Low-income fathers' involvement in their toddlers' lives: biological fathers from the Early Head Start research and evaluation study.

In this article, we use data on biological fathers (n = 597) and mothers (N = 1,550) from 12 sites of the Early Head Start Research and Evaluation Project (EHS study) to examine the type and frequency of father involvement. We use a three-part model of father involvement to examine whether fathers participating in the EHS study are accessible to, engaged with, and show responsibility for their two-year-old children. We also examine patterns of reported father involvement by relationship status and residency. We find that more than 80% of all two-year-old children in the EHS study have accessible biological fathers, with the majority of nonresident boyfriends and nonresident friends and more than a third of fathers in no relationship with the mother seeing their children at least once in three months. These accessible fathers are engaged in a range of activities and show responsibility for their children, although patterns vary by the father-mother relationship status and father residency. More specifically, fathers who had at least a romantic relationship with the mother were more involved with their children across types of involvement than those in no relationship. Associations between relationship status and father engagement and responsibility remained after controlling for demographic variation among fathers in different relationship groups. A significant proportion of fathers who had no relationship with the mother of their child had some contact with the child, suggesting that the relationship between mother and father is not the only factor helping fathers stay involved in their children's lives. Finally, fathers report doing a lot more caregiving than has been suggested by other studies.

Keywords: father involvement, Early Head Start Research and Evaluation Project (EHS study), father-mother relationship status, father residency

**********

Over the past three decades, research on fathers has shown consistently positive associations between father involvement and children's academic achievement, peer relations, cognitive development, and behavioral and emotional regulation (e.g., Amato & Rivera, 1999; Cabrera, Tamis-LeMonda, Bradley, Hofferth, & Lamb, 2000; Lamb, 2004; Tamis-LeMonda & Cabrera, 2002). These findings also hold true for nonresidential fathers, among whom the quantity and quality of father involvement has been linked to better social, emotional, and cognitive outcomes for children (Furstenberg & Harris, 1993; Hetherington, 1991). The bulk of the work on paternal involvement in married families, however, has been conducted with middle-class samples, while research on unmarried fathers tends to use mothers as proxy respondents for fathers and one-dimensional definitions of father involvement (present vs. absent, mother divorced vs. married, never- vs. ever-married) (Garfinkel, McLanahan, Meyer, & Seltzer, 1998; Garfinkel, McLanahan, & Robins, 1994). Although recent qualitative studies of low-income men present a more balanced picture of nonresident fathers (Coley, 2001; Jarrett, Roy, & Burton, 2002; Nelson, Clampet-Lundquist, & Edin, 2002), the little survey research that exists tends to emphasize low rates of child support payment and contact with children (Cabrera et al., 2000; McLanahan & Sandefur, 1994; Wu, Bumpass, & Musick, 2001).

These representations stem in part from the lack of national data on low-income, unmarried fathers collected from the men themselves. National surveys, where mothers are the respondents, are typically designed with the traditional two-parent household as the sampling unit. These surveys ignore demographic shifts in family structure, particularly the rise in single motherhood and nonresidential fatherhood (McLanahan & Sandefur, 1994; Wu et al., 2001). In national surveys, undercount rates are reported to be higher for unmarried and nonresident fathers, who are disproportionately low-income and non-White (Cherlin & Griffith, 1998), partly because nonresident fathers are more likely to have unstable housing, be incarcerated (Cherlin & Griffith, 1998; Western & McLanahan, 2000), and not know of or acknowledge their paternity (Peters, Argys, Brooks-Gunn, & Smith, 1998). These circumstances hinder their participation in national fatherhood studies, which are often used by researchers to report on family trends and child well-being. As a result, most information on father involvement among nonresident men comes from simplistic, mother-report data.

This gap in research became the highlight of a major federal initiative to reorganize the way national datasets collect data on fathers and families (Cabrera, Brooks-Gunn, Moore, West, Boiler, & Tamis-LeMonda, 2002). This initiative included a federal mandate to involve fathers in research, programs, and services. The result is the availability of unprecedented data on men and their interactions and relationships with their partners and children (see Cabrera et al., 2002, for an in-depth description of these datasets). For example, the Fragile Families and Child Wellbeing Study (FF) follows a birth cohort of approximately 3,700 nonmarital and 1,200 marital births and will provide information on unwed mothers, fathers, and their relationships (Brooks-Gunn, Berlin, Leventhal, & Fuligni, 2000). The Early Head Start Research and Evaluation Project (EHS study) is a study of approximately 3,000 families enrolled in Early Head Start and uses an experimental design to understand the effects of an early intervention program on parents and children. This is one of the first national evaluation studies that includes both mothers and fathers of young children (Cabrera et al., 2002). The Early Childhood Longitudinal Birth Cohort Study (ECLS-B) is the first study in the United States to track a nationally representative sample of 10,688 children from infancy through school entry.

Most of these datasets are newly available to the public, and researchers have used preliminary data to explore fathering patterns, especially among low-income, unmarried fathers (Carlson & McLanahan, 2002a; Carlson & McLanahan, 2002b; Johnson, 2001; Wilson & Brooks-Gunn, 2001). In addition, new analyses from other nationally representative samples such as the National Survey of Families and Households and the National Survey of American's Families suggest that low-income nonresident fathers see their children more often than previously believed (Cooksey & Craig, 1998; Manning, Stewart, & Smock, 2003; Mincy & Oliver 2003). These findings provide valuable information about visitation patterns among low-income, nonresident fathers but do not speak to how poor resident and nonresident fathers interact with their children.

In this article, we use data on the fathers in the EHS study to explore biological father involvement in the lives of young children. To this end, we briefly discuss the literature on father involvement among poor fathers. In the results section, we report that the majority of two-year-old children in the EHS study have fathers who are engaged with them in positive, supportive ways, although patterns vary by father-mother relationship and residency. We end the article with a discussion of the implications of our findings for research, policy, and practice.

This study advances the field in a number of ways. First, it provides a demographic profile of a large, diverse sample of low-income biological fathers. Previous studies on father involvement among low-income fathers were mainly qualitative and/or focused on small samples of urban, predominantly African-American men (Black, Dubowitz, & Starr, 1999; Coley & Chase-Lansdale, 1999; Furstenberg & Harris, 1993; Jarrett et al., 2002; Nelson et al., 2002). Second, this study provides information on ways men in different family structures interact with their children and sets the stage for further work on the impact of this involvement on children's outcomes. Third, it uses data collected from fathers themselves rather than using mothers' reports about the fathers' activities. While mothers' perspectives are important, they do not capture how men see themselves as fathers. Using father reports was a specific recommendation from a national report on how to improve the knowledge base on fatherhood (Nurturing Fatherhood: Improving Data and Research on Male Fertility, Family Formation and Fatherhood, 1998).

What we know about how father involvement influences child development comes mostly from studies about the effects of father absence on children's outcomes. These studies largely reported negative effects of father absence due to non-marital childbirth, divorce, or more general nonresidency. From these findings, researchers inferred the converse--that present fathers might have positive developmental influences on their children (Lamb, 2004; Lewis, 1997). By the mid-1990s, a large body of work consistently associated father absence, mainly due to divorce, with their children's poor school achievement, diminished involvement in the labor force, early childbearing, and heightened levels of risk-taking behavior (Nurturing Fatherhood: Improving Data and Research on Male Fertility, Family Formation and Fatherhood, 1998). Children in father-absent homes are more likely to live in communities characterized by high levels of unemployment and crime, poor-quality housing, schools, and health care, and a lack of resources than children from intact homes, all of which are associated with poorer child outcomes (Duncan & Brooks-Gunn, 1997). In particular, boys growing up without fathers seem especially prone to exhibit problems in the areas of sex-role and gender-identity development, school performance, psychosocial adjustment, and self-control (Hetherington & Stanley-Hagan, 1995; Hetherington & Henderson, 1997). Girls are affected by father-absence as well, although the effects on girls may be less enduring, dramatic, and consistent than the effects on boys (Lewis, 1997). Girls growing up without a father are at risk for early sexual activity and adolescent pregnancy (Ellis et al., 2003; McLanahan & Sandefur, 1994; Newcomer & Udry, 1987).

What we know about the effect of positive father involvement on children comes mostly from studies of middle-class, two-parent families. Positive father-child activities and interactions are consistently found to correlate with positive outcomes for children (Lamb, 2004; Lewis, 1997; Palkovitz, 2002; Shannon, Tamis-LeMonda, London, & Cabrera, 2002). It is assumed that involvement can influence children's outcomes both directly and indirectly (Lamb, 1997). For example, fathers' speech patterns directly affect their children's linguistic and cognitive abilities. Fathers who engage their children in challenging verbal exchanges promote linguistic aptitude (Perlmaan & Gleason, 1993; Tamis-LeMonda, Shannon, Cabrera, & Lamb, 2004). Overall, greater paternal involvement and investment is also associated with a range of other positive child outcomes, including cognitive development, academic achievement, self-esteem, emotional and behavioral regulation, and social competence (see Lamb, 1997, 2004; Marsiglio, Day, & Lamb, 2000; Palkovitz, 2002; Pleck, 1997; Radin, 1994). Although we do not understand the processes behind these associations, there is strong and consistent evidence suggesting that fathers can make a unique contribution to their children's development.

Few studies examine paternal involvement among nonresident and low-income fathers (Cabrera et al., 2000; Coley, 2001; Shannon et al., 2002; Cabrera & Garcia-Coll, 2004). Research on paternal involvement among nonresident fathers in particular tends to emphasize low rates of parent-child contact, particularly between fathers and older children. Using larger and smaller datasets with mother-report data, studies have found that approximately 40% of nonresident fathers visit their preschool age children weekly, a proportion that drops to 22% when children are school-aged (Coley & Chase-Lansdale, 1999; Lerman, 1993). Other studies report that between 30% and 43% of nonresident fathers have no contact with their older children and adolescents (Furstenberg, Morgan, & Allison, 1987; King, 1994). A more recent study reported higher rates of contact between nonresident fathers and children of all ages, with 60% of fathers reporting some contact with their children at least monthly (Cooksey & Craig, 1998). Still, most of this research uses simplistic variables to capture involvement, omitting more nuanced measures such as the nature and amount of activities fathers engage in with their children, fathers' responsibility in the parenting role, and the quality of father-child interactions (Coley & Chase-Lansdale, 1999).

Recent qualitative studies offer a more realistic view of poor men that goes beyond "the deadbeat dad" perception. For example, a comprehensive review of the qualitative data on poor African-American fathers shows that fathers' role in a family is subject to negotiations with other family members (Jarrett et al., 2002). If they cannot provide financially, many men offer their time and other in-kind support; if they cannot commit to a marital relationship, they opt for a nonmarital romantic or friendship-based relationship with the child's mother. Nelson and his colleagues' (2002) qualitative study of 40 African-American fathers revealed that for many men having children greatly improved their lives. Although many had good intentions, they often felt they had failed as fathers because of unemployment, personal problems, and a negative relationship with the child's mother. This personal sense of failure often resulted in less involvement with their children (Nelson et al., 2002). Findings from quantitative studies also suggest that the involvement of these non-custodial fathers has a positive effect on their children's cognitive development (Black et al., 1999; Shannon et al., 2002).

Other recent findings using national quantitative datasets suggest that the majority of fathers from low-income backgrounds stay in contact with their children, even if they do not reside with them (Mincy & Oliver, 2003), are positively involved in their partners' pregnancy, are present at the birth of their children, and are committed to helping raise their children (Carlson & McLanahan, 2002b). These findings are important steps toward understanding the lives and influence of low-income fathers, but they also lead to other questions, such as how low-income men engage with their young children. This study builds on this research by using a multidimensional model of father involvement to examine how low-income men interact with their children in a large and diverse sample of low-income families.

A noteworthy challenge to understanding research on father involvement is the lack of theory linking specific aspects of father involvement to child developmental outcomes (Tamis-LeMonda & Cabrera, 2002), While "positive" father involvement is linked with positive outcomes for children, it is unclear what "positive" means in a diverse economic and cultural/ethnic context (Marsiglio et al., 2000). In lieu of a grand theory, the field has relied on different models of fatherhood that conceptualize father involvement as multidimensional, including behaviors that extend beyond the role of provider (Cabrera, in press; Palkovitz, 2002). These substantial theoretical issues are outside the scope of this article. However, to frame our findings we use Lamb, Pleck, Charnov, & Levine's (1987) father involvement model. This model provides a flexible and common understanding of the types of behaviors that constitute father involvement. Moreover, it was used in the conceptualization and design of several cutting-edge studies, including the EHS study (Cabrera et al., 2002). Lamb et al. (1987) propose the following three types of father involvement: (1) accessibility--a father's presence and availability to the child, irrespective of the nature or extent of interactions between them; (2) responsibility--a father's understanding and meeting of his child's needs, including the provision of economic resources to the child; and (3) engagement--a father's experience of direct contact, caregiving, and shared interactions with his child.

The EHS evaluation includes a study of the implementation and impacts of the EHS program and a longitudinal study of infants and toddlers in low-income families. EHS is a comprehensive, two-generation program that provides intensive services from before birth until age three years. The EHS study includes approximately 3,000 families living in 17 communities across the country who met the EHS income eligibility criteria at enrollment and had a child younger than 12 months of age. To determine eligibility based on income, programs use the poverty guideline, which is a version of the federal poverty measure. The guidelines are used to determine financial eligibility for certain federal programs (e.g., $14,494 for a family of three in 2002) (see Administration for Children and Families, 2002). For Early Head Start, income eligibility is determined at the time of enrollment and eligibility is maintained until children finish the program at age three.

As part of the experimental research design, families who applied to EHS were randomly assigned either to the EHS study group and received program services or to the control group and received other services available in their communities. Data on family demographics and service needs were collected prior to random assignment. The evaluation included measures of a broad range of child and family outcomes and obtained extensive information about the programs and the individual families' experiences with them. Families were interviewed at 7, 16, and 28 months after random assignment and at program exit. Child assessments, parent interviews, and parent-child interaction assessments were conducted when the children were 14, 24, and 36 months old.

The EHS families and the control group were enrolled first, and fathers were recruited after mothers identified them. (The primary caregiver in 95% of the cases was the child's mother.) Questions for the mother about father and father figure involvement with the child were included at baseline and in the parent interviews conducted when the children were 14, 24, and 36 months old. (1) In 1997, data collection was expanded in 12 of the 17 research sites to include interviews with fathers and, in eight of the 12 sites, to include father-child videotaped interactions. The result of these efforts was four strands of father research: (1) the Father Involvement with Toddlers Study (FITS), (2) the Father and Child Interaction During Toddlerhood Study (FACITS), (3) the Father and Newborn Study (FANS), and (4) the Participation in Fatherhood Programs and Services Use Study (PASS) (Cabrera et al., 2002). (See Administration for Children and Families [2002] report on overall program effects). The present study draws its data from the FITS strand of the EHS project.

This study seeks to address the following research question: What is the frequency and type of father involvement among fathers whose children participated in the EHS study? To address this question, we ask more specifically: (1) Are fathers whose children participated in the EHS study accessible to their children? (2) How do they engage with their children? (3) In what ways are they responsible for their children? And (4) how do these patterns of paternal involvement vary by mother-father relationship status and residency?

Some caveats are necessary. First, our study focuses only on biological fathers. We exclude other types of fathers, such as stepfathers and social fathers, because the patterns of investment in and relationship to children of these men can be different from those of biological fathers (Hetherington & Henderson, 1997; McLanahan & Sandefur, 1994). Moreover, most child and family public policies are directed toward biological fathers, particularly in the case of recent policy initiatives on family formation and child support. In the absence of research on low-income fathers, we run the risk of promoting policies based on too little or even misleading information. Policies based on untested assumptions may cause more social harm than good and may divert limited funds away from policies that enhance family functioning.

METHOD

PROCEDURES

This study is based on data collected on mothers and fathers participating in the Early Head Start Research and Evaluation project and the Father Involvement with Toddlers Study (FITS) data (from 12 of the 17 EHS research sites). In the 12 FITS sites, mothers who identified fathers were asked for written permission to contact them. If the biological father did not live with the mother and child, the mother was also asked about any man in the child's life who might be "like a father" to the child. Fathers were contacted by telephone, and home visits were scheduled at times convenient to the fathers (including late evenings and weekends). Written consent was obtained during the home visits. Father visits involved interviews (in 12 sites) and videotaped parent-child play interactions (in eight sites) when children were 24 and 36 months old. After the visits, parents were given $20 and small gifts for their participation.

PARTICIPANTS

The present study focused on the data collected from the biological mother and biological father interviews at 24 months. The sample includes mothers at 24 months (N = 1,550) who participated in the EHS Research and Evaluation Project and fathers who participated in the FITS (n = 597). (2) To present a demographic profile of the families participating in the EHS study, we use maternal reports of family demographic characteristics from the larger sample (N = 1,550) (see Table 1).

According to mothers participating in EHS, almost half of the biological fathers resided with their children (were married or cohabiting). About three quarters of resident fathers were married to their biological children's mothers, while 27% were cohabiters (not married but living with the biological child's mother). Of the nonresident fathers, 14% had a romantic relationship with their biological children's mother (boyfriend), 37% were just friends with their biological child's mother, and almost half had no relationship with the mother. Overall, 36% of parents in the sample were married, 14% cohabited, 7% were romantically involved but nonresident, 18% were just friends, and 25% had no relationship at all.

Families came from diverse racial/ethnic backgrounds. Among mothers, 47% were White, 25% Black, 21% Latino, and 7% other race [i.e., Asian, American Indian, or biracial]; among fathers, 46% were White, 26% Black, 25% Latino, and 4% other race [i.e., Asian, American Indian, or biracial]. Mothers were relatively young and poorly educated; 25% were adolescents (18 years or younger) at the time of their child's birth, and 42% had less than a high school education at that time. Fathers were older than mothers, and they were less likely to have graduated from high school. Most fathers (78%) reported working full-time or part-time. Overall, the families were low-income: 39% of mothers received AFDC at some point during the 15 months after random assignment, and only 38% lived above the poverty line at that time. Almost half of families were from urban neighborhoods. Children's ages averaged 25.1 months (SD = 1.42) at the time of the mother visit, and about half were boys.

Given the design of the study (mothers identified fathers, but not all identified fathers were willing to participate in the study), the fathers who ultimately participated in the EHS FITS study are a select group of men (see Table 1). Moreover, not all research sites participated in the Father Study, adding a site-level source of selection bias. Compared with those who did not participate in the FITS study, participating fathers and their children's mothers were more likely to be married and/or cohabiting, White or Latino, and living above the poverty level or less likely to be receiving welfare. Participating fathers were also more likely to be educated beyond high school and employed. Fathers' age, mothers' age and education, and children's gender, age, and first-born status were unrelated to fathers' participation.

DEMOGRAPHICS BASED ON PARENT RELATIONSHIP STATUS

Resident fathers (see Table 2). Analysis of data from the mothers' reports showed that married families were more likely to be living in rural areas and above the poverty level. Married men were older at the time of the child's birth, had higher levels of education, and were more likely to be employed than all unwed groups. Following national trends, married parents were more likely to be White or Latino than Black, and a lower proportion of mothers were primarily English speaking. Cohabiting fathers were slightly less advantaged than were married fathers in terms of age, education, and employment, while cohabiting mothers were more likely to receive welfare and less likely to live above the poverty line than married women. However, racially and ethnically, cohabiting men resembled married men in that they were most likely to be White or Latino and less likely to be Black than nonresident fathers.

Nonresident fathers. They more closely resembled cohabiting than married men; they were younger, less educated, and less employed than married fathers and similar to cohabiters on these variables. Among the three groups of nonresident fathers, boyfriends were youngest (22% were adolescents) and least likely to have high levels of education (11% had more than a high school education, versus 18% of friends and fathers with no relationship) but had higher rates of employment than fathers who were just friends (71% vs. 66%, respectively) and fathers who had no relationship at all with the mother (71% vs. 35%, respectively). However, their relative youth may explain their lower levels of education. Boyfriends and friends were more likely to be Black, and men with no relationship with the mother were most likely to be White.

MEASURES

Measures for this study include demographic characteristics and father involvement--accessibility, engagement, and responsibility--as conceptualized by Lamb et al.'s (1987) model. In addition to this model, we include an array of demographic variables that are derived from Paikovitz's (1997) conceptualization of father involvement. Demographic characteristics. We used mothers' reports to identify the mother-father relationship status as married, cohabiting, boyfriend (nonresident but romantically involved with the mother), friend (nonresident but divorced/separated), or no relationship (mother reports having no relationship, romantic or otherwise, with the biological father). We also used mothers' reports to collect data on the mothers' and fathers' ages, race/ethnicity, and education; AFDC status; mothers' and fathers' urban/rural residency; fathers' and mothers' employment status; and children's age, gender, and first-born status.

Father involvement. Using Lamb et al.'s (1987) model, three types of father involvement were assessed: accessibility, engagement, and responsibility. Accessibility was measured from mother reports because using data from father reports would underestimate the father's accessibility to the child. For resident fathers, we use mothers' reports of fathers' residency in the house as an index of fathers' general accessibility to the child; thus all resident fathers were considered accessible by definition. For nonresident fathers, we use frequency of father contact as a proxy for father accessibility to his children or presence in their lives. Mothers were asked, "In the last three months, how often has (child) seen his father?" Responses were rated using a five-point Likert scale ranging from 1 = never to 5 = every day or almost every day. If the mother said that the child saw his father at least once in three months, responses were coded yes. Responses were also coded once to a few times a month and a few times a week to every day, depending on the mothers' specific answer.

We used father reports for the constructs of father engagement and responsibility, rather than mother reports, because one of the goals of this paper is to report on fathers' activities from their own perspectives. In addition, we asked fathers more questions about their involvement with their children than we asked mothers. Furthermore, because mothers and fathers were asked slightly different questions about father involvement, we cannot assess concordance rates between the two groups' responses. We recognize the problems inherent in using self-report data. Although father-report data may better represent men's perspectives, fathers can be influenced by social desirability bias to present themselves as more involved with their children. It is methodologically challenging to establish the validity of either mother or father reports of father involvement, particularly when parents do not agree. In this study, we use father-report data in order to represent men's perceptions of their involvement with their children.

Engagement was measured by asking fathers: "In the past month, how often did you take care of your child while the child's mother did other things?" This item was rated on a scale from 1 = never to 5 = every day or almost every day. We recoded the item into a dichotomous variable with 1 = every day or almost every day and 0 = less than every day as values. We then asked, "In a typical day, how much help do you give in caring for your child?" This item was rated on a scale from 1 = none to 3 = a lot. This item was recoded into a dichotomous variable with 1 = a lot and 0 = less than a lot as values.

Finally, we asked a series of 30 questions about the activities the fathers did with their children in the past month. The questions included the following: "How often did you read to your child?" "How often did you take your child to visit relatives?" "How often did you help your child get dressed?" and "How often did you play chasing games with your child?" We rated these items on a scale from 1 = never to 6 = several times a day. We then grouped each of the 30 items into one of four meaningful engagement activities: caregiving (8 items), socialization (9 items), physical play (8 items), and didactic (5 items). All four scales demonstrated good internal consistency, with alphas ranging from .77 to .86 (see Appendix A).

Responsibility was measured by asking fathers the following questions: (1) "How much influence do you have in making major decisions about such things as education, religion, and health care for your child?" (2) "In the past month how often have you taken your child to the doctor?" and (3) "In the past month how often have you taken care of your child when s/he was ill?" Responses to the first question were rated using a three-point Likert scale (1 = no influence to 3 = a great deal of influence) and were then recoded into a dichotomous variable with 1 = a great deal of influence and 0 = less than a great deal of influence as values. The last two items were rated on a scale from 1 = never to 6 = more than once a day, so mean values are reported. Nonresident fathers were also asked who paid for the following items: the child's clothing, toys and diapers; medical expenses; and everyday household items, food, and babysitting services. Paid items were coded as 1 =father only paid, 2 = mother only paid, 3 = both mother and father paid, 4 = someone else paid. This item was recoded into a dichotomous variable with 1 =father paid and 0 =father did not pay as values.

RESULTS

To address the overarching question of how the fathers whose children are enrolled in the EHS study are involved with their children, we present our findings in terms of Lamb et al.'s (1987) model. First, using maternal reports on the full sample (N = 1,550), we examine fathers' accessibility as a group and then across the five categories of parent relationship status: (a) married; (b) cohabiters; (c) boyfriends; (d) nonresident fathers who were just friends with the mother (friends); and (e) nonresident fathers who had no relationship with the mother. We present these results in terms of the percentage of mothers reporting that biological fathers had contact with their children at least once in the previous three months, a few times a month, and a few times a week (Table 3).

We then examine fathers' reports (n = 597) of aspects of their engagement with and responsibility toward their children as a group and by relationship status.3 As indicated in Table 2, parent relationship status is associated with a variety of demographic measures. Resident parents are more advantaged than parents who live apart, and among resident parents married parents are more advantaged than cohabiters (demographic characteristics of the father sample by relationship group are reported in Appendix B; patterns by group are parallel to those for the full sample). We account for the potentially confounding effects of these associations by reporting percentages or means for each engagement and responsibility measure adjusted for relevant father and child demographic characteristics. Specifically, ANCOVA models were run for all engagement and responsibility measures with mean values of father race (Black, Hispanic, or other race), education (less than high school/GED, more than high school/GED), employment (employed, missing employment data), urbanicity and age at child's birth, and child age in months, gender, and first-born status entered as covariatas. Adjusted percentages or means (depending on the measure) are reported beneath unadjusted percentages or means in Table 4.

ARE FATHERS ACCESSIBLE TO THEIR CHILDREN? HOW DOES FATHER ACCESSIBILITY VARY BY RELATIONSHIP STATUS?

Approximately 80% of the 1,550 EHS mothers interviewed at 24 months reported that the biological father was present or accessible in their children's lives when we define accessibility as "seeing the child once in the past three months" (see Table 3). Of the nonresident fathers who were boyfriends of the mother, more than 95% saw their children at least this often, while nearly 90% of nonresident fathers who were just friends with the mother contacted their children at least once during the three-month period. And a notable 34% of nonresident fathers with no relationship at all with the mother still had some contact with their children during the three months. Overall, 61% of nonresident fathers had contact with their children at least once in three months.

The story is different when we define accessibility more conservatively as "contact with the child at least a few times a week." Using this definition, the percentage of nonresident fathers accessible to their children drops to 32% across types of nonresident father. This difference is attributable largely to lower rates of accessibility among fathers who had no relationship of any kind with their child's mother. Forty-eight percent of nonresident dads who were just friends with the mother contacted their children a few times a week, but only 6% of fathers with no relationship at all with the mother had weekly contact. Notably, more than 80% of nonresident boyfriends still see their children a few times a week.

In sum, rates of fathers' accessibility to their children vary by father's residence, relationship with the mother, and definition of accessibility. The interesting finding is that of the fathers who had no relationship at all with the child's mother, more than one-third still had some contact with their children, which is somewhat inconsistent with the literature suggesting that father involvement is mediated by the father-mother relationship (Lamb, 1997). Among nonresident men, however, rates of accessibility were much higher for fathers still romantically involved with the mother, suggesting that the father-mother relationship does impact father availability.

HOW DO FATHERS ENGAGE IN MULTIPLE ACTIVITIES WITH THEIR CHILDREN? HOW DO THEIR ENGAGEMENTS VARY BY RELATIONSHIP STATUS AND RESIDENCY?

Engagement was measured by asking fathers to report on the frequency ("How often did you take care of your child while child's mother did other things?"), the amount ("How much help do you give in caring for your child?"), and the activities they engaged in with their children (caregiving, socialization, physical play, and didactic). The fathers' perceptions of the amount of caregiving they provided varied with the question's specificity (see Table 4). Half the fathers reported looking after the child every day or almost every day, and 65% claimed they provided "a lot" of help. Only 6% said that they never looked after their child.

The amount of reported caregiving differed among relationship groups, although the differences were much larger for the frequency of looking after the child than for the amount of reported caregiving help provided. In terms of the unadjusted percentages, most married fathers and most cohabiting fathers reported looking after the child every day (frequency) (54% and 62%, respectively), as did a large percentage (49%) of boyfriends. Considerably smaller percentages of fathers who were just friends with the mother (24%) or who had no relationship with her (14%) reported this activity. When asked more generally, however, how much help (amount) they provided the mother in caregiving, more than half of all fathers and equal proportions of four groups of fathers reported providing "a lot of help." As expected, fathers who had no relationship with the mother were least likely to provide "a lot" of help.

When we adjusted the percentages of father-reported caregiving for father and child demographic characteristics, patterns by relationship group were largely unchanged. Married and cohabiting fathers looked after the child every day or almost in equal proportions, while friends and fathers with no relationship with the mother were the least likely to report this frequency. After controlling for demographic variation, however, fathers in different relationship groups reported providing "a lot" of help in relatively equal proportions.

In terms of fathers' engagement activities (i.e., caregiving, physical play, didactic, and socialization), on average, fathers reported spending a few times a week engaging in caregiving and physical play and a few times a month or more engaged in socialization and didactic play activities with their children (see Table 4). An examination of the frequency of involvement within these four types of engagement activities (see Appendix A) revealed that fathers reported spending most of their time trying to tease their children to get them to laugh, changing their children's diapers or helping them with the toilet, playing chasing games, and helping prepare meals or a bottle. Not surprisingly, the fathers reported spending the least amount of time taking their children to an activity in a community center, taking their child to the zoo or public event, taking their child to a religious event, or getting up in the middle of the night to provide care because the child had awakened.

Fathers' reported level of involvement in the four engagement activities differed based on the parent relationship status (see Table 4). According to the unadjusted means, married fathers and cohabiting fathers engaged in significantly more physical play and socialization activities than did fathers identified as friends or as having no relationship with the child's mother. Among nonresident fathers, boyfriends reported more physical play and socialization activities than fathers in no relationship with the mother. Surprisingly, cohabiting fathers reported engaging in more caregiving activities than did married men. In addition, married fathers and boyfriends reported engaging in more caregiving activities than did fathers who reportedly had no relationship with the mother. Cohabiting fathers also engaged in significantly more didactic activities than did fathers who were just friends or in no relationship with the child's mother.

Patterns of father engagement in activities were similar when we adjusted for father and child characteristics; however, some contrasts became nonsignificant. Demographic characteristics appear to account for the higher rate of reported physical play and socialization activities among boyfriends when compared to fathers in no relationship with the child's mother. Similarly, when demographics are held constant, cohabiting fathers report the same level of caregiving as married fathers and boyfriends, while both resident groups report higher levels of caregiving than friends and fathers in no relationship with the mother. Variation in the didactic scale is unchanged, with cohabiting fathers reporting higher levels of teaching activities than friends and fathers in no relationship with the child's mother.

HOW DO FATHERS SHOW RESPONSIBILITY FOR THEIR CHILDREN? HOW DOES THEIR RESPONSIBILITY VARY BY RELATIONSHIP STATUS AND RESIDENCY?

Responsibility was measured by asking fathers about the amount of influence they had in making major decisions concerning their child, the frequency with which they took their sick child to the doctor, and how often they took care of their sick child. Nonresident fathers also were asked about their economic contributions for the child's clothing, food, and gifts.

Fathers were asked to rate themselves on how much influence they had in making major decisions about education, religion, health care, and similar issues of importance in their child's life. As indicated in Table 4, the majority of fathers (71%) reported having a great deal of influence. Interestingly, married fathers reported having a great deal of influence over these decisions in greater proportions (81%) than did cohabiting fathers (66%), boyfriends (68%), fathers who were just friends (48%), and fathers in no relationship with the mother (35%). In turn, cohabiters reported having a great deal of influence in greater proportions than fathers who are friends or in no relationship with the mother but in equal proportions to boyfriends. These patterns were identical when percentages of reporting a great deal of influence were adjusted for demographic differences among the groups.

On average, fathers rarely took their ill children to the doctor or took care of them at home when their children were sick (M = 1.91 and 2.08, respectively). There were significant differences in how frequently fathers took their ill children to the doctor and took care of their ill children based on relationship status (Fs = 2.74 and 5.83, p < .05). Specifically, fathers who were married, cohabiting, or boyfriends were more likely to take their ill children to the doctor than fathers in no relationship with their child's mother. Also, cohabiting fathers reported taking care of their ill children significantly more often than fathers in no relationship with their child's mother. These differences were unaffected when father and child demographic characteristics were held constant.

Examination of nonresidential fathers' economic contributions to their children revealed that they reported most frequently paying for their children's clothing, toys, and childcare items (85%, 84%, and 83%, respectively) and less frequently paying for household items and baby-sitting (63% and 65%, respectively). Before adjusting for demographic variation, there were almost no significant differences in fathers' reports of financial support toward their children and families across the three nonresidential groups of fathers. The exception was that fathers who were boyfriends with their children's mother were more likely to pay for food than those in no relationship. Once demographic differences are controlled, however, boyfriends are more likely to contribute for household items than friends or fathers in no relationship with the mother and are even more likely to pay for food when compared to other nonresident groups (according to significance levels).

In summary, the majority of fathers, regardless of residence status, were accessible to their 24-month-old children. However, the majority of those who frequently saw their children had at least a romantic relationship with the child's mother. A closer examination of the fathers' engagements with and responsibility for their children showed that more than half of fathers, most of whom had at least a romantic relationship with the mother, reported looking after their child by themselves and providing a substantial amount of caregiving help. Fathers reported engaging in caregiving and physical play activities more frequently than engaging in socialization activities. Relative to nonresident fathers as a whole, both married fathers and cohabiting fathers perceived themselves as engaging in more activities with their children. In addition, marfled fathers perceived themselves as having much more influence over major decisions regarding their children than did cohabiting fathers and all nonresident fathers.

DISCUSSION

In this paper, we address the question of how low-income men are involved with their toddlers by examining father characteristics and involvement in a sample of families participating in the EHS study. As an improvement over previous methodologies, which have used only mothers' reports of father involvement, we used the men's own voices to understand the roles that these fathers report playing in their children's lives. Although these self-reports of father-child activities are not unbiased, they represent the men's perceptions of their involvement. Before discussing our findings, we emphasize that since our sample does not represent a random group of men, we cannot generalize findings from this study to larger low-income populations. However, this study provides meaningful information about how a diverse group of low-income men are involved with their children.

We find that the large majority of the fathers in the sample made themselves regularly accessible to their children, engaged with their children in positive ways, and assumed responsibility for their children's well-being--all of which is captured by Lamb et al.'s (1987) model of father involvement. Overall, the fathers who participated in our study had higher levels of education, had higher rates of employment, and were more likely to be married or living with the mother than were other low-income fathers in the EHS study. As a result, the study families appear to be more stable and might thus have better relationships with their children than low-income parents in general. Despite these limitations, the families do represent a diverse sample of low-income families, and their data can be used to challenge assumptions about low-income fathers and generate research hypotheses that can be tested with emerging datasets.

We find that the vast majority of fathers, regardless of residence status, are accessible to their children at 24 months, the age at which other studies have reported a drop-off in paternal involvement (Shannon et al., 2002). Furthermore, rates of father accessibility among nonresident fathers in our study exceed those reported in other studies of father involvement. However, there were significant differences in the fathers' involvement with their children when we examined the sample's relationship status and residency. In general, compared with their counterparts, married fathers and cohabiting fathers were more accessible to their children, were more engaged with them, and assumed more responsibility for them. Moreover, fathers who were romantically involved with the child's mother were more accessible to their children than nonresident fathers who were merely friends with the child's mother or had no relationship with her at all. Our findings suggest that nonresident fathers, who are more likely to drift away from their families, are accessible to their children, particularly if they still are romantically involved with the mother. This finding is consistent with a robust body of work that points to the quality of the marital relationship as an important mediator of positive parenting (Belsky, 1984, 1990; Cowan, Cowan, Heming, & Miller, 1994; Lamb, 1999) and the quality of parental relationships as a mediator of nonresident father involvement (Coley & Chase-Lansdale, 1999; Cooksey & Craig, 1998; Furstenberg & Harris, 1993).

This study has limitations. Since our sample is selective, these findings must be replicated with other datasets. More importantly, the issue of how fathers are involved with their children should be addressed using a nationally representative study, such as the ECLS-B. That study will provide a more definite answer to the question "How involved are poor, nonresident fathers in their children's lives, and what is the nature and impact of that involvement?"

The findings reported in this paper do not settle the debate about whether marriage causes better parent behavior, but they do support the notion that married parents are different from unmarried parents along important demographic dimensions that have been robustly linked to financial and relationship stability. Higher education, better employment, and lack of major health problems are general traits of married people compared to the unwed (Wilson & Brooks-Gunn, 2001). Married fathers in the EHS study, in turn, report that they have more influence over child-rearing decisions than other fathers. It is important to note, however, that although cohabiting fathers resemble nonresident fathers demographically and in terms of their perceived influence, they report being just as accessible to and engaged with their children as married fathers do.

Among nonresident fathers, rates of accessibility and engagement are lower but still moderately high for those fathers who continue to be romantically involved with the mother. Parents who have a romantic relationship with their partners tend to be more accessible and engaged with their children than those who are not romantically involved. This finding suggests that programs should consider targeting not only both parents, mothers and fathers, but also nonresident fathers, with permission of the custodial parent, regardless of the status of the romantic relationship to the child's mother.

This study makes several contributions to the literature on father involvement. Despite the effect that poverty may have on men's relationships, housing, physical health, and mental health, low-income fathers in this study are accessible, engaged, and assume responsibility for their children. Nonresident fathers in our sample are more involved with their children than other studies suggest. Fathers who have some type of relationship with the mother are more involved with their children across types of father involvement than those in no relationship. A significant proportion of fathers who have no relationship with the mother of their child had some contact with the child, however, suggesting that the relationship between mother and father is not the only factor helping men stay involved in their children's lives. Finally, accessible fathers report doing a lot more caregiving than has been suggested by other studies. These low-income men are engaged in activities that are considered to be mainly done by mothers. The consequences of this type of involvement for children merit future research. This study provides a steppingstone toward understanding the nature and frequency of father involvement in poor families.

Table 1

Parent and Child Characteristics at 24 Months

                                          Mother-Only
Demographics                      Full      Sample

Mother-Father Relationship (%)
  Married                         35.5       23.0
  Cohabiter                       14.1        9.9
  Boyfriend (nonresident)          7.2        6.5
  Friend                          17.9       21.6
  No Relationship                 25.4       39.0
Father Characteristics
Race (%)
  White                           45.6       44.8
  Black                           25.7       28.9
  Latino                          24.7       21.6
  Other                            4.0        4.8
Age at Child's Birth (M)          25.7       25.5
Teenager at Child's Birth (%)     10.9       11.3
Education Level (%)
  <High School                    62.9       71.0
  High School                     15.3       11.1
  >High School                    21.8       17.9
Working (%)                       78.0       73.6
Mother Characteristics
Race (%)
  White                           47.4       46.3
  Black                           24.9       28.3
  Latino                          20.8       18.0
  Other                            7.0        7.3
Age at Child's Birth (M)          22.8       22.6
Teenager at
Child's Birth (%)                 25.1       25.7
Education Level (%)
  <High School                    41.8       42.9
  High School                     30.5       30.5
  >High School                    27.7       26.5
Primary Language English (%)      81.7       83.7
Ever worked (%)                   78.4       78.1
AFDC (%)                          38.5       44.5
Above FPL (%)                     33.7       33.0
Urban (%)                         46.3       48.1
Child Characteristics
Male (%)                          50.6       49.8
First Born (%)                    60.2       60.2
Age in Months
(at 24mths) (M)                   25.1       25.1
N =                               1550       953

Demographics                      Father Sample    F/[chi square]

Mother-Father Relationship (%)                       344.40 ***
  Married                             55.4
  Cohabiter                           20.9
  Boyfriend (nonresident)              8.2
  Friend                              11.9
  No Relationship                      3.5
Father Characteristics
Race (%)                                              14.34 **
  White                               46.6
  Black                               22.1
  Latino                              28.3
  Other                                3.0
Age at Child's Birth (M)              25.9             1.14
Teenager at Child's Birth (%)         10.2             0.43
Education Level (%)                                   52.23 ***
  <High School                        52.3
  High School                         20.8
  >High School                        27.0
Working (%)                           83.4            18.10 ***
Mother Characteristics
Race (%)                                              21.64 ***
  White                               49.1
  Black                               19.4
  Latino                              25.1
  Other                                6.4
Age at Child's Birth (M)              23.2             1.95 *
Teenager at
Child's Birth (%)                     24.1              .46
Education Level (%)                                    2.20
  <High School                        39.8
  High School                         30.4
  >High School                        29.7
Primary Language English (%)          78.7             6.00 (+)
Ever worked (%)                       78.9              .12
AFDC (%)                              28.8            63.17 ***
Above FPL (%)                         44.6            17.68 ***
Urban (%)                             43.6             3.00 (+)
Child Characteristics
Male (%)                              51.9             0.64
First Born (%)                        60.3              .00
Age in Months
(at 24mths) (M)                       24.9             2.00 *
N =                                    597

Note. Pull sample: for father variables, Ns range from 1,283 for race
to 1,482 for age; for mother variables, Ns range from 1,283 for poverty
level to 1,550 (total N) for race. Father Sample: For father variables,
ns range from 590 for employment to 597 (total N) for race. Father
Sample: For father variables, ns range from 590 for employment to 597
(total n) for age; and for mother variable, ns range from 520 for
poverty level to 597 (total n) for race.

(+) p<.10. *p<.05. ** p<.01. ***p<.001.

Table 2

Parent and Child Characteristics by Relationship Status at 24
Months--Full Sample

                                           Resident
Demographics                    Married     Cohab.     Boyfriend

Father Characteristics
Race (%)
  White                          53.4        42.1        18.8
  Black                          11.3        27.2        55.2
  Latino                         30.9        28.2        22.9
  Other                           4.4        2.6          3.1
Age at Child's Birth (M          27.1        25.6        24.1
Teenage at Child's Birth (%)      3.9        12.2        21.8
Education Level (%)
  <High School                   50.3        65.7        73.6
  High School                    19.8        19.5        15.1
  >High School                   29.9        14.8        11.3
Working (%)                      92.5        77.2        71.2
Mother Characteristics
Race (%)
  White                          55.5        47.5        22.5
  Black                          10.7        21.5        47.7
  Latino                         26.4        23.3        21.6
  Other                           7.5        7.8          8.1
Age at Child's Birth (M)         24.2        22.1        20.8
Teenager at Child's Birth (%)    13.7        30.1        41.4
Education Level (%)
  <High School                   34.9        44.2        58.6
  High School                    30.8        32.1        27.0
  >High School                   34.3        23.7        14.4
Primary Language English (%)     70.7        82.2        87.4
Ever Worked (%)                  78.0        81.4        73.4
AFDC (%)                         19.1        36.5        52.3
Above FPL (%)                    44.2        32.0        16.2
Urban (%)                        28.9        59.4        66.7
Child Characteristics
Male (%)                         49.8        53.0        48.6
First Born (%)                   51.7        64.8        65.8
Age in Months (at 24mths) (M)    25.0        25.2        24.9
Total N=                         550         219         111

                                     Nonresident

                                                No
Demographics                                Relation-     [F/[chi
                                 Friend        ship       square]

Father Characteristics
Race (%)                                                  154.22 ***
  White                           36.9         48.9
  Black                           44.4         28.5
  Latino                          15.7         17.4
  Other                            3.0          5.2
Age at Child's Birth (M           24.4         25.0        11.44 ***
Teenage at Child's Birth (%)      16.5         13.3        52.20 ***
Education Level (%)                                        80.27 ***
  <High School                    69.4         75.7
  High School                     12.1          6.0
  >High School                    18.5         18.3
Working (%)                       66.1         35.1       571.53 ***
Mother Characteristics
Race (%)                                                  157.87 ***
  White                           36.1         50.9
  Black                           43.0         27.5
  Latino                          15.9         14.8
  Other                            5.1          6.9
Age at Child's Birth (M)          22.0         22.4        15.09 ***
Teenager at Child's Birth (%)     30.3         29.8        65.34 ***
Education Level (%)                                        34.32 ***
  <High School                    46.5         41.9
  High School                     28.6         31.5
  >High School                    24.9         26.6
Primary Language English (%)      91.6         88.3        74.90 ***
Ever Worked (%)                   79.6         78.0         2.64
AFDC (%)                          48.4         55.7       187.82 ***
Above FPL (%)                     25.3         21.1       100.58 ***
Urban (%)                         55.2         51.4       113.44 ***
Child Characteristics
Male (%)                          56.3         47.1         6.37
First Born (%)                    62.5         66.3        26.49 ***
Age in Months (at 24mths) (M)     25.1         25.0         1.37
Total N=                          277          393

Note. Percents represent % of total N (1,550), although Ns vary as
noted in Table 1.

*** p <.001.

Mother-Report of Biological Father Accessibility by Relationship Status
at 24 Months

                               Resident              Nonresident

                       Full    Married     Cohab.    Boyfriend

Contact w/ child in
  last 3 months (%)    80.2     100.0      100.0       95.5
Contact a few
  times a month (%)    72.3     100.0      100.0       91.0
Contact a few
  times a week (%)     64.4     100.0      100.0       82.9

                         Non-
                       resident

                                     No            [chi
                       Friend    Relationship     square]

Contact w/ child in
  last 3 months (%)     86.6         33.9        754.29***
Contact a few
  times a month (%)     70.4         16.1        952.40***
Contact a few
  times a week (%)      47.7         5.6         1105.39***

Note. Percents represent % of total N (1,550), although 61 cases have
missing data on contact with bio-lo&al father. Chi-square statistics
were computed including missing cases as a separate category.

*** p <001.

Table 4
Father-Report of Father Engagement and Responsibility by Relationship
Status

                                        Resident

                              Full     Married       Cohab.

Engagement
Looks after Child
Every Day/Almost
  Unadjusted Percentage (a)   50.3     53.9 (a)     61.6 (a)
  Adjusted Percentage (b)              56.7 (a)     59.2 (a)
Provides a Lot of
  Caregiving Help
  Unadjusted Percentage       64.6     62.0         73.2
  Adjusted Percentage                  65.0         71.3
Activities (c)
Physical Play
  Unadjusted Mean (a)         3.88      3.95 (a)     4.08 (a)
  Adjusted Mean (b)                     4.00 (a)     4.03 (a)
Caregiving
  Unadjusted Mean             3.94      3.93 (c)     4.32 (a)
  Adjusted Mean                         3.98 (a)     4.28 (a)
Didactic
  Unadjusted mean             3.43      3.47         3.61 (a)
  Adjusted Mean                         3.50         3.59 (a)
Socialization
  Unadjusted Mean             2.86      2.91 (a)     2.95 (a)
  Adjusted Mean                         2.94 (a)     2.91 (a)
Responsibility
A Lot of Influence in
  Major Decisions
  Unadjusted Percentage       71.0     81.2 (a)     65.6 (b)
  Adjusted Percentage                  80.9 (a)     66.4 (b)
Take Child to the Doctor
Unadjusted Mean               1.91      1.90 (a)     2.03 (a)
  Adjusted Mean                         1.93 (a)     2.00 (a)
Take Care of Ill Child
Unadjusted Mean               2.08      2.03         2.27 (a)
  Adjusted Mean                         2.09         2.21 (a)
Financially Contributes
  for the Following:
Clothing
  Unadjusted Percentage       85.3       --          --
  Adjusted Percentage                              84.7
Toys
  Unadjusted Percentage       84.2       --          --
  Adjusted Percentage
Medicine
  Unadjusted Percentage       71.8       --          --
  Adjusted Percentage
Household Items
  Unadjusted Percentage       63.1       --          --
  Adjusted Percentage

Childcare Items
  Unadjusted Percentage       83.3       --          --
  Adjusted Percentage
Food
  Unadjusted Percentage       72.5       --          --
  Adjusted Percentage
Baby-sitting
  Unadjusted Percentage       64.5       --          --
  Adjusted Percentage
Total N =                    597      331         129

                                   Nonresident

                              Boyfriend      Friend

Engagement
Looks after Child
Every Day/Almost
  Unadjusted Percentage (a)   49.0          24.3 (b)
  Adjusted Percentage (b)     43.0          21.8 (b)
Provides a Lot of
  Caregiving Help
  Unadjusted Percentage       71.4          61.4
  Adjusted Percentage         63.7          57.3
Activities (c)
Physical Play
  Unadjusted Mean (a)          3.83 (a,b)    3.45 (c)
  Adjusted Mean (b)            3.72          3.41 (b)
Caregiving
  Unadjusted Mean              3.99 (a,b)    3.57 (c)
  Adjusted Mean                3.89 (a,b)    3.49 (c)
Didactic
  Unadjusted mean              3.33          3.15 (b)
  Adjusted Mean                3.25          3.13 (b)
Socialization
  Unadjusted Mean              2.83 (a,b)    2.67 (c)
  Adjusted Mean                2.76          2.61 (b)
Responsibility
A Lot of Influence in
  Major Decisions
  Unadjusted Percentage       67.3 (a,b,c)   46.5 (c)
  Adjusted Percentage         67.6 (a,b,c)   46.2 (c)
Take Child to the Doctor
Unadjusted Mean                2.15 (a)       1.74
  Adjusted Mean                2.08 (a)       1.71
Take Care of Ill Child
Unadjusted Mean                2.26           2.03
  Adjusted Mean                2.16           1.96
Financially Contributes
  for the Following:
Clothing
  Unadjusted Percentage       83.3           90.2
  Adjusted Percentage         90.1           70.4
Toys
  Unadjusted Percentage       77.8           92.0
  Adjusted Percentage         81.4           90.8
Medicine
  Unadjusted Percentage       75.0           74.5
  Adjusted Percentage         79.8           73.3
Household Items
  Unadjusted Percentage       77.8           56.9
  Adjusted Percentage         84.4           54.0

Childcare Items
  Unadjusted Percentage       86.1           86.3
  Adjusted Percentage         88.4           86.1
Food
  Unadjusted Percentage       77.8 (a)       76.5
  Adjusted Percentage         84.8 (a)       73.8 (a)
Baby-sitting
  Unadjusted Percentage       75.0           63.0
  Adjusted Percentage         78.9           62.5
Total N =                     45             71

                                                 F

                                   No
                              Relationship

Engagement
Looks after Child
Every Day/Almost
  Unadjusted Percentage (a)     14.3 (b)     10.07 ***
  Adjusted Percentage (b)        7.2 (b)     11.89 ***
Provides a Lot of
  Caregiving Help
  Unadjusted Percentage         47.4          2.18 (+)
  Adjusted Percentage           43.4          1.88
Activities (c)
Physical Play
  Unadjusted Mean (a)            3.18        10.65 ***
  Adjusted Mean (b)              3.14 (b)    11.21 ***
Caregiving
  Unadjusted Mean                3.18 (c)    10.41 ***
  Adjusted Mean                  3.07 (c)    11.45 ***
Didactic
  Unadjusted mean                2.98 (b)     3.40 **
  Adjusted Mean                  2.91 (b)     3.98 **
Socialization
  Unadjusted Mean                2.36 (c)     5.86 ***
  Adjusted Mean                  2.38 (b)     7.10 ***
Responsibility
A Lot of Influence in
  Major Decisions
  Unadjusted Percentage         33.3         14.71 ***
  Adjusted Percentage           33.9         12.03 ***
Take Child to the Doctor
Unadjusted Mean                  1.33 (b)     5.83 ***
  Adjusted Mean                  1.28 (b)     5.51 ***
Take Care of Ill Child
Unadjusted Mean                  1.57 (b)     2.74 *
  Adjusted Mean                  1.44 (b)     2.57 *
Financially Contributes
  for the Following:
Clothing
  Unadjusted Percentage         73.3          1.40
  Adjusted Percentage            1.5
Toys
  Unadjusted Percentage         73.3          2.40 (+)
  Adjusted Percentage           68.8          2.09
Medicine
  Unadjusted Percentage         56.3          1.13
  Adjusted Percentage           49.4          2.23
Household Items
  Unadjusted Percentage         50.0          2.75 (+)
  Adjusted Percentage           44.2          4.99 **

Childcare Items
  Unadjusted Percentage         66.7          1.77
  Adjusted Percentage           61.8          2.51 (+)
Food
  Unadjusted Percentage         46.7 (b)      3.06 *
  Adjusted Percentage           38.8 (b)      5.56 **
Baby-sitting
  Unadjusted Percentage         46.7         1.85
  Adjusted Percentage           40.1         2.90 (+)
Total N =                       21

Note. Ns for engagement variables range from 563-596. Ns for
nonresidential fathers' financial contribu-tions for child and family
items range from 93 to 103.

(a) Unadjusted probabilities/means are derived from Anovas computed
without covariates; groups with different letter subscripts differ in
Bonferroni post-hoc comparisons at p < .05.

(b) Adjusted percentages/means are derived from Ancovas computed with
father race (Black, Latino, other), education (<HS, >HS), employment
(employed, missing employment data), urbanicity, age at child's birth,
and child gender, parity and age-in-months entered as as covariates
(mean values); groups with differ-ent letter subscripts differ in
Bonferroni post-hoc comparisons at p < .05.

(c) Engagement scales range from 1 to 6.

(+) p<.10. * p<.05. ** p<.Ol. *** p<.001.

Appendix A Descriptive Statistics for Engagement Items

Engagement Items                                                   M

Physical Play
  Play outside in the yard, a park or a playground with child     3.84
  Play chasing games with child                                   4.24
  Try to tease child to get child to laugh                        4.57
  Take child for a ride on your shoulders or back                 4.01
  Turn child upside down or toss him in the air                   3.62
  Roll a ball, toss a ball, or play games with a ball             4.12
  Go for a walk with child                                        3.54
  Bounce child on your knee                                       3.37
  Dance with child                                                3.62
Caregiving
  Put child to bed                                                4.22
  Give child a bath                                               3.55
  Help get child dressed                                          4.20
  Help change child's diaper or help with toilet                  4.55
  Help child brush teeth                                          3.71
  Prepare meals or bottles for child                              4.29
  Assist child with eating or give child a bottle                 4.07
  Get up with child when he/she wakes up during night             2.98
Didactic
  Sing nursery rhymes like "Jack and Jill" with child             3.08
  Sing songs with child                                           3.80
  Read stories to child                                           3.48
  Tell stories to child                                           3.14
  Play together with toys for building things                     3.66
Socialization
  Have relatives visit you and child                              3.30
  Take child with you to visit relatives                          3.32
  Take child shopping with you                                    3.47
  Take child with you to a religious service or religious
    event                                                         2.30
  Take child with you to an activity at a community center        1.81
  Take child to a restaurant or eat out with child                3.08
  Take child to a public place like a zoo or museum               2.05
  Visit friends with child                                        3.08
  Take child to play with other children                          3.36

Engagement Items                                                   SD

Physical Play
  Play outside in the yard, a park or a playground with child     1.21
  Play chasing games with child                                   1.31
  Try to tease child to get child to laugh                        1.42
  Take child for a ride on your shoulders or back                 1.36
  Turn child upside down or toss him in the air                   1.65
  Roll a ball, toss a ball, or play games with a ball             1.17
  Go for a walk with child                                        1.24
  Bounce child on your knee                                       1.61
  Dance with child                                                1.44
Caregiving
  Put child to bed                                                1.20
  Give child a bath                                               1.36
  Help get child dressed                                          1.14
  Help change child's diaper or help with toilet                  1.40
  Help child brush teeth                                          1.60
  Prepare meals or bottles for child                              1.34
  Assist child with eating or give child a bottle                 1.57
  Get up with child when he/she wakes up during night             1.53
Didactic
  Sing nursery rhymes like "Jack and Jill" with child             1.52
  Sing songs with child                                           1.45
  Read stories to child                                           1.39
  Tell stories to child                                           1.41
  Play together with toys for building things                     1.41
Socialization
  Have relatives visit you and child                              1.27
  Take child with you to visit relatives                          1.12
  Take child shopping with you                                    0.99
  Take child with you to a religious service or religious
    event                                                         1.12
  Take child with you to an activity at a community center        0.93
  Take child to a restaurant or eat out with child                0.85
  Take child to a public place like a zoo or museum               0.94
  Visit friends with child                                        1.24
  Take child to play with other children                          1.30

Engagement Items                                                  Range

Physical Play
  Play outside in the yard, a park or a playground with child      1-6
  Play chasing games with child                                    1-6
  Try to tease child to get child to laugh                         1-6
  Take child for a ride on your shoulders or back                  1-6
  Turn child upside down or toss him in the air                    1-6
  Roll a ball, toss a ball, or play games with a ball              1-6
  Go for a walk with child                                         1-6
  Bounce child on your knee                                        1-6
  Dance with child                                                 1-6
Caregiving
  Put child to bed                                                 1-6
  Give child a bath                                                1-6
  Help get child dressed                                           1-6
  Help change child's diaper or help with toilet                   1-6
  Help child brush teeth                                           1-6
  Prepare meals or bottles for child                               1-6
  Assist child with eating or give child a bottle                  1-6
  Get up with child when he/she wakes up during night              1-6
Didactic
  Sing nursery rhymes like "Jack and Jill" with child              1-6
  Sing songs with child                                            1-6
  Read stories to child                                            1-6
  Tell stories to child                                            1-6
  Play together with toys for building things                      1-6
Socialization
  Have relatives visit you and child                               1-6
  Take child with you to visit relatives                           1-6
  Take child shopping with you                                     1-6
  Take child with you to a religious service or religious
    event                                                          1-6
  Take child with you to an activity at a community center         1-6
  Take child to a restaurant or eat out with child                 1-6
  Take child to a public place like a zoo or museum                1-6
  Visit friends with child                                         1-6
  Take child to play with other children                           1-6

Appendix B

Parent and Child Characteristics by Relationship Status at 24
Months--Father Sample

                                 Resident         Non-Resident

Demographics                Married     Cohab.     Boyfriend

Father Characteristics
Race (%)
  White                       55.6       40.8         14.3
  Black                       11.2       28.0         51.0
  Latino                      29.6       29.6         30.6
  Other                        3.6        1.6          4.1
Age at Child's Birth (M)      27.0       25.0         24.0
Teenage at Child's
  Birth (%)                    3.9       14.4         24.5
Education Level (%)
  <High School                44.7       60.8         65.3
  High School                 19.6       22.4         24.5
  >High School                35.6       16.8         10.2
Working (%)                   91.8       80.0         73.5
Race (%)
  White                       56.5       48.0         20.4
  Black                       10.9       21.6         40.8
  Latino                      25.4       25.6         32.7
  Other                        7.3        4.8          6.1
Age at Child's Birth (M)      24.3       21.9         21.2
Teenage at Child's
  Birth (%)                   12.7       35.2         46.9
Education Level (%)
  <High School                34.2       46.7         59.6
  High School                 29.9       28.7         28.6
  >High School                35.8       24.6         12.2
Primary Language
  English (%)                 73.5       82.8         77.6
Ever Worked (%)               78.3       82.4         72.5
AFDC (%)                      19.3       31.2         51.0
Above FPL (%)                 45.3       41.6         20.4
Urban (%)                     29.3       60.8         77.6
Child Characteristics
Male (%)                      49.2       55.2         42.9
First Born (%)                54.7       68.0         65.3
Age in Months
  (at 24mths) (M)             24.9       25.3         25.0
Total N =                    331        129           45

                              Non-
                            Resident
                                             No
Demographics                 Friend     Relationship    F/[chi square]

Father Characteristics
Race (%)                                                  72.57 ***
  White                       38.0          42.9
  Black                       39.4          33.3
  Latino                      21.1          19.0
  Other                        1.4           4.8
Age at Child's Birth (M)      24.0          24.9           5.44 ***
Teenage at Child's
  Birth (%)                   21.1          14.3          37.13 ***
Education Level (%)                                       30.94 ***
  <High School                59.2          66.7
  High School                 21.1          19.0
  >High School                19.7          14.3
Working (%)                   81.7          57.1          63.93 ***
Race (%)                                                  57.70 ***
  White                       35.2          52.4
  Black                       38.0          28.6
  Latino                      22.5           9.5
  Other                        4.2           9.5
Age at Child's Birth (M)      21.6          22.9           7.95 ***
Teenage at Child's
  Birth (%)                   35.2          47.6          57.07 ***
Education Level (%)                                       22.76 **
  <High School                43.5          28.6
  High School                 31.9          47.6
  >High School                24.6          23.8
Primary Language
  English (%)                 91.4          95.2          16.74 **
Ever Worked (%)               81.7          73.7           2.43
AFDC (%)                      45.1          57.1          48.32 ***
Above FPL (%)                 23.9          14.3          38.87 ***
Urban (%)                     52.1          57.1          69.19 ***
Child Characteristics
Male (%)                      64.8          52.4           7.81 (+)
First Born (%)                69.0          61.9          10.25 *
Age in Months
  (at 24mths) (M)             25.0          24.9           1.80
Total N =                     71            21

Note. Total N = 597. Ns for demographic variables range from 597 (for
all child variables, father's race, mother's race, father's age,
mother's age, father's education, mother's education, and urbanicity)
to 508 (for mother's employment).

(+) p < .10. * p < .05. ** p < .01. *** p < .001.

The findings reported here are based on research conducted as part of the national Early Head Start Research and Evaluation Project funded by the Administration for Children and Families (ACF), U.S. Department of Health and Human Services, under contract 105-95-1936 to Mathematica Policy Research, Princeton, NJ, and Columbia University's National Center for Children and Families, Teachers College, in conjunction with the Early Head Start Research Consortium. The consortium consists of representatives from 17 programs participating in the evaluation, 15 local research teams, the evaluation contractors, and the Administration for Children, Youth, and Families (ACYF). Research institutions in the Consortium (and principal researchers) include ACF (Rachel Chazan Cohen, Judith Jerald, Esther Kresh, Helen Raikes, and Louisa Tarullo); Catholic University of America (Michaela Farber, Lynn Milgram Mayer, Harriet Liebow, Christine Sabatino, Nancy Taylor, Elizabeth Timberlake, and Shavaun Wall); Columbia University (Lisa Berlin, Christy Brady-Smith, Jeanne Brooks-Guan, and Alison Sidle Fuligni); Harvard University (Catherine Ayoub, Barbara Alexander Pan, and Catherine Snow); Iowa State University (Dee Draper, Gayle Luze, Susan McBride, and Carla Petorson); Mathematica Policy Research (Kimberly Boiler, Ellen Eliason Kisker, John M. Love, Diane Paulsell, Christine Ross, Peter Schochet, Cheri Vogel, and Welmoet van Kammen); Medical University of South Carolina (Richard Faldowski, Gui-Young Hong, and Susan Pickrel); Michigan State University (Hiram Fitzgerald, Tom Reischl, and Rachel Schiffman); New York University (Mark Spellman and Catherine Tamis-Lemonda); University of Arkansas (Robert Bradley, Mark Swanson, and Leanne Whiteside-Mansell); University of California, Los Angeles (Carollee Howes and Claire Hamilton); University of Colorado Health Sciences Center (Robert Emde, Jon Koffmacher, JoAnn Robinson, Paul Spicer, and Norman Watt); University of Kansas (Jane Atwater, Judith Carta, and Jean Ann Summers); University of Missouri-Columbia (Mark Fine, Jean Ispa, and Kathy Thornburg); University of Pittsburgh (Beth Green, Carol McAllister, and Robert McCall); University of Washington School of Education (Eduardo Armijo and Joseph Stowitschek); University of Washington School of Nursing (Kathryn Barnard and Susan Spieker), and Utah State University (Lisa Boyce and Lori Roggman).

The authors wish to thank the consortium reviewer (Hiram Fitzgerald), who commented on earlier drafts of this article under the guidelines of the Early Head Start Research Consortium publications policies. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government. We are also indebted to the fathers, mothers, and children who participated in this study.

NOTES

(1.) Father figures were men who were not biological fathers but whom the mother designated as "like a father" to the focus child. They included partners, friends, and family members such as uncles and grandfathers.

(2.) Although 1,568 biological mothers were interviewed for the EHS project at the FITS sites at 24 months, 28 were eliminated from these analyses either because the biological fathers were deceased by that time or because of missing data on relationship status. Similarly, 599 biological fathers were interviewed at those sites at 24 months, but two were eliminated because of missing data on relationship status data.

(3.) There were eight mothers with missing data on the mother-father relationship variable at 24 months. However, these data were recoded to match mother-father relationship status at 14 or 36 months and were retained in all analyses.

REFERENCES

Administration for Children and Families. (2002). Making a difference in the lives of infants and toddlers and their families: The impacts of Early Head Start. Washington, DC: U.S. Department of Health and Human Services. Retrieved January 19, 2004, from http://www.mathematica-mpr.com/PDFs/ehsfinalvol1.pdf

Amato, P.R., & Rivera, F. (1999). Paternal involvement and children's behavioral problems. Journal of Marriage and the Family, 61,375-384.

Belsky, J. (1984). The determinants of parenting: A process model. Child Development, 55, 83-96.

Belsky, J. (1990). Parental and nonparental care and children's socio-emotional development: A decade in review. Journal of Marriage and the Family, 52, 885-903.

Black, M.M., Dubowitz, H., & Starr, R.H. (1999). African American fathers in low-income, urban families: Development, behavior, and home environment of their three-year-olds. Child Development, 70, 967-978.

Brooks-Guan, J., Berlin, L.J., Leventhal, T., & Fuligni, A. (2000). Depending on the kindness of strangers: Current national data initiatives and developmental research. Child Development, 71, 257-267.

Cabrera, N. (in press). Father involvement: Models and theoretical frameworks. In C.B. Fisher & R.M. Learner (Eds.), Applied developmental science; Encyclopedia of research, policy, and program. Thousand Oaks, CA: Sage.

Cabrera, N., Brooks-Gunn, J., Moore, K., West, J., Boiler, K., & Tamis-LeMonda, C. (2002). Bridging research and policy: Including fathers of young children in national studies. In C. Tamis-LeMonda & N. Cabrera (Eds.), Handbook of father involvement: Multidisciplinary perspectives (pp. 489-524). Mahwah, NJ: Lawrence Erlbaum.

Cabrera, N., & Garcia-Coll, C. (2004). Latino fathers: Uncharted territory in need of much exploration. In M.E. Lamb (Ed.), The role of father in child development (4th ed., pp. 98-120). Hoboken, NJ: John Wiley & Sons.

Cabrera, N.J., Tamis-LeMonda, C.S., Bradley, R.H., Hofferth, S., & Lamb, M.E. (2000). Fatherhood in the twenty-first century. Child Development, 71, 127-136.

Carlson, M., & McLanahan, S. (2002a). Fragile families, father involvement and public policy. In C.S. Tamis-LeMonda & N. Cabrera (Eds.), Handbook of father involvement: Multidisciplinary perspectives (pp. 555-597). Mahwah, NJ: Lawrence Erlbaum Associates.

Carlson, M., & McLanahan, S. (2002b). Characteristics and antecedents of involvement by young, unmarried fathers. Center for Research on Child Wellbeing Working Paper, #02-09-FF.

Cherlin, A., & Griffith, J. (1998). Methodological issues in improving data on fathers. Nurturing Fatherhood Report (pp. 177-212), Washington, DC: Federal Interagency Forum on Child and Family Statistics.

Coley, R.L. (2001). (In)visible men: Emerging research on low-income, unmarried, and minority fathers. American Psychologist, 56, 743-753.

Coley, R.L., & Chase-Lansdale, P.L. (1999). Stability and change in paternal involvement among urban, African American fathers. Journal of Family Psychology, 13, 416-435.

Cooksey, E., & Craig, T.H. (1998). Parenting from a distance: The effects of paternal characteristics on contact between nonresidential fathers and their children. Demography, 35(2), 187-201.

Cowan, C.P., Cowan, P.A., Heming, G., & Miller, N.B. (1994). The transition from couple to family: Adaptation and distress in parents and children. In P.A. Cowan & E.M. Hetherington (Eds.), Family transitions (pp.79-109). Hillsdale, NJ: Erlbaum.

Duncan, G., & Brooks-Gunn, J. (Eds.). (1997). The consequences of growing up poor. New York: Russell Sage.

Ellis, B.J., Bates, J.E., Dodge, K.A., Fergusson, D.M., Horwood, L.J., Pettit, G.S., & Woodward, L. (2003). Does father absence place daughters at special risk for early sexual activity? Child Development, 74, 801-822.

Federal Interagency Forum on Child and Family Statistics. (1998). Nurturing Fatherhood: Improving Data and Research on Male Fertility, Family Formation and Fatherhood. Washington, DC.

Furstenberg, F.F., Jr., & Harris, K.J. (1993). When and why fathers matter: Impacts of father involvement on the children of adolescent mothers. In R.I. Lerman & T.J. Ooms (Eds.), Young unwed fathers: Changing roles and emerging policies (pp.117-138). Philadelphia: Temple University Press.

Furstenberg, F.F., Morgan, S.P., & Allison, P.A. (1987). Paternal participation and children's well-being after marital dissolution. American Sociological Review, 52, 695-701.

Garfinkel, I., McLanahan, S., Meyer, D.R., & Seltzer, J. (1998). (Eds.), Fathers under fire. New York: Russell Sage Foundation.

Garfinkel, I., McLanahan, S., & Robins, P.K. (1994). (Eds.), Child support and child well-being. Washington: Urban Institute Press.

Hetherington, E.M. (1991). The role of individual differences and family relationships in children's coping with divorce and remarriage. In P.A. Cowan & E.M. Hetherington (Eds.), Family transitions (pp. 165-194). Hillsdale, N J: Erlbaum.

Hetherington, E., & Stanley-Hagan, M. (1995). Parenting in divorced and remarried families. In M. Bornstein (Ed.), Handbook of parenting; Status and social conditions of parenting (Vol. 3). Mahwah, NJ: Lawrence Erlbaum Associates.

Hetherington, E.M., & Henderson, S.H. (1997). Fathers in stepfamilies. In M.E. Lamb (Ed.), The role of the father in child development (pp. 212-226). New York: John Wiley & Sons.

Jarrett, R.L., Roy, K.M., & Burton, L. (2002). Fathers in the "hood": Insights from qualitative research on low-income, African American men. In C.S. Tamis-LeMonda & N. Cabrera (Eds.), Handbook of father involvement: Multidisciplinary perspectives (pp. 211-248). Hillsdale, NJ: Lawrence Erlbaum Associates.

Johnson, W.E. (2001). Paternal involvement among unwed fathers. Children & Youth Services Review, 23, 513-536.

King, V. (1994). Nonresident father involvement and child well-being: Can dads make a difference? Journal of Family Issues, 15, 78-96.

Lamb, M.E. (Ext.). (1997). The role of the father in child development (3rd ed.). New York: John Wiley & Sons.

Lamb, M.E. (Ed.). (1999). Parenting and child development in "'nontraditional" families. Mahwah, NJ: Lawrence Erlbaum.

Lamb, M.E. (Ed.). (2004). The role of the father in child development (4th ed.). Hoboken, NJ: John Wiley & Sons.

Lamb, M.E., Pleck, J.H., Charnov, E.L., & Levine, J.A. (1987). A biosocial perspective on paternal care and involvement. In J.B. Lancaster, J. Altmarm, A. Rossi, & L. Sherrod (Eds.), Parenting across the lifespan: Biosocial perspectives (pp. 111-142). Chicago: Aldine.

Lerman, R.I. (1993). A national profile of young unwed fathers. In R.I. Lerman & T.J. Ooms (Eds.), Young unwed fathers: Changing roles and emerging policies (pp. 27-51). Philadelphia: Temple University Press.

Lewis, C. (1997). Fathers and preschoolers. In M.E. Lamb (Ed.), The role of the father in child development (3rd ed., pp. 121-142). New York: John Wiley & Sons.

Manning, W.D., Stewart, S.D., & Smock, P.J. (2003). The complexity of fathers' parenting responsibilities and involvement with nonresident children. Journal of Family Issues, 24, 645-667.

Marsiglio, W., Day, R.E., & Lamb, M.E. (2000). Exploring fatherhood diversity: Implications for conceptualizing father involvement. Marriage and Family Review, 29(4), 269-294.

McLanahan, S., & Sandefur, G. (1994). Growing up with a single parent: What hurts, what helps. Cambridge: Harvard University Press.

Mincy, R.B., & Oliver, H. (2003). Age, race, and children's living arrangements: Implications for TANF reauthorization (Policy Brief B-53). Washington, DC: Urban Institute Publications.

Newcomer, S., & Udry, R.J. (1987). Parental marital status effects on adolescent sexual behavior. Journal of Marriage and the Family, 49, 235-240.

Nelson, T., Clampet-Lundquist, S., & Edin, K. (2002). Sustaining fragile fatherhood: Father involvement among low-income, noncustodial African-American fathers in Philadelphia. In C.S. Tamis-LeMonda & N. Cabrera (Eds.), Handbook of father involvement: Multidisciplinary perspectives (pp. 525-554). Hillsdale, NJ: Lawrence Erlbaum Associates.

Palkovitz, R. (1997). Reconstructing "involvement": Expanding conceptualizations of men's caring in contemporary families. In A.J. Hawkins & D.C. Dollahite (Eds.), Generative fathering: Beyond deficit perspectives (pp. 200-216). Thousand Oaks, CA: Sage.

Palkovitz, R. (2002). Involved fathering and child development: Advancing our understanding of good fathering. In C.S. Tamis-LeMonda & N. Cabrera (Eds.), Handbook of father involvement: Multidisciplinary perspectives (pp. 119-140). Hillsdale, NJ: Lawrence Erlbaum Associates.

Perlmann, R.Y., & Gleason, J.B. (1993). The neglected role of fathers in children's communicative development. Seminars in Speech and Language, 14(4), 314324.

Peters, E., Argys, L., Brooks-Gunn, J., & Smith, J. (1998). The impact of child support on cognitive outcomes of young children. Demography, 35(2), 159-173.

Pleck, J.H. (1997). Paternal involvement: Levels, origins, and consequences. In M.E. Lamb (Ed.), The role of the father in child development (3rd ed., pp. 66103). New York: Wiley.

Radin, N. (1994). Primary caregiving fathers in intact families. In A. Gottfried & A. Gottfried (Eds.), Redefining families: Implications for children's development (pp. 173-204). Hillsdale, NJ: Lawrence Erlbaum Associates.

Shannon, J.D., Tamis-LeMonda, C.S., London, K., & Cabrera, N. (2002). Beyond rough and tumble: Low-income fathers' interactions and children's cognitive development at 24 months. Parenting: Science and Practice, 2(2), 77-104.

Tamis-LeMonda, C.S., & Cabrera, N. (2002). (EAs.) Handbook of father involvement: Multidisciplinary perspectives. Hillsdale, N J: Lawrence Erlbaum Associates.

Tamis-LeMonda, C.S., Shannon, J.D., Cabrera, N., & Lamb, M.E. (2004). Fathers and mothers at play with their 2- and 3-year-olds: Contributions to language and cognitive development. New York: New York University Press.

U.S. Department of Health and Human Services, Nurturing fatherhood: Improving data and research on male fertility, family formation, and fatherhood. (2000). Washington, DC: U.S. Government Printing Office.

Western, B., & McLanahan, S. (2000, June). Fathers behind bars: The impact of incarceration on family formation. Contemporary Perspectives in Family Research, 2, 309-324.

Wilson, M., & Brooks-Gunn, J. (2001). Health status and behaviors of unwed fathers. Children and Youth Services Review, 23(4-5), 377-410.

Wu, L.L., Bumpass, L.L., & Musick, K. (2001). Historical and life course trajectories of nonmarital childbearing. In L.L. Wu & B. Wolfe (Eds.), Out of wedlock: Trends and consequences of nonmarital fertility (pp. 1-48). New York: Russell Sage Foundation.

NATASHA J. CABRERA

University of Maryland

JACQUELINE D. SHANNON

New York University

CHERI VOGEL

Mathematica Policy Research, Inc.

CATHERINE TAMIS-LEMONDA

New York University

REBECCA M. RYAN

Columbia University

JEANNE BROOKS-GUNN

Columbia University

HELEN RAIKES

Society for Research in Child Development, Consultant Administration for Children and Families

RACHEL COHEN

Administration for Children and Families

Correspondence concerning this article should be addressed to Natasha J. Cabrera, University of Maryland at College Park, 3304 Benjamin Building, Room 117E, College Park, MD 20742. Electronic mail: [e-mail].