K., Wasik, H. B., & Garcia, V. H. (1996). An integrative model for the study of developmental competencies in minority children. Child Development, 67, 1891-1914.

Garcia-Coll, C. T., Meyer, E. C., & Brillon, L. (1995). Ethnic and minority parenting. In M. Bornstein (Ed.), Handbook of Parenting: Biology and Ecology of Parenting (Vol. 2, pp. 189-209). Hillsdale, NJ: Earlbaum.

George, C., & Solomon, J. (1999). Attachment and caregiving: The caregiving behavioral system. In J. Cassidy and P. R. Shaver (Eds.), Handbook of Attachment: Theory, Research, and Clinical Applications (pp. 649-670). New York, NY: The Guilford Press.

hooks, b. (1984). Feminist Theory: From Margin to Center. Boston, MA: South End Press.

Hunter, A. G. (1997). Counting on grandmothers: Black mothers' and fathers' reliance on grandmothers for parenting support. Journal of Family Issues, 18(3), 251-269.

Hunter, A. G., Pearson, J. L., Ialongo, N. S., & Kellam, S. G. (1998). Parenting alone to multiple caregivers: Child care and parenting arrangements in black and white urban families. Family Relations, 47, 343-353.

Hutcheson, J. J. & Black, M. M. (1996). Psychometric properties of the parenting stress index in a sample of low-income African-American mothers of infants and toddlers. Early Education and Development, 7(4), 381-400.

Jackson, A. P. (1998). The role of social support in parenting for low-income, single, black mothers. Social Service Review, 72, 365-379.

Jackson, F. J. (1993). Multiple caregiving among African americans and infant attachment: The need for an emic approach. Human Development, 36, 87-102.

Jackson, A. P., Gyamfi, P., Brooks-Gunn, J., & Blake, M. (1998). Employment status, psychological well-being, social support, and physical discipline practices of single black mothers [41pp]. Journal of Marriage and the Family [Online serial], 60(4). Available: ProQuest Direct [1999, Sept. 18].

McAdoo, H. P. (1978). Factors related to stability in upwardly mobile black families. Journal of Marriage and the Family, 40, 761-776.

McLoyd, C. V. (1990). The impact of economic hardship on black families and children: Psychological distress, parenting, and socioemotional development. Child Development, 61, 311-346.

McLoyd, C. V. & Randolph, M. S. (1984). The conduct and publication of research on afro-american children: A content analysis. Human Development, 27, 6575.

Murray, V. M., Bynum, M. S., Brody, G. H., Willert, A. & Stephens, D. (2001). African American single mothers and children in context: A review of studies on risk and resilience. Clinical Child and Family Psychology Review, 4(2), 133-155.

Myers, H. F., Rana, P. G., & Harris, M. (1979). Black Child Development in America, 1927-1977. Westport, CT: Greenwood.

Nunnally, J. C. & Bernstein, I. H. (1994). Psychometric Theory (3rd ed.). New York: McGraw-Hill.

Parker, G., Tupling, H. & Brown, B. L. (1979). A parental bonding instrument. British Journal of Medical Psychology, 32, 1-10.

Pearson, L. J., Hunter, G. A., Ensminger, E. M., & Kellam, G. S. (1990). Black grandmothers in multigenerational households: Diversity in family structure and parenting involvement in the woodlawn community. Child Development, 61, 434-442.

Randolph, M. S. (1995). African american children in single-mother families, in B. J. Dickerson (Ed.), African American Single Mothers: Understanding Their Lives and Families (pp. 117-145). Thousand Oaks, CA: Sage Publications.

Raymond, J. H., Jones, F. & Cooke, V. (1998). African american scholars and parents cannot blame current harsh physical punishment of black males on slavery: A response to "cultural interpretations of childellipsis" [55pp]. Family Journal [Online serial], 6(4). Available: ProQuest Direct [1999, Sept. 181.

Rhodes, P. J. (1994). Race-of-interviewer effects: a brief comment. Sociology, 28(2), 547-559.

Solomon, J. & George, C. (1996). Defining the caregiving system: Toward a theory of caregiving. Infant Mental Health Journal, 17(3), 183-197.

Spencer, M. B. (1990). Development of minority children: An introduction. Child Development, 61, 267-269.

Tolson, J. F. T., & Wilson, N. M. (1990). The impact of two- and three-generational black family structure on perceived family climate. Child Development, 61, 416-428.

Wakschlag, S. L., Chase-Lansdale, L. P., & Brooks-Gunn, J. (1996). Not just "ghosts in the nursery": Contemporaneous intergenerational relationships and parenting in young african-american families. Child Development, 67, 2131-2147.

Washington, E., & McLoyd, V. C. (1982). The external validity of research involving American minorities. Human Development, 25, 324-339.

Wilson, M. N. (1989). Child development in the context of the black extended family. American Psychologist, 44(2), 380-385.

Wilson, M. N., Kohn, L. P., Curry-El, J. & Hinton, I. D. (1995). The influence of family structure characteristics on the child-rearing behaviors of African-American mothers. Journal of Black Psychology, 21(4), 450-462.

Young, H. V. (1970). Family and childhood in a southern negro community. American Anthropologist, 72, 269-288.

DAPHNE S. CAIN

Louisiana State University

TERRI COMBS-ORME

University of Tennessee