Infant Social Interactions With Multiple Caregivers

Abstract

Most studies of diverse populations of families within the United States have either focused predominantly on ethnicity or socioeconomic status (SES), and those that have examined both ethnicity and SES have noted difficulties in disentangling the effects of SES and ethnicity. In order to achieve a greater understanding of variation in infant experiences with parental and nonparental caregivers in differing socioeconomic and ethnic contexts, 41 infants from African American and 40 infants from European American families of lower and middle SES were observed for 12 hours each in and around their home environments. Ethnic differences were evident in the infants’ overall experiences with caregivers, maternal availability, affection, caregiving, and stimulation by nonnuclear relatives; SES differences were identified for maternal and paternal holding, maternal carrying, and paternal caregiving. When caregiver availability was taken into account, variations in interactional and care experiences were predominantly predicted by ethnicity. These results underscore the need to study both ethnicity and socioeconomic variation rather than either one alone. Furthermore, the caregiving behaviors of African American mothers and fathers may be misrepresented when multiple SES contexts are not considered.

Keywords

family/child rearing infant/child cultural psychology socioeconomic

Ethnicity and socioeconomic status (SES) are frequently viewed as important factors affecting caregiving and other interactive experiences of infants and children (e.g., Bartz & Levine, 1978; Fong et al., 2004; Henly & Lyons, 2000; Hirshberg, Huang, & Fuller, 2005; Hogan, Hao, & Parish, 1990; Huston, Chang, & Gennetian, 2002; Li-Grining & Coley, 2006; Miller-Loncar, Erwin, Landry, Smith, & Swank, 1998). However, studies examining these factors have often overrepresented minority families in lower income categories and European American families in higher income categories, complicating attempts to disentangle the independent and combined effects of ethnicity and income (e.g., Benin & Keith, 1995; Bradley, Corwyn, McAdoo, & Garcia Coll, 2001; Early & Burchinal, 2001; Kontos, Howes, Shinn, & Galinsky, 1997). Emphasizing the need to consider ethnicity and socioeconomic context simultaneously, Kontos et al. (1997) found no statistical interaction between ethnicity and income in their study of children’s experiences in family child care and relative care suggesting that ethnicity acts as a main effect rather than as a moderator of socioeconomic effects. They indicate that children’s family and relative child care experiences may vary according to their socioeconomic status and ethnicity. Likewise, Super and Harkness (1997) argue that families of different cultural and socioeconomic backgrounds may value different interactional styles. Thus, in the current study we examine the social experiences of 3- to 4-month-old infants in African American and European American families in lower and middle socioeconomic contexts and consider the forms and extent to which multiple caregivers (mothers, fathers, siblings, nonnuclear relatives, and nonrelatives) interacted with infants in relation to the SES and ethnicity of the families.

Whether or not they focus on ethnicity and SES, most studies of young infants tend to focus on parents, and especially mothers, rather than on nonparental caregivers (e.g., Ainsworth, Blehar, Waters, & Wall, 1978; Bakermans-Kranenburg, van Ijzendoorn, & Kroonenberg, 2004; Field & Widmeyer, 1981; Fracasso, Lamb, Schölmerich, & Leyendecker, 1997; Gottfried, 1984; Ispa et al., 2004; Leyendecker, Lamb, Schölmerich, & Fricke, 1997; Roopnarine, Fouts, Lamb, & Lewis, 2005). It is widely suggested that African American families rely more heavily on extended kin networks than European Americans, with many scholars considering the availability of extended family support to be a prominent cultural practice among African Americans (Dilworth-Anderson, 1992; Heath, 1983; Lareau, 2003; Stack, 1974; Ward, 1971; Wilson, 1989) even though this proposition has not been documented persuasively in studies controlling for SES (Benin & Keith, 1995; Hofferth, 1984; Hogan et al., 1990; Vandell, McCartney, Owen, Booth, & Clarke-Stewart, 2003). Some researchers have reported that, in nationally representative samples, lower SES families rely more than middle-income families on extended kin for child care regardless of ethnicity (Hofferth, 1995; Lehrer, 1983). Fouts, Roopnarine, and Lamb (2007) found that lower SES African American infants interacted more with extended kin than African American infants in upper SES contexts. Some of the inconsistencies in the findings may be attributable to methodological differences (i.e., observations of families, national surveys, qualitative interviews). In sum, comparative research examining the variety of caregivers available to and involved with young infants in European American and African American families in different SES contexts is rare.

Conceptual Framework and Defining SES and Ethnicity

It has long been recognized that socioeconomic circumstances and ethnicity both influence children’s experiences with caregivers and that SES and ethnicity are intertwined (Low, 2005; Ogbu, 1981; Super & Harkness, 1997). The relative importance of SES and ethnicity effects on children’s experiences and family care patterns has been conceptualized in many different ways (Kuhlthau & Mason, 1996; Lareau, 2003). In the present study, we adopted a cultural-contextual, adaptive-resilient approach in examining families from two ethnic groups (African American and European American) equally sampled across two SES contexts (lower and middle). Like others who adopt the same conceptual approach, we do not assume that families who have particular characteristics are deficient but instead propose that patterns of childcare are adapted to particular cultural and socioeconomic ecologies. Instead of asking whether socioeconomic circumstances or ethnicity have a greater influence on early care patterns and infant experiences, therefore, we consider that both may be influential and that family patterns change over time in response to cultural demands and different socioeconomic circumstances (Low, 2005; Ogbu, 1981; Super & Harkness, 1997).

Because we assume that family patterns adapt to cultural and socioeconomic constraints, we also expect that families from different cultural and socioeconomic backgrounds may value different styles of interaction and care (Super & Harkness, 1997). Accordingly, in the present study, a broad range of infant-caregiver interactions were examined, including vocalizing, stimulating, physical affection, vocal affection, physical soothing, vocal soothing, physical caretaking, holding, and carrying.

A family’s SES is often defined either using measures of income (e.g., low-, moderate-, and upper-income) or a combination of indices, including factors such as income, wealth, parental education levels, status of parental occupation(s), or income-to-needs assessments (Entwisle & Astone, 1994). In the present study, we assessed SES using a combination of indices (income, parent education, and status of parent occupation). Due to the various ways that SES has been defined, in the review that follows, we focus on studies that consider one or more aspects of SES. For example, we assume that studies focused on the impact of economic hardship help us understand the broader significance of SES.

Following Phinney’s (1996) exploration of the concept of ethnicity, we view ethnicity as being made up of one’s culture, ethnic identity, and minority or majority status. We view culture as both beliefs and practices that are shared by a group and are socially transmitted. This is consistent with the cultural-contextual adaptive-resilient approach, which places emphasis on child care practices, beliefs about child care, and social and physical ecologies in defining the child care niche (Super & Harkness, 1997). Researchers have noted that both ethnicity and SES (and more broadly, social class) may relate to childrearing practices that can be considered “cultural practices” (Lareau, 2003; Morelli, Rogoff, & Angelillo, 2003; Rogoff, 2003). Rogoff (2003) defined cultural communities as “groups of people who have some common and continuing organization, values, understanding, history, and practices” (p. 80). Thus, a cultural community may or may not be linked to ethnicity and SES, but there are multiple factors that may define cultural communities. Rogoff (2003) warned that a categorical approach to culture based solely on ethnic identity does not acknowledge the dynamic nature of cultural communities, including variation within communities. We agree that ethnic identity is not synonymous with cultural community. Because we did not define the SES and ethnic groups in this study according to their cultural beliefs and practices, we are not aware of the extent to which the groups in our study may represent cultural communities.

In addition to the cultural component of ethnicity, we also view ethnicity as self-defined in relation to larger societal structures, including majority and minority statuses. In acknowledging the possible impact of minority status on individual identity, Ogbu (1978) discussed how identification as a “voluntary” minority as opposed to a “forced” minority can influence the interpretation of opportunities in the environment. On this point, Ogbu (1981) has warned that comparing majority with minority groups can make minority groups appear deficient. Ogbu (1981) suggested that researchers should study within-group variability before making ethnic group comparisons, and we thus built this study upon previous attempts to assess intra-ethnic variations in childcare and socialization patterns (Fouts et al., 2007; Roopnarine et al., 2005). Informed by prior work, five types of caregivers were chosen for study (mothers, fathers, siblings, nonnuclear relatives, and nonrelatives) in order to avoid implicit assumptions that nuclear family members (parents and siblings) as opposed to extended family members (nonnuclear relatives) or friends (nonrelatives) play the most prominent roles in infant care.

Although we attempt to distinguish between socioeconomic status and ethnicity for analytic purposes, these concepts are certainly interrelated because families and individuals may experience socioeconomic conditions differently depending on their ethnicity, and socioeconomic context may be a part of how families identify with their cultural communities. For example, McLoyd (1990) has explained that it is problematic to generalize the experiences of economic loss among European American families to African American families, because “differences in economic resources, employment opportunities, perceptions of the causes of economic loss, and other factors associated with race may modify responses to economic loss” (McLoyd, 1990, p. 313). Murray, Smith, and Hill (2001) have also noted that ethnicity interacts with economic loss and poverty and that these interactions may hinder opportunities in life, because discrimination and prejudice experienced by people of color perpetuate disparities “in terms of economic power, political influence, civil rights, and access to resources” (p. 911).

Socioeconomic Context and Infant and Child Care

Most studies that address socioeconomic contexts of early care and education have focused on toddlers and older children, ignoring the possibility that socioeconomic differences may be evident during the early infancy period. Studies involving preschoolers have reported that low-income and poor families are more likely to rely on informal care networks (friends, family, and neighbors) rather than on formal child care facilities and that low-income and poor families tend to rely on relatives for child care more often than wealthier families (Hofferth, 1995; Lehrer, 1983). Kontos et al. (1997) have also found that very low-income infants and young children (i.e., between 10 and 65 months) are often cared for by providers who were less sensitive than those caring for moderate-income children (Kontos et al., 1997). Furthermore, regardless of ethnicity, socioeconomic circumstances appear to be related to the level of vocal stimulation that children experience, with young children in lower SES contexts receiving less vocal stimulation than those in higher SES contexts (Gottfried, 1984; Hart & Risley, 1995; Hoff, 2003). Other researchers have not found differences in levels of vocalizations associated with SES, however (Fouts et al., 2007; Leyendecker et al., 1997; Roopnarine et al., 2005). These discrepancies could possibly be due to cultural differences in emphasis on early verbal stimulation. Based on an examination of free-standing conversations between adults and toddlers in three cultural communities (middle-class European American, Efe foragers in the Democratic Republic of Congo, and indigenous Maya in Guatemala), Morelli et al. (2003) have suggested that the prevalence of adult conversations with toddlers among middle-class European Americans may represent “middle-class” cultural practices geared to preparing children for success in school.

Two major limitations of the studies attempting to assess the relationship between socioeconomic circumstances and caregiver-child interactions are the widespread focus on very poor families and inattention to the number of individuals involved in the care and socialization of children in diverse ethnic groups (see McLoyd, 1990). This is especially problematic because variations in socioeconomic circumstances (e.g., economic disadvantage) are likely to prevent parents from providing care of consistently high quality to infants and children and because such parental behavior may adversely affect children’s social and emotional functioning (McLoyd, 1990; McLoyd, Hill, & Dodge, 2005). The restricted focus on mothers rather than all caregivers is certainly not limited to studies of any one ethnic group, but the increased attention to fathers in the last 30 years (Lamb, 2010) has not been matched by a broader focus on all the individuals playing significant roles in children’s social and cognitive development.

Ethnicity and Infant and Child Care

Turning to the role of ethnicity in shaping infant and child experiences with caregivers, Field and Widmeyer (1981) found that interactions involving low-income African American mothers and infants were characterized by less talking, “infantized” behavior (e.g., baby talk), responsivity to infant cues, and game playing than among low-income Cuban, Puerto Rican, or South American immigrant mothers and infants in the United States. These researchers suggested that the infrequent occurrence of these behaviors may have reflected the African American mothers’ concern about not “spoiling” children. Similarly, in a survey of Black, Chicano, and Anglo American parents in a low-income working-class neighborhood, Bartz and Levine (1978) found that Black parents were distinguished by beliefs that children should have more responsibility for their body (e.g., toilet training) and feelings earlier, concerns about children “wasting” time (e.g., unoccupied leisure time), preferences for high control and monitoring of children’s behavior, and declarations that the parents were more emotionally supportive of their children than those in other ethnic groups. Others too have found that African American parents express less warmth or affection while monitoring or controlling their children than European American parents, leading them to suggest that these parenting patterns may have developed in response to hostile or dangerous environments and that the differences in warmth and vocal interactions should be interpreted as adaptive rather than deficient (Berlin, Brooks-Gunn, Spiker, & Zaslow, 1995; Garcia Coll, 1990; Hofferth, 2003; Ispa et al., 2004).

Analyzing data from the National Longitudinal Survey of Youth, Bradley et al. (2001) found that African American infants were less likely to receive physical affection from their mothers than Asian American, European American, and Hispanic American infants. Affection was also predicted by income: More affluent mothers expressed affection more frequently than less affluent mothers in all ethnic groups. African American mothers were also least likely to report “no spanking,” but parents below the poverty line spanked more frequently than parents above the poverty line regardless of ethnicity. African American infants were least likely to spend time with their fathers, but poor African American infants spent much less time with fathers than nonpoor African American infants. European American mothers were most likely to read to their infants, and this tendency was especially pronounced among nonpoor European American mothers. Although Bradley et al. considered poverty status (poor versus nonpoor) in their analyses, they cautioned that ethnicity and SES were confounded.

As noted above, many researchers have suggested that African American families tend to have more access to and rely more on kin support for the direct care of their children than European American families (Benin & Keith, 1995; Early & Burchinal, 2001; Jackson, 1993; Hays & Mindel, 1973; Wilson, 1989). According to Dilworth-Anderson (1992), this reliance originated in a “mutual aid system” that “emerged out of the idea of survival in a hostile and oppressive society where blacks viewed themselves as ‘making it’ only through the concerted efforts of groups of people” (p. 30). In view of the current economic hardships faced by many families in the United States, African American families may continue to access kin networks for childcare support (Dilworth-Anderson, 1992). In a review of research related to social support among urban poor parents, Belle (1983) suggested that different social support networks, including kin networks, buffer the negative consequences of poverty and thus appear adapted to economic hardship regardless of ethnicity.

In analyses of the 1995 National Household Education Survey data, Early and Burchinal (2001) found that ethnicity predicted families’ use of relative care for infants and children because African American families with infants and children were more likely to rely on relative care than were European American or Latino American families. In this large sample (7,553 families with children between the ages of 0 and 5 years), poverty status (“poor” [at or below the poverty line], “near-poor” [between 100% and 200% of the poverty line], and “not-poor” [above 200% of the poverty line]) did not predict reliance on relative care, but unfortunately African American and Latino Americans were overrepresented in the poor category, limiting the ability to reach strong conclusions about ethnicity.

After examining nationally representative data (National Longitudinal Survey of Labor Force Behavior, Youth Survey), Hogan et al. (1990) reported that young African American mothers (age 19 to 26) had better access to kin than young European American mothers and that African American mothers were more likely to rely on kin for child care than were European American mothers. Similarly, in a national survey of African American and European American families with children under the age of 18, Hofferth (1984) showed that African American families were more likely to have extended kin living in their household and that European American families were more likely to receive financial support from their extended kin. It is difficult to identify whether families in either group received more support from kin because support was provided in qualitatively different forms. Although Hofferth (1984) did not consider family income in her analyses, she found that mothers who had more education were more likely to participate in a kinship network than were less educated mothers. Two other large-scale studies also found that women of color (defined as all ethnicities other than non-Hispanic White) were more likely to report relying on grandparent care than non-Hispanic White women (Vandell et al., 2003) and that African American mothers were more likely to receive help from relatives when their young children were ill (Benin & Keith, 1995).

In sum, most researchers have focused on ethnic rather than SES variations in early care and socialization. Researchers who have focused on ethnicity have sometimes failed to distinguish clearly between ethnic minority groups (Vandell et al., 2003), and many have not considered variation in SES as a factor in their analyses (Hofferth, 1984; Hogan et al., 1990; Vandell et al., 2003). Moreover, most of the studies have involved parental surveys, with few researchers observing infants’ actual experiences. Nonetheless, some fairly consistent ethnic differences are apparent. African American families are often reported to have extended kin living in their households and extended kin are more likely to provide child care than in European American families (Hogan et al., 1990; Vandell et al., 2003). However, some studies suggest that European American families are more likely to receive financial support from extended kin than African American families (Hofferth, 1984) and that African American caregivers (especially mothers) often interact with children in distinctive ways (for examples, see Bartz & Levine, 1978; Berlin et al., 1995; Field & Widmeyer, 1981; Garcia Coll, 1990; Hofferth, 2003; Ispa et al., 2004). Low-income families are more likely to rely on informal child care (friends, family, and neighbors) than on formal child care (Hofferth et al., 1991; IWPR, 1996), although some researchers have not found a relationship between SES and reliance on relative care (Early & Burchinal, 2001) or between SES and vocal interaction (Fouts et al., 2007; Gottfried, 1984; Hart & Risley, 1995; Hoff, 2003; Leyendecker et al., 1997; Roopnarine et al., 2005).

The Present Study

The main purpose of the present study was to achieve greater understanding of variations and similarities in interactional patterns of infants and caregivers in differing socioeconomic (lower and middle) and ethnic (European American and African American) groups and to consider the forms and extent to which multiple caregivers (mothers, fathers, siblings, nonnuclear relatives, and nonrelatives) interacted with infants. Noting the paucity of observational research on families in different socioeconomic and ethnic contexts, we used lengthy (12-hour) observations to document the typical daily experiences of infants and their caregivers. The same approach has been adopted in many different cultural and socioeconomic contexts in Central Africa, Central America, Germany, Canada, and the United States (Fouts, Hewlett, & Lamb, 2005; Fouts et al., 2007; Fracasso et al., 1997; Hewlett, Lamb, Shannon, Leyendecker, & Scholmerich, 1998; Leyendecker et al., 1997; Roopnarine et al., 2005).

In line with the cultural-contextual/adaptive-resilient perspective, we attempted to study a broad range of interactions with infants. There has been far more emphasis on vocal interaction and stimulation than on physical interaction in the past (Gottfried, 1984; Hart & Risley, 1995; Hoff, 2003). Families from different cultural and socioeconomic backgrounds may value different styles of interaction and care, however (Super & Harkness, 1997). Thus, we chose to examine both vocal and physical interactions and a variety of behaviors including vocalizing, stimulating, physical affection, vocal affection, physical soothing, vocal soothing, physical caretaking, holding, and carrying (Leyendecker et al., 1997; Roopnarine et al., 2005). The display of affection and soothing are important behaviors to consider because they illustrate caregiver warmth and responsiveness, which have been associated with secure caregiver-child relationships (Ahnert, Pinquart, & Lamb, 2006; Ainsworth et al., 1978; de Wolff & van Ijzendoorn, 1997; Isabella & Belsky, 1991) and socioeconomic circumstances (Bakermans-Kranenburg et al., 2004). Physical caretaking, holding, and carrying are very demanding aspects of care during early infancy, and involvement in these activities may vary according to type of caregiver and culture (see Lamb, 2002, for a review of mother-father differences; Super & Harkness, 1997, for a review of cultural variations).

In this study, we sought to address the following research questions: (a) Who is available to interact with and provide care to infants during the day? (b) Do SES and/or ethnicity predict variations in the availability of various individuals to infants? (c) Do SES and/or ethnicity predict variations in patterns of interaction between various individuals and infants? (d) Do SES and/or ethnicity predict how different individuals interact with infants when they are present? Because prior studies (see McLoyd et al., 2005) have suggested that African American families tend to have more extended family networks than European American families, we expected that nonnuclear relatives would be more available to African American infants than European American infants, especially in low-income circumstances. Furthermore, we expected that different styles of caregiver-infant interactions would be associated with both ethnicity (African American and European American) and SES but did not make more specific predictions because of the contradictory prior findings about the associations among caregiving styles, ethnicity, and SES.

Method

Participants

Forty-one African American and 40 European American families with 3- to 4-month-old infants living in a midsize northeastern city in the United States participated in the study. These families, all of whom were born in the United States and spoke English as their first language, were recruited through advertisement in neighborhood newspapers, health clinics, churches, childcare centers, and other community centers, as well as by word-of-mouth. Each family received $60 for participating in the study.

Ethnicities were identified by infants’ parents and socioeconomic status was defined according to parental educational attainment, occupation, and income (Entwisle & Astone, 1994). Of the 41 African American families, 20 were of lower socioeconomic status and 21 were of middle socioeconomic status. Their data were collected between 1999 and 2002. Of the European American families, 20 were of lower socioeconomic status and 20 were of middle socioeconomic status. Their data were collected between 2002 and 2004. See Table 1 for demographic information.

Table 1.

Family Demographics

	African American		European American
	Lower SES n = 20	Middle SES n = 21	Lower SES n = 20	Middle SES n = 20
Female infant	10	12	10	12
Male infants	10	9	10	8
Mother’s age (mean)	23.05	30.29	27.10	33.90
Father’s age (mean)	25.58	34.25	24.55	35.55
% of first born infants	35.00	19.05	20.00	40.00
% of infants with one or more siblings	65.00	80.95	80.00	60.00
% parents married to each other	0.00	76.00	70.00	95.00
% father resident	20.00	76.00	75.00	95.00
% both parents employed	30.00	42.86	30.00	65.00
% father only employed	30.00	42.86	50.00	30.00
% mother only employed	15.00	14.29	10.00	5.00
% no employment	25.00	0.00	10.00	0.00
% mother, high school or less	55.00	14.00	40.00	0.00
% father, high school or less	69.00	25.00	35.00	15.00
% mother, some college	35.00	43.00	25.00	5.00
% father, some college	16.00	45.00	40.00	25.00
% mother, college degree	10.00	43.00	35.00	95.00
% father, college degree	16.00	30.00	25.00	60.00

All of the lower SES families were at or below the federal guidelines for low-income families (i.e., 150% of the poverty level or below). The middle SES families all had incomes above 150% of the poverty level; they averaged between 287% and 313%. As shown in Table 1, the middle SES families were more likely to have one or more parents employed than were lower SES families.

African American and European American parents in the lower SES groups were more likely to have graduated from or have less than a high school education than were families in the middle SES context. Lower SES parents were predominantly employed in jobs such as retail positions, various trades (e.g., builder, mechanic, maintenance), delivery, factory workers, waiter/waitresses, secretaries, child care providers, or barbers. Middle SES parents were mostly employed as civil servants, in various trades (e.g., builder, mechanic, maintenance), retail sales, bus drivers, fitness trainers, engineers, teachers, therapists, secretaries, librarians, architects, human resources, or managerial positions.

Middle SES parents were also more likely to be married than were lower SES parents. In middle SES families, fathers were more likely to be resident, although in both SES groups fewer African American than European American families had resident fathers. Likewise, in both SES groups European American parents were more likely to be married to each other than were African American parents, though this difference was most striking in the lower SES group because none of those parents were married to each other.

Procedure

The observations of African American families were conducted by five female graduate and undergraduate students who identified themselves as African American and were majoring in child development and early childhood special education. The observations of European American families were conducted by three female graduate students studying child development who self-identified as European Americans. Observer and participant ethnicity was matched in an effort to increase the comfort level of participants.

Each focal infant was observed on 4 different days for 3 hours (8 a.m. to 11 a.m., 11 a.m. to 2 p.m., 2 p.m. to 5 p.m., and 5 p.m. to 8 p.m.), so that 12 daytime hours were covered. Prior to the first observation, an observer visited the family to collect demographic data, explain the observation procedure, and answer questions about the procedures. The families were asked to go about their typical daily activities during the observations. Even though mothers were not asked to be present during observations, most mothers were present, and most of the observations took place in or near the homes.

Coding

Three-hour observation periods were divided into hourly segments, with 15-minute rest periods after 45 minutes to prevent observer fatigue. Observers recorded infant and caregiver behaviors on-the-mark at 30-second intervals using a detailed behavioral checklist. Observers listened to an audio recording via earphones instructing them to “observe” for a 20-second period and then “record” within a 10-second period the presence of observed behaviors indicated on a checklist. Each 30-second “observe and record” segment is hereafter referred to as an observation interval.

The behavioral checklist and operational definitions utilized in the present study were originally derived and adapted from Belsky, Gilstrap, and Rovine’s (1984) coding scheme and have been previously described in Hewlett et al. (1998) and Leyendecker et al. (1997). For each observation interval, caregivers available to or interacting with the focal 3- to 4-month-olds were identified as mother, father, sibling, nonnuclear relative, or nonrelative; caregivers’ age and gender were not recorded. The coding categories of the behavioral checklist included the following:

Individual available

All individuals present in the same room as the infant were recorded in each observational interval.

Vocalization to infant

A person vocalized to the infant in a typical talking manner (i.e., not “baby talk”) or using short, simple dialogue, vocalizing slowly or “baby talking” some words so that the infant would pay attention. The person could make animal sounds for the infant, sing to her or him, or change the intonation as a way of calling or attracting attention.

Stimulate/arousal of infant

A person attempted to draw or hold the infant’s attention in order to maintain alertness or the level of positive affect. The person might touch the infant with a finger to call attention, play with her or him, or make other efforts to stimulate the infant (e.g., by making him or her jump on lap; moving him or her, pulling limbs, etc.). When stimulation involved vocalizing to the infant, “vocalization to infant” was coded simultaneously. Stimulation was not coded if the individual was soothing the infant.

Affection to infant (vocal)

A person expressed positive affect to the infant verbally (e.g., praise, terms of endearment). This variable was reserved for clear demonstrations of love and affection, which went beyond simple statements of pleasure.

Affection to infant (physical)

A person expressed affect physically (e.g., touching, hugging, kissing, makes loving gestures, or patting infant affectionately) in clear demonstrations of love and affection.

Soothing of infant (vocal)

Using verbal expressions to distract the infant, a person tried to quiet or calm an irritable or crying infant. If the infant calmed for at least 20 seconds, then “vocal soothe” was no longer coded.

Soothing of infant (physical)

A person tried to physically quiet or calm an irritable or crying infant by, for example, rocking, patting, picking up, pushing the infant in a carriage, or showing the infant a toy. This behavior was only coded while the infant was irritable or crying. If the infant calmed for a period of 20 seconds, then “physical soothe” was no longer coded.

Infant fusses

The infant was awake and showed signs of agitation or being upset; she or he was bothered and emitted moans, whines, and whimpers but did not cry.

Infant cries

The infant was awake and made vocalizations, gestures, and motor behaviors that left no doubt that the infant was crying (i.e., tears or intense vocal cry). This variable was not coded if the infant was only whimpering, complaining, or restless.

Physical caregiving of infant

An individual performs physical care activities such as burping infant, changing diapers, dressing-undressing, brushing infant hair, wiping infant mouth or nose, or bathing the infant.

Holding infant

Infant is located on the person’s lap.

Carrying infant

Infant is being carried in an individual’s arms.

Several of the above codes were mutually exclusive. Soothing (physical or vocal) was not coded with affection (physical or vocal) or stimulate/arousal. Also, fussing and crying were not coded with affection (physical or vocal) or stimulate/arousal. All of the other codes could be coded simultanously.

Reliability

Observers were trained for 3 weeks before conducting observations in the field. During this time, observers viewed and reviewed together two 45-minute naturalistic video-recorded follows of two 3- to 4-month-old infants in their homes; observers then coded the videotapes independently before discussing disagreements. One training video was of a European American infant in a middle-SES family, and one video was of an African American infant in a lower- to middle-SES family. Observers continued practicing with the videos on their own. Once observers reached 90% agreement on every code with at least one other observer, they were permitted to conduct field observations. Over the course of the study, reliability checks were performed on eighteen 45-minute periods of six African American and six European American families, with reliability calculations conducted separately for each ethnic group, given that observers differed across the ethnic groups.

For the African American sample, Cohen’s kappa coefficients were infant fuss = .83, infant cry = .84, vocalization to infant = .85, stimulate/arousal of infant = .72, vocal affection = .81, physical affection = .63, vocal soothe = .88, physical soothe = .84, physical caregiving = 91.7, holding = 1.00, carrying = 1.00, and caregiver available = 1.00. For the European American sample, Cohen’s kappa coefficients were infant fuss = .58, infant cry = .75, vocalization to infant = .76, stimulate/arousal of infant = .93, vocal affection = .81, physical affection = .55, vocal soothe = .75, physical soothe = .82, physical caregiving = .85, holding = 1.00, carrying = 1.00, and caregiver available = .98. The low coefficients obtained for infant fuss and physical affect are probably due to the co-occurrence of these behaviors with other behaviors. For example, infant fuss often preceded crying, and physical affection often co-occurred with vocal affection. Nevertheless, it is unclear why the coding of fuss was only problematic in the European American sample.

Results

For all analyses, Cohen’s f was used to quantify effect sizes, and Cohen’s (1988) categorizations of effect sizes (i.e., small effect sizes range from .10 to .24, medium from .25 to .39, and large .40 or greater) were used to interpret the results. For the sake of brevity, only significant (p ≤ .05) and marginally significant (p < .06) effects and interactions are described. Complete descriptive data are presented in Tables 2, 3, and 4.

Table 2.

Caregiver Availability to Infants

	African American				European American
	Lower SES n = 20		Middle SES n = 21		Lower SES n = 20		Middle SES n = 20
Variables	M	SD	M	SD	M	SD	M	SD
Availability
Mother	82.17	19.52	80.53	27.93	72.56	16.85	60.97	24.56
Father	7.86	14.08	10.97	19.37	11.06	13.24	20.00	17.14
Sibling	26.85	29.85	29.62	23.44	40.73	29.91	31.73	32.73
Nonnuclear relatives	21.51	19.00	21.67	23.17	20.71	26.30	8.37	11.76
Nonrelatives	8.73	17.27	7.00	21.50	8.64	15.21	7.40	11.36

Note. Means are percentages of total observation.

Table 3.

	African American				European American
	Lower SES n =20		Middle SES n =21		Lower SES n =20		Middle SES n =20
Variables^a	M	SD	M	SD	M	SD	M	SD
Mother
Vocalizing	22.82	12.22	26.45	12.49	24.15	11.48	26.54	12.62
Stimulating	5.10	3.30	4.80	3.46	5.47	5.83	5.54	5.25
Verbal affection	1.32	1.28	2.82	5.40	4.00	3.60	3.99	3.57
Physical affection	10.54	6.56	9.27	6.89	16.75	12.99	14.87	9.79
Verbal soothing	3.27	2.47	2.96	1.69	4.39	3.21	2.62	2.04
Physical soothing	3.55	2.50	3.25	3.69	3.91	2.99	2.85	1.77
Caregiving	13.34	7.37	12.02	7.24	8.12	4.59	9.13	4.58
Hold	7.20	7.09	14.09	13.74	11.23	9.68	12.81	7.92
Carry	19.55	8.28	12.68	11.02	19.56	14.73	12.72	7.16
Father
Vocalizing	2.15	4.01	3.97	6.50	2.76	4.73	7.35	9.55
Stimulating	1.03	2.92	1.16	2.22	.74	1.73	2.86	5.51
Verbal affection	.13	.20	.77	2.45	.44	.83	.87	1.53
Physical affection	2.04	4.88	1.69	3.65	1.61	2.10	3.13	3.69
Verbal soothing	.29	.48	.59	1.18	.40	1.00	.88	1.26
Physical soothing	.28	.55	.71	1.43	.31	.85	.89	1.35
Caregiving	.62	1.14	1.70	3.52	.50	.85	2.27	3.26
Hold	.14	.43	.98	2.23	1.07	2.32	2.56	4.34
Carry	2.86	5.01	1.24	3.81	1.24	2.18	3.51	4.67
Sibling
Vocalizing	.31	.64	1.37	3.22	.75	1.06	.65	1.37
Stimulating	.13	.27	.68	1.36	.30	.64	.32	.57
Verbal affection	.00	.00	.05	.16	.09	.17	.18	.52
Physical affection	.23	.47	.90	1.51	1.00	1.41	.72	1.06
Verbal soothing	.05	.23	.21	.62	.04	.11	.00	.00
Physical soothing	.15	.47	.27	.60	.08	.17	.01	.06
Caregiving	.14	.35	.42	1.19	.09	.20	.06	.14
Hold	.18	.79	.43	.97	.33	1.08	.07	.22
Carry	.18	.42	.62	1.52	.03	.13	.10	.36
Nonnuclear relative
Vocalizing	5.95	7.86	2.79	5.02	3.83	6.46	2.25	4.42
Stimulating	1.65	2.22	1.08	2.96	.46	.82	.56	1.22
Verbal affection	.32	.56	.17	.69	.91	1.75	.63	1.87
Physical affection	2.77	4.64	1.63	4.68	2.68	5.65	2.09	4.01
Verbal soothing	.85	1.08	.51	1.06	.55	1.28	.28	.81
Physical soothing	1.18	1.73	.41	.67	.63	1.60	.33	.82
Caregiving	1.79	2.71	.94	1.89	.51	.93	.41	.78
Hold	1.37	3.21	1.21	3.36	1.14	3.22	1.10	2.23
Carry	2.92	3.51	2.05	5.20	2.10	4.82	1.64	3.84
Nonrelative
Vocalizing	2.22	5.57	.59	1.61	.76	1.50	1.84	4.48
Stimulating	.50	1.00	.11	.25	.25	.53	.40	1.14
Verbal affection	.04	.10	.02	.07	.07	.14	.14	.45
Physical affection	.73	2.07	.06	.14	.29	.62	.49	1.53
Verbal soothing	.20	.47	.27	1.13	.06	.15	.23	.67
Physical soothing	.24	.58	.01	.03	.08	.27	.15	.35
Caregiving	.98	3.16	.42	1.40	.06	.15	.50	1.33
Hold	.52	1.37	.00	.00	.53	1.59	.52	1.28
Carry	1.58	3.68	.67	2.18	.43	1.27	1.04	2.26
Relative rate of soothing to fussing and crying	.50	.16	.61	.24	.57	.22	.55	.16

Note. For vocalizing, stimulating, affection, soothing, and caregiving, means are percentages of intervals focal infants were awake and available for interaction. For holding and carrying, means are percentages of the total observation.

Table 4.

	African American				European American
Variables	Lower SES n =20		Middle SES n =21		Lower SES n =20		Middle SES n =20
Mother	M	SD	M	SD	M	SD	M	SD
Vocalizing	18.67	10.14	21.31	7.12	25.07	11.01	31.95	12.43
Stimulating	4.13	2.66	4.15	3.31	5.45	5.38	6.27	5.31
Verbal affection	1.09	1.12	2.05	3.88	4.42	4.13	4.75	3.56
Physical affection	8.40	5.29	7.57	4.91	17.58	11.18	18.31	10.97
Verbal soothing	2.49	1.66	2.58	1.54	4.50	3.42	3.22	2.63
Physical soothing	2.65	1.65	2.64	2.47	3.87	2.75	3.56	2.50
Caregiving	10.40	5.58	9.41	4.28	9.30	4.96	10.79	6.93
Hold	9.08	8.79	15.00	16.53	27.00	18.00	23.82	13.41
Carry	25.17	9.94	17.70	14.21	15.18	12.29	21.16	9.84
	n =10		n =13		n =15		n =19
Father	M	SD	M	SD	M	SD	M	SD
Vocalizing	4.76	9.07	12.75	14.20	14.27	13.21	17.70	16.17
Stimulating	1.44	4.56	3.97	7.30	3.00	4.87	5.89	8.95
Verbal affection	2.56	7.89	.28	.68	4.07	7.34	2.03	2.68
Physical affection	9.03	21.57	6.14	6.42	8.41	8.22	8.29	8.24
Verbal soothing	.19	.61	1.73	3.10	3.09	8.61	2.08	2.68
Physical soothing	.17	.55	2.60	6.77	2.54	7.33	2.20	3.06
Caregiving	.27	.85	5.35	13.81	2.41	3.26	4.85	5.62
Hold	.00	.00	7.14	17.19	9.79	15.17	13.03	14.94
Carry	11.24	24.74	4.46	6.86	7.04	7.70	9.49	15.39
	n =13		n =17		n =18		n =12
Sibling	M	SD	M	SD	M	SD	M	SD
Vocalizing	2.80	9.89	4.23	12.21	1.09	1.31	1.49	2.05
Stimulating	2.78	9.89	1.88	4.87	.39	.77	.84	1.12
Verbal affection	.00	.00	.14	.33	.14	.22	.48	.86
Physical affection	.02	.07	2.62	5.21	1.57	1.80	2.34	3.01
Verbal soothing	.00	.00	.79	2.63	.04	.13	.00	.00
Physical soothing	.00	.00	1.11	2.60	.15	.39	.02	.07
Caregiving	1.10	3.96	1.31	4.43	.26	.55	.13	.22
Hold	.00	.00	1.34	2.73	.06	.18	.31	.94
Carry	4.95	17.83	3.94	11.83	.58	1.87	.39	.62
	n =19		n =15		n =11		n =10
Nonnuclear relative	M	SD	M	SD	M	SD	M	SD
Vocalizing	14.92	15.66	9.96	13.87	21.16	23.81	17.81	17.80
Stimulating	3.93	4.71	4.75	9.42	3.85	6.26	4.24	6.78
Verbal affection	.58	.94	.64	2.37	4.52	5.12	2.86	5.57
Physical affection	5.38	7.18	3.84	4.56	12.18	14.20	15.96	15.79
Verbal soothing	2.11	2.62	1.70	2.45	2.22	3.45	1.77	2.65
Physical soothing	2.80	3.65	1.51	2.16	2.42	3.84	2.47	2.88
Caregiving	3.36	4.28	1.86	3.10	3.28	4.82	3.36	4.29
Hold	4.39	8.92	5.31	14.46	11.20	13.00	17.17	18.53
Carry	12.14	10.86	8.49	14.51	8.30	19.31	15.47	24.25
	n =14		n = 9		n =15		n =14
Nonrelative	M	SD	M	SD	M	SD	M	SD
Vocalizing	10.59	13.94	9.17	18.63	9.10	13.60	9.01	12.95
Stimulating	4.45	11.17	5.93	11.47	2.75	6.26	2.07	4.08
Verbal affection	.60	2.22	3.70	11.11	3.44	10.25	.64	1.59
Physical affection	3.21	5.59	4.22	11.02	2.90	6.07	1.73	3.57
Verbal soothing	1.14	2.57	.60	1.32	.51	1.10	.89	1.60
Physical soothing	1.96	3.67	.00	.00	.96	1.76	.58	.90
Caregiving	1.36	3.62	2.56	7.69	.55	.98	2.10	3.52
Hold	7.47	16.67	.00	.00	4.18	8.58	7.42	12.69
Carry	8.86	17.12	9.91	29.74	4.58	10.39	3.72	6.48

Note. Means are percentages of intervals focal infants were in the presence of a particular caregiver.

Who Was Available to Infants in the Different Ethnic and SES Groups? Did SES and/or Ethnicity Predict Variations in the Availability of Various Individuals?

Household composition and availability of caregivers

A 2 × 2 ANOVA of the number of extended kin living in households revealed a main effect for SES, F(1, 77) = 9.38, p ≤ .01, Cohen’s f = .35, as lower SES families had more extended kin living in their households than middle SES families did.

A 2 × 2 repeated measures ANOVA was conducted on the proportion of total observation intervals that each type of caregiver was available to infants (see Table 2). Only the main effect for ethnicity was significant, Wilks’ = .87, F(4, 74) = 2.89, p ≤ .05, Cohen’s f = .40. Univariate analyses indicated that only the availability of mothers varied significantly by ethnicity, F(1, 79) = 7.00, p ≤ .01, Cohen’s f = .30, with mothers present more frequently in African American than European American families regardless of SES.

Do SES and/or Ethnicity Predict Variations in the Infants’ Experiences With Different Caregivers?

To determine whether SES and/or ethnicity predicted infants’ experiences with various caregivers, types of caregiver interactions were grouped for a series of analyses and each caregiver was examined separately because the purpose of these analyses was to assess infant experience rather than to compare caregiver involvement. Conceptual groupings of the codes included (a) social interactions (vocalizing, stimulating, physical affection, vocal affection), (b) responses to fussing or crying (physical and vocal soothing), and (c) caretaking behaviors (caregiving, holding, carrying). The engagement and responsive behavioral codes (vocalizing, stimulating, physical affection, vocal affection, physical soothing, vocal soothing, caregiving) were prorated to represent the proportion of observational intervals that the infant was awake and therefore available for engagement; holding and carrying often occurred when the infant was awake or asleep and were prorated relative to the amounts of time each infant was observed. In order to avoid inflating proportions of interactions, behavioral codes within each group were not compiled into combined scores but instead entered separately in each analysis. For example, in the social interaction analysis, the codes vocalizing, stimulating, physical affection, and vocal affection were entered separately.

Descriptive statistics for all behavioral variables (vocalize, stimulate, physical affection, vocal affection, physical soothing, vocal soothing, caregiving, holding, carrying) are shown in Table 3.

Mothers

A multivariate analysis of the social interactions involving mothers (vocalizing, stimulate, physical affection, vocal affection; see Table 3) revealed a main effect for ethnicity, Wilks’ = .87, F(4, 74) = 2.88, p ≤ .05, Cohen’s f = .40. Univariate analyses revealed that European American infants experienced more physical and vocal affection from mothers than did African American infants, F(1, 77) = 7.99, p ≤ .01, Cohen’s f = .32, F(1, 797) = 5.16, p ≤ .05, Cohen’s f = .26, respectively.

Similarly, a multivariate analysis of the caretaking involving mothers (caregiving, holding, carrying; see Table 3) revealed a main effect for ethnicity, Wilks’ = .87, F(3, 75) = 3.87, p ≤ .01, Cohen’s f = .39. Univariate analyses revealed that African American infants received more caregiving from mothers than European American infants did, F(1, 77) = 8.88, p ≤ .001, Cohen’s f = .34. There was also a significant multivariate main effect for SES, Wilks’ = .89, F(3, 75) = 3.00, p ≤ .05, Cohen’s f = .35. Infants in lower SES families were carried more by mothers than infants in middle SES families, F(1, 77) = 8.30, p ≤ .01, Cohen’s f = .33, whereas infants in middle SES families were held by their mothers more often than infants in lower SES families, though this difference was marginally significant, F(1, 77) = 3.62, p ≤ .06, Cohen’s f = .22.

Fathers

The multivariate analysis of caretaking involving fathers (caregiving, hold, carry; see Table 3) revealed a significant main effect for SES, Wilks’ = .90, F(3, 75) = 2.73, p ≤ .05, Cohen’s f = .33. Univariate analyses revealed that infants in middle-SES families were more likely to receive caregiving from or be held by their fathers than infants in lower SES families (caregiving: F[1, 77] = 6.51, p ≤ .01, Cohen’s f = .29; holding: F[1, 77] = 3.80, p ≤ .05, Cohen’s f = .22).

Nonnuclear relatives

Multivariate analysis of the social interactions involving nonnuclear relatives (vocalizing, stimulate, physical affection, vocal affection; see Table 3) revealed a main effect for ethnicity, Wilks’ = .84, F(4, 74) = 3.49, p ≤ .01, Cohen’s f = .43. Univariate analyses showed that African American infants experienced more stimulation from nonnuclear relatives than European American infants did, though this difference was marginally significant, F(1, 77) = 3.66, p ≤ .06, Cohen’s f = .22.

For caretaking (caregiving, holding, carrying; see Table 3) provided by nonnuclear relatives, there was a significant multivariate main effect for ethnicity, Wilks’ = .89, F(3, 75) = 2.99, p ≤ .05, Cohen’s f = .35. Univariate analyses revealed that infants in African American families were more likely to receive caregiving from nonnuclear relatives than infants in European American families, F(1, 77) = 5.30, p ≤ .05, Cohen’s f = .26.

Did SES and/or Ethnicity Predict Differences in How Individuals Interacted With Infants as a Function of Caregiver Presence?

To examine how different individuals interacted with the infants, caregiver involvement was examined as a function of presence by prorating each type of involvement (vocalize, stimulate, physical affection, vocal affection, physical soothing, vocal soothing, caregiving, holding, carrying) relative to the number of observation intervals during which each caregiver was present. By viewing caregiver involvement as a function of presence, we were able to identify styles of caregiving while controlling for variations in the availability of each caregiver. Thus, for the analysis involving each type of caregiver, only infants who had been observed in the presence of that particular type of caregiver were included. Because all infants were observed with their mothers, the “mother analysis” includes the entire sample. Table 4 provides descriptive data for each caregiver as a function of presence, and the truncated samples for each caregiver analysis are specified there. As with the previous set of analyses, three separate 2 × 2 (SES × Ethnicity) MANOVAs were conducted for each caregiver including a MANOVA for social interactions, soothing responses, and caretaking. As in the last set of analyses, behavioral codes within each group were not compiled into combined scores but instead entered separately in each analysis.

Mothers

A multivariate analysis of social interactions involving mothers (vocalizing, stimulate, physical affection, vocal affection; see Table 4) as a function of presence revealed a main effect for ethnicity, Wilks’ = .70, F(4, 74) = 8.00, p ≤ .001, Cohen’s f = .43. Univariate analyses showed that European American mothers were more likely to vocalize and show vocal and physical affection to infants than African American mothers (vocalize: F[1, 77] = 13.77, p ≤ .001, Cohen’s f = .42; physical affection: F[1, 77] = 27.24, p ≤ .001, Cohen’s f = .59; vocal affection: F[1, 77] = 15.88, p ≤ .001, Cohen’s f = .45).

A multivariate analysis of mothers’ caretaking (caregiving, holding, carrying; see Table 4) of infants as a function of presence showed a significant main effect for ethnicity, Wilks’ = .81, F[3, 75] = 5.76, p ≤ .001, Cohen’s f = .48. When present, European American mothers were more likely to hold infants than were African American mothers, F[1, 77] = 16.88, p ≤ .001, Cohen’s f = .47.

Siblings

The multivariate analysis of sibling social interactions (see Table 4) as a function of presence revealed a significant main effect for ethnicity, Wilks’ = .83, F(4, 53) = 2.81, p ≤ .05, Cohen’s f = .46. Siblings in European American families were more likely to show vocal affection to infants than were siblings in African American families, F(1, 56) = 4.37, p ≤ .05, Cohen’s f = .28.

Nonnuclear relatives

A multivariate analysis of relative’s social interactions (see Table 4) as a function of presence revealed a main effect for ethnicity, Wilks’ = .71, F(4, 48) = 4.87, p ≤ .01, Cohen’s f = .41. Univariate analyses showed that nonnuclear relatives in European American families tended to spend more of their time with infants engaging them in physical and vocal displays of affection than nonnuclear relatives in African American families (physical: F[1, 53] = 10.76, p ≤ .01, Cohen’s f = .46; vocal: F[1, 53] = 9.81, p ≤ .01, Cohen’s f = .44).

Discussion

These analyses make clear that both ethnicity and SES are related to child care practices. For example, African American mothers were observed to be with their infants more often than were European American mothers, while infants in lower SES families were more likely to have co-resident extended family members in their households than were infants in middle-SES families. There were also ethnic differences in maternal caregiving and affection as well as in levels of caregiving and stimulation by nonnuclear relatives, while SES differences were identified for paternal caregiving and holding, and maternal holding and carrying. When we focused only on the times caregivers were present, there were ethnic differences in levels of maternal vocalizing and affection, holding and affection from nonnuclear relatives, and affection from siblings.

Further underscoring the importance of both ethnicity and SES, there were equal numbers of effects for each variable (each variable predicted variation in five behavioral patterns), suggesting that neither SES nor ethnicity is the most important factor but that child care practices are fitted to economic constraints in concert with cultural contexts. Like Kontos et al. (1997), we found no statistical interactions between ethnicity and SES, a finding which supports their claim that ethnicity acts as a main effect rather than as a moderator of socioeconomic effects. Some scholars have suggested that there may be a “middle-class” cultural practice that emphasizes adult-child conversations or vocal interactions and child-focused activities related to scholastic goals (Lareau, 2003; Morelli et al., 2003). Our examination of infant care practices does not precisely reflect these characteristics, because SES was not a predictor of vocal interactions. Perhaps this inconsistency is due to variation in methods and contexts of observation and because we only focused on 3- to 4-month-old infants, whereas Morelli et al. (2003) and Lareau (2003) studied children in early and middle childhood, respectively. In our study, middle-SES families exemplified higher levels of parents holding infants in a stationary manner rather than carrying infants around. Holding in one position may be indicative of focused adult attention and is perhaps consistent with a middle-class practice of child-focused activities. However, observational studies of later infancy and toddlerhood that distinguish between SES and ethnicity are needed to identify continuity between these infant care practices and those in early and middle childhood.

Ethnic differences, especially between African American and European American families, have been emphasized for decades, first from a deficit perspective and then with greater appreciation of differences and strengths (for a review of deficit and comparative models, see McLoyd & Randolph, 1985; McLoyd et al., 2005), but in many studies, SES and ethnicity were confounded. In addition, most researchers have focused only on mothers. Our extended observations of lower and middle-SES families in two ethnic groups thus provide unique information about patterns of interactions with diverse individuals in the home environment.

To our knowledge, no other study has shown that African American mothers are more available to infants than European American mothers and that African American infants receive more care from their mothers than European American infants do, regardless of socioeconomic circumstances. The greater availability of African American mothers to infants does not appear to be related to mothers’ work status or number of other children in the household, because these scores were similar in the two ethnic groups. As in previous studies, infants in European American families experienced more maternal affection than infants in African American families (Berlin et al., 1995; Hofferth, 2003; Ispa et al., 2004). A cultural-contextual perspective would suggest that differences in availability, along with caregiving (burping, dressing, cleaning, grooming, hair combing and arranging, etc.) and affection, may be mediated by cultural expectations about how infants should be cared for, how they should appear, and how they develop. For example, African American mothers tend to be concerned about not “spoiling” children and expect children to be responsible for their feelings earlier than European American mothers (Bartz & Levine, 1978; Field & Widmeyer, 1981).

As with the ethnic differences in infants’ experiences with mothers, the SES differences appear to reflect different styles of caregiving, because infants in lower SES families were carried by their mothers more, whereas infants in middle-SES families were held by their mothers more. These SES differences have been reported in previous research on African American families (Fouts et al., 2007; Roopnarine et al., 2005) and may reflect that caregivers in lower SES families care for infants while also attending to other social, family, or work responsibilities; frequent carrying may thus be a strategy adapted to the ecological constraints encountered by lower SES families.

It is noteworthy that there were no ethnic differences with respect to paternal interactions with infants even though African American fathers in both SES groups were more likely to be nonresident and less likely to be married than were European American fathers. This finding is not consistent with the suggestion that nonresidential and unmarried fathers are less involved than resident fathers (for a review of how nonresident African American fathers have been depicted as uninvolved, see Roopnarine, 2004; Tamis-LeMonda & McFadden, 2010). By contrast, socioeconomic differences in infants’ experiences with fathers were evident with respect to physical caregiving and holding: Infants in middle-SES families experienced more paternal caregiving and holding than infants in lower SES families. This is consistent with other findings showing that poverty is associated with reduced paternal involvement (for a review, see Nelson, 2004).

As expected, infants in African American families were more likely to experience physical caretaking and stimulation by nonnuclear relatives than were infants in European American families. Given the prominence of kin networks reported in other studies of African American families, however, it was somewhat surprising that there were so few ethnic differences in involvement by nonnuclear relatives. Most previous studies have involved interviews and surveys of parents (mostly mothers) (Hogan et al., 1990; Vandell et al., 2003), and it is possible that nonnuclear relatives provide support in ways (e.g., providing care to older children and providing financial and emotional support) that would not be evident during home observations.

When caregiver interactions with infants were viewed as a function of caregiver presence, only ethnic differences were found, with higher levels of vocalizing, affection, and holding by mothers and higher levels of affection by siblings and nonnuclear relatives in European American than in African American families. These findings are consistent with the results of previous studies showing that European American caregivers tend to show more vocal warmth to children than African American caregivers do (Berlin et al., 1995; Garcia Coll, 1990; Hofferth, 2003; Ispa et al., 2004). However, these findings should be interpreted with caution, because they are largely inconsistent with our analyses involving the infants’ daily experiences, rather than those that controlled for the presence of particular caregivers. Our findings suggest that limiting the focus of research to those times when caregivers are present may mask important variations while exaggerating ethnic differences.

Although the current study provides some novel and unique findings with at least moderate effect sizes, it has some limitations. The sample size was relatively small and our results should not be generalized to other ethnic groups or SES contexts. Instead, they should be used to inform the design of larger observational studies examining ethnic and SES differences in infants’ and children’s everyday experiences. Because we did not collect detailed data on the identity of various caregivers, we cannot address hypotheses related to the age and gender of caregivers either. Furthermore, even though several significant main effects were identified for both ethnicity and SES, we were not able to determine how meaningful these differences were from the families’ perspectives.

These limitations aside, there are several implications of this study that apply to cross-cultural and child- and family-focused research. First, our findings raise doubts about the focus on either participants’ ethnic or socioeconomic backgrounds. For example, presuming that socioeconomic patterns generalize across ethnicities or studying one ethnicity within one context and suggesting that those patterns are characteristic of that ethnic group in general is not advisable. It is also important for researchers to consider many types of interactions, rather than focusing primarily on one type of behavior such as vocal interactions. Finally, limiting observations to the presence of certain caregivers may mask the variation in child care practices within ethnic groups, and thus, we recommend that researchers avoid presuming maternal or parental primacy and instead document the caregiving experiences that infants and children have with a variety of caregivers.

Footnotes

Acknowledgements

This study was supported by funds from the National Institute of Child Health and Human Development awarded to Michael E. Lamb. We are grateful to the families who participated in the present study.

The authors declared that they had no conflicts of interests with respect to their authorship or the publication of this article.

This research was supported in part by the Intramural Research Program at the National Institute of Child health and Human Development (Section on Social and Emotional Development).

References

Ahnert

Pinquart

Lamb

M. E.

(2006). Security of children’s relationships with nonparental care providers: A meta-analysis. Child Development, 74, 664-679.

Ainsworth

M. D.

Blehar

Waters

Wall

(1978). Patterns of attachment. Hillsdale, NJ: Erlbaum.

Bakermans-Kranenburg

van Ijzendoorn

M. H.

Kroonenberg

P. M.

(2004). Differences in attachment security between African-American and White children: Ethnicity or socioeconomic status? Infant Behavior and Development, 27, 417-433.

Bartz

K. W.

Levine

B. S.

(1978). Child rearing by Black parents: A description and comparison to Anglo and Chicano parents. Journal of Marriage and the Family, 40, 709-720.

Belle

D. E.

(1983). The impact of poverty on social networks and supports. Marriage & Family Review, 5, 89-103.

Belsky

Gilstrap

Rovine

(1984). The Pennsylvania Infant and Family Development Project 1: Stability and change in mother-infant and father-infant interaction in a family setting at one, three, and nine months. Child Development, 55, 692-705.

Benin

Keith

V. M.

(1995). The social support of employed African American and Anglo mothers. Journal of Family Issues, 16, 275-297.

Berlin

L. J.

Brooks-Gunn

Spiker

Zaslow

(1995). Examining observational measures of emotional support and cognitive stimulation in Black and White mothers of preschoolers. Journal of Family Issues, 16, 664-686.

Bradley

R. H.

Corwyn

R. F.

McAdoo

H. P.

García Coll

(2001). The home environments of children in the United States Part I: Variations by age, ethnicity, and poverty status. Child Development, 72, 1844-1867.

10.

Cohen

(1988). Statistical power analysis for the behavioral sciences (2nd ed.). New York: Academic Press.

11.

de Wolff

M. S.

van IJzendoorn

M. H.

(1997). Sensitivity and attachment: A meta-analysis on parental antecedents of infant attachment. Child Development, 68, 571-591.

12.

Dilworth-Anderson

(1992). Extended kin networks in black families. Generations, 16, 29-32.

13.

Early

D. M.

Burchinal

M. R.

(2001). Early childhood care: Relations with family characteristics and preferred care characteristics. Early Childhood Research Quarterly, 16, 475-497.

14.

Entwisle

D. R.

Astone

N. M.

(1994). Some critical guidelines for measuring youths’ race/ethnicity and socioeconomic status. Child Development, 65, 1521-1540.

15.

Field

T. M.

Widmeyer

S. M.

(1981). Mother-infant interactions among lower SES black, Cuban, Puerto Rican, and South American immigrants. In Field

T. M.

Sostek

A. M.

Vietze

Liederman

P. H.

(Eds.), Culture and early interactions (pp. 41-60). Hillsdale, NJ: Erlbaum.

16.

Fong

G. F.

DeBaryshe

Yuan

S. C. W.

Yuen

Caulfield

Nemoto

M. A.

Hisatake

(2004). Factors affecting choice of kith and kin care by families receiving child-care subsidies. Hülili: Multidisciplinary Research on Hawaiian Well-Being, 1, 93-106.

17.

Fouts

H. N.

Hewlett

B. S.

Lamb

M. E.

(2005). Parent-offspring weaning conflicts among the Bofi farmers and foragers of Central Africa. Current Anthropology, 46, 29-50.

18.

Fouts

H. N.

Roopnarine

J. L.

Lamb

M. E.

(2007). Social experiences and daily routines of African American infants in different socioeconomic contexts. Journal of Family Psychology, 21, 655-664.

19.

Fracasso

M. P.

Lamb

M. E.

Schölmerich

Leyendecker

(1997). The ecology of mother-infant interaction in Euro-American and immigrant Central American families living in the United States. International Journal of Behavioral Development, 20, 207-217.

20.

Garcia Coll

C. T.

(1990). Developmental outcome of minority infants: A process-oriented look into our beginnings. Child Development, 61, 270-289.

21.

Gottfried

A. W.

(1984). Home environment and early cognitive development: Integration, meta-analyses, and conclusions. In Gottfried

A. W.

(Ed.), Home environment and early cognitive development: Longitudinal research (pp. 329-342). New York: Academic Press.

22.

Hart

Risley

T. R.

(1995). Meaningful differences in the everyday experience of young American children. Baltimore, MD: Paul H. Brookes Publishing.

23.

Hays

W. C.

Mindel

C. H.

(1973). Extended kinship relations in black and white families. Journal of Marriage and the Family, 35, 51-57.

24.

Heath

S. B.

(1983). Ways with words: Language, life, and work in communities and classrooms. New York: Cambridge University Press.

25.

Henley

J. R.

Lyons

(2000). The Negotiations of child care and employment demands among low-income parents. Journal of Social Issues, 56, 683-706.

26.

Hewlett

B. S.

Lamb

M. E.

Shannon

Leyendecker

Scholmerich

(1998). Culture and infancy among central African foragers and farmers. Developmental Psychology, 334, 653-661.

27.

Hirshberg

Huang

D. S.

Fuller

(2005). Which low-income parents select child-care? Family demand and neighborhood organizations. Children and Youth Services Review, 27, 1119-1148.

28.

Hoff

(2003). Causes and consequences of SES-related differences in parent-child speech. In Bornstein

M. C.

Bradley

R. H.

(Eds.), Socioeconomic status, parenting, and child development (pp. 147-160). Mahwah, NJ: Lawrence Erlbaum Associates.

29.

Hofferth

S. L.

(1984). Kin networks, race, and family structure. Journal of Marriage and Family, 46, 791-806.

30.

Hofferth

S. L.

(1995). Caring for children at the poverty line. Children and Youth Services Review, 17, 61-90.

31.

Hofferth

S. L.

Brayfield

A. A.

Deich

S. G.

Holcomb

P. A.

(1991). National child care survey, 1990. Washington D.C.: Urban Institute Press.

32.

Hofferth

S. L.

(2003). Race/ethnic differences in father involvement in two-parent families: Culture, context, or economy? Journal of Family Issues, 24, 185-216.

33.

Hogan

D. P.

Hao

Parish

W. L.

(1990). Race, kin networks, and assistance to mother-headed families. Social Forces, 68, 797-812.

34.

Huston

A. C.

Chang

Y. E.

Gennetian

(2002). Family and individual predictors of child care use by low-income families in different policy contexts. Early Childhood Research Quarterly, 17, 441-469.

35.

Institute for Women’s Policy Research (IWPR). (1996). Child care usage among low-income and AFDC families. Washington, DC: Author.

36.

Isabella

Belsky

(1991). Interactional synchrony and the origins of infant-mother attachment: A replication study. Child Development, 62, 373-384.

37.

Ispa

J. M.

Fine

M. A.

Halgunseth

L. C.

Harper

Robinson

Boyce

Brooks-Gunn

Brady-Smith

(2004). Maternal intrusiveness, maternal warmth, and mother-toddler relationship outcomes: Variations across low-income ethnic and acculturation groups. Child Development, 75, 1613-1631.

38.

Jackson

J. F.

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

39.

Kontos

Howes

Shinn

Galinsky

(1997). Children’s experiences in family child care and relative care as a function of family income and ethnicity. Merrill-Palmer Quarterly, 43, 386-403.

40.

Kuhlthau

Mason

K. O.

(1996). Market child care versus care by relatives: Choices made by employed and nonemployed mothers. Journal of Family Issues, 17, 561-578.

41.

Lamb

M. E.

(2002). Infant-father attachments and their impact on child development. In Tamis-LeMonda

C. S.

Cabrera

(Eds.), Handbook on father involvement: Multidisciplinary perspectives (pp. 93-117). Mahwah, NJ: Erlbaum.

42.

Lamb

M. E.

(Ed.). (2010). The role of the father in child development (5th ed.). Hoboken, NJ: John Wiley & Sons, Inc.

43.

Lareau

(2003). Unequal childhoods: Class, race, and family life. Berkeley: University of California Press.

44.

Lehrer

(1983). Determinants of child care mode choice: An economic perspective. Social Science Research, 12, 69-80.

45.

Leyendecker

Lamb

M. E.

Schölmerich

Fricke

D. M.

(1997). Contexts as moderators of observed interactions: A study of Costa Rican mothers and infants from differing socio-economic backgrounds. International Journal of Behavioral Development, 21, 15-34.

46.

Li-Grining

C. P.

Coley

R. L.

(2006). Child care experiences in low-income communities: Developmental quality and maternal views. Early Childhood Research Quarterly, 21, 125-141.

47.

Low

(2005). Families: An evolutionary anthropological perspective. In Roopnarine

J. L.

Gielen

(Eds.), Families in global perspectives (pp. 14-32). Boston: Allyn & Bacon.

48.

Miller-Loncar

C. L.

Erwin

L. J.

Landry

S. H.

Smith

K. E.

Swank

P. R.

(1998). Characteristics of social support networks of low socioeconomic status African American, Anglo American, and Mexican American mothers of fullterm and preterm infants. Journal of Community Psychology, 26, 131-143.

49.

McLoyd

V. C.

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

50.

McLoyd

V. C.

Hill

N. E.

Dodge

K. A.

(Eds.). (2005). African American family life: Ecological and cultural diversity. New York: Guilford.

51.

McLoyd

Randolph

(1985). Secular trends in the study of Afro-American children: A review of child development, 1936-1980. Monographs of the Society for Research in Child Development: History and Research in Child Development, 50, 78-92.

52.

Morelli

G. A.

Rogoff

Angelillo

(2003). Cultural variation in young children’s access to work or involvement in specialised child-focused activities. International Journal of Behavioral Development, 27, 264-274.

53.

Murray

V. M.

Smith

E. P.

Hill

N. E.

(2001). Introduction to the special section: Race, ethnicity, and culture in studies of families in context. Journal of Marriage and Family, 63, 911-914.

54.

Nelson

T. J.

(2004). Low-income fathers. Annual Review of Sociology, 30, 427-451.

55.

Ogbu

J. U.

(1978). Minority education and caste: The American system in cross-cultural perspective. San Diego, CA: Academic Press.

56.

Ogbu

J. U.

(1981). Origins of human competence: A cultural-ecological perspective. Child Development, 52, 413-429.

57.

Phinney

J. S.

(1996). When we talk about American ethnic groups what do we mean? American Psychologist, 51, 918-927.

58.

Rogoff

(2003). The cultural nature of human development. New York: Oxford University Press.

59.

Roopnarine

J. L.

(2004). African American and African Caribbean fathers: Levels, quality and meaning of involvement. In Lamb

M. E.

(Ed.), The role of the father in child development (4th ed., pp. 58-97). Hoboken, NJ: Wiley.

60.

Roopnarine

J. L.

Fouts

H. N.

Lamb

M. E.

Lewis

T. E.

(2005). Mothers’ and fathers’ behaviors toward their 3-4-month-old infants in low-, middle- and upper-socioeconomic African American families. Developmental Psychology, 41(5), 723-732.

61.

Stack

C. B.

(1974). All our kin: Strategies for survival in a Black community. New York: Harper & Row, Publishers, Inc.

62.

Super

Harkness

(1997). The cultural structuring of child development. In Berry

Dasen

Saraswathi

T. S.

(Eds.), Handbook of cross-cultural psychology: Basic processes and human development (pp. 1-39). Needham, MA: Allyn & Bacon.

63.

Tamis-LeMonda

C. S.

McFadden

K. E.

(2010). Fathers from low-income backgrounds: Myths and evidence. In Lamb

M. E.

(Ed.), The role of the father in child development (5th ed.). Hoboken, NJ: John Wiley & Sons, Inc.

64.

Vandell

D. L.

McCartney

Owen

M. T.

Booth

Clarke-Stewart

(2003). Variations in child care by grandparents during the first three years. Journal of Marriage and Family, 65, 375-381.

65.

Ward

M. C.

(1971). Them children: A study in language and learning. New York: Holt, Rinehart, and Winston, Inc.

66.

Wilson

M. N.

(1989). Child development in the context of the Black extended family. American Psychologist, 44, 380-385.