Nowadays, with the high rates of non-marriage, separation, and divorce, many children spend part of their childhoods living apart from at least one of their biological parents (Zill, 1996). Extensive research has been conducted on the effects of divorce for children's well-being (Kelly, 1993; Furstenberg and Cherlin, 1991; Wallerstein, 1991; Chase-Lansdale and Hetherington, 1990; Hetherington, 1981, 1979) and the problems experienced by children growing up in single-parent families (McLanahan and Sandefur, 1994). Such research has found that children are better off financially, psychologically, and emotionally when they are raised by two parents. However, reviewers of the research also note that the effects of divorce should not be overstated. The majority of children whose families are disrupted by divorce show no adverse signs several years later. For a small proportion of children, however, the consequences may be longer lasting. Researchers have found effects of marital disruption 12 to 22 years later in such outcomes as poor relationships with parents, increased levels of problem behavior, increased likelihood of dropping out of school and receiving psychological help, and lower likelihood of attending college (McLanahan and Sandefur, 1994; Zill, Morrison, and Coiro, 1993). Although the overall likelihood of these outcomes was relatively small, for some outcomes, such as dropping out of school, the risk was doubled.
It is not uncommon following the breakup of a family for one of the parents to become increasingly detached, paying little or no child support and visiting only infrequently, if at all (Seltzer, 1991; Furstenberg and Nord, 1985; Furstenberg et al., 1983). Because mothers are more likely than fathers to retain custody of the children in the event of divorce, most of the extant research has focused on noncustodial fathers and the consequences for children of living apart from their fathers (McLanahan and Sandefur, 1994; Amato, 1993). Existing research is mixed about whether the continuing involvement of nonresident fathers is important to children's lives. Several large-scale studies have found no association between the amount of contact a non-custodial father has with his children and an assortment of measures of child well-being (King, 1994; Furstenberg, Morgan, and Allison, 1987). Other studies, however, have found continued contact to be related to improved psychological scores, fewer behavioral problems, and better peer relationships (Peterson and Zill, 1986; Wallerstein and Kelly, 1980). Most of the studies that examine the influence of paternal contact on children examine the amount of contact that nonresident fathers have with their children. However, a simple count of days may not accurately reflect the role that fathers play in their children's lives. There is a need to consider other aspects of fathers' presence in children's lives, such as the quality of the relationship between fathers and their children, the types of activities that fathers share with their children, whether the visits are forced upon or welcomed by the children, whether the children's lives are disrupted by the contact (e.g., by having to travel long distances away from friends and family in order to visit their fathers), and whether the contact reflects the continuing presence of committed and involved fathers in their children's lives.
In this section, information on the involvement of nonresident fathers in their children's schools is presented and discussed. Some data on the involvement of nonresident mothers are also presented as a contrast to that of fathers. This section also examines factors that are associated with nonresident fathers' involvement and whether such involvement is linked to selected student outcomes. According to data from the NHES:96, approximately 16.8 million (34 percent) children in kindergarten through 12th grade have fathers who live apart from them. A much smaller number, 4.1 million (8 percent), have mothers who live outside the home.
According to the reports of the custodial parents, approximately one-quarter of the nonresident fathers had not had contact with their children in kindergarten through 12th grade in more than a year. Similarly, 10 percent of nonresident mothers had had no contact with their children in more than a year. These figures indicate, however, that three-quarters of nonresident fathers and 90 percent of nonresident mothers did have contact with their school-aged children in the previous year. The finding that nonresident mothers are more likely than nonresident fathers to have had at least some contact with their children in the previous year is consistent with previous research (Nord and Zill, 1996).
The fact that 75 percent of the students in the NHES:96 have had contact with their nonresident fathers in the past year is notable. In the early 1980s, it was estimated that just over half of children ages 6 to 17 years with nonresident fathers had had contact with their fathers in the past year (Furstenberg et al., 1983). School-aged children in 1996 were more likely to have at least some contact with their nonresident fathers than were children 15 years ago.
Custodial parents who reported that the nonresident parent has had contact with their children in the past year reported on the involvement of the nonresident parent in four school activities: attended a general meeting, attended a regularly scheduled parent-teacher conference, attended a class or school event, or volunteered at the school.
The most common activity of nonresident fathers is attending a school or class event (figure 15 ). Twenty-two percent of nonresident fathers who had seen their children in the last year attended at least one such event according to the reports of custodial mothers or the children's guardians. This compares with 53 percent of fathers in two-parent families. Approximately 18 percent of nonresident fathers attended a general school meeting and 15 percent attended a parent-teacher conference. In contrast, 55 percent of fathers in two-parent families attended a general school meeting since the beginning of the school year and 39 percent attended a parent-teacher conference. Clearly, involvement in schools by nonresident fathers is substantially lower than that of fathers in two-parent families; however, the proportion of nonresident fathers participating in school activities is by no means trivial.
Nonresident mothers are more likely than nonresident fathers to attend a class or school event, a parent-teacher conference, or a general school meeting. Just under one-third of nonresident mothers attended each of these events according to the reports of the custodial parent or guardian (figure 15). However, as with nonresident fathers, their participation in school activities is substantially lower than that of resident mothers.
A large proportion of nonresident parents have not participated in any of the school activities, even though they have seen their children in the previous year. Approximately 69 percent of the nonresident fathers and 56 percent of the nonresident mothers had participated in none of the activities (figure 16). In contrast, only 25 percent of resident fathers and 8 percent of resident mothers in two-parent families had not participated in any of the school activities since the beginning of the school year. One explanation may be that some of these parents do not live nearby and, thus, find it difficult to participate. According to data from the 1990 Survey of Income and Program Participation conducted by the U.S. Bureau of the Census, 38 percent of nonresident parents live in the same city or county as their children (Nord and Zill, 1996). This same study found that contact decreased substantially as parents moved away from the city or county in which their children lived. Such information was not collected in the NHES:96, so it is not possible to determine the extent to which distance is interfering with the involvement of nonresident parents in their children's schools.
In spite of the large proportion of nonresident parents with no involvement in their children's schools, 31 percent of nonresident fathers and 44 percent of nonresident mothers who have had contact with their children in the past year have attended at least one of the four activities. Eighteen percent of
nonresident fathers and 31 percent of nonresident mothers have participated in at least two of the four activities. And 9 percent of nonresident fathers and 20 percent of nonresident mothers have attended at least three of the school activities.
Several factors have been found to be associated with fathers' continued contact with their children following the disruption of families. These factors include whether fathers pay child support, the custodial mothers' education and household income, and ages of the children (Nord and Zill, 1996; Furstenberg and Cherlin, 1991; Seltzer, Schaeffer, and Charng, 1989). Previous studies have found that children who are younger, whose fathers pay child support, who have well-educated mothers, and who have higher family incomes are more likely to have seen their fathers within the past year than are other children. The results based on the NHES:96 are similar. Children are more likely to have had some contact with their fathers in the past year if their fathers have paid some child support, if the custodial mothers are more educated, and if their families are not experiencing economic difficulties (figure 17). There were no differences in NHES:96 in the proportion who had seen the children in the previous year by the grade level of the children.
These factors were entered together in multivariate models to determine whether their influence remained important after controlling for the other factors in the models (table 8). Three separate models are shown for nonresident fathers and for nonresident mothers. These models show the factors that are associated with the odds that nonresident parents have had contact in the past year with their children in grades K-12 (model 1), grades K-5 (model 2), and grades 6-12 (model 3). The results remain generally the same for children at all grade levels. The strongest influence on whether nonresident fathers have had contact with their children in the past year is whether they have paid any child support. Mothers' education and household income are also positively associated with continued father contact, as others studies have found. The one exception is among children in kindergarten through 5th grade. For these children, household income has no influence on whether their fathers have had contact with them or not in the last year. Children's grade level also has no significant influence on whether fathers have had contact with them. However, fathers are more likely to have had contact with their children if the mothers have not remarried.
These same factors are also associated with nonresident fathers being highly involved in their children's schools (figure 18). Nonresident fathers who have paid any child support are more likely than those who have paid none to be highly involved in their children's schools (10 percent versus 7 percent). High involvement by the nonresident fathers also tends to increase as the custodial mothers' education increases and if the custodial mothers are not experiencing economic difficulties. As with custodial parents, there is a tendency for nonresident fathers to decrease their involvement in their children's schools as the children move from elementary to middle to high school.
These factors were entered together in multivariate models to determine whether they remained important influences on nonresident fathers' involvement in their children's schools after controlling for the other factors in the models. Because only a small proportion of nonresident fathers are highly involved in their children's schools, for the multivariate models the dependent variable used was whether the nonresident fathers were moderately to highly involved in their children's schools, that is, whether they had attended two or more of the school activities since the beginning of the school year. In addition to the above factors, one set of models included the resident mothers' level of involvement in the children's schools.
Even after controlling for the other factors in the model, the tendency for nonresident fathers to become less involved in their children's schools as the children grow older is evident. The adjusted odds that children's fathers are moderately to highly involved in their schools are 36 percent less if the children are in grades 6 through 8 and 46 percent less if they are in grades 9 through 12 compared to if they are in kindergarten through 5th grade (table 9). Household income, mothers' education, whether the children live in single-mother families or in stepfather families, mothers' involvement in their schools, and whether the nonresident fathers have paid any child support are all significant influences on nonresident fathers' involvement among children in kindergarten through 12th grade. The specific factors that are important influences on nonresident fathers' involvement, however, differ somewhat by the grade level of the children.
Children in kindergarten-5th grade. Child support remains an important influence on nonresident fathers' involvement in their children's schools among children in kindergarten through 5th grade, though it loses some of its influence when mothers' involvement in school is added to the model. Moreover, the influence of child support on nonresident fathers' involvement in their children's schools is considerably smaller than it was on nonresident fathers' contact with their children. Other important influences on nonresident fathers' involvement among children in these grades are mothers' education and mothers' involvement in the children's schools. The likelihood that nonresident fathers are moderately to highly involved in their children's schools is also higher if the mothers have not remarried (significant at the 0.10 level). Household income has no influence on whether nonresident fathers are involved in the children's schools.
Children in 6th-12th grade. Whether nonresident fathers pay child support has no influence on their level of involvement in the schools of 6th through 12th graders, nor does mother's education. However, nonresident fathers are more likely to be moderately to highly involved in their children's schools as the resident families' household incomes and mothers' involvement in the children's schools increase. The strongest influence on nonresident fathers' involvement in their 6th through 12 graders' schools, however, is mother's marital status. Nonresident fathers are much more likely to be moderately to highly involved in their children's schools if the mothers are single than if there is a stepfather present. It is not clear whether the reason for the association is due to the time elapsed since the mothers and fathers separated or to the presence of a stepfather.
In this section, the influence of nonresident fathers' involvement in their children's schools on five student outcomes is examined. All children with nonresident fathers are included so that contrasts can be made between children with and without any contact with their nonresident fathers and among those whose fathers show different levels of involvement in their schools. Results are presented for children in grades 1 through 12 and in grades 1 through 5 and grades 6 through 12 so that differences in the influence of nonresident fathers involvement on student outcomes can be examined for different grade levels. As described below, there is an association between nonresident fathers' involvement in their children's schools and all five outcomes, though the influence is stronger for some outcomes than for others and is more apparent among children in grades 6 through 12 than among children in grades 1 through 5.
Get mostly A's. As table 10 indicates, nonresident fathers' involvement in their children's schools is associated with increased odds that children in grades 1 through 12 get mostly A's (models 1 and 2). This influence is reduced somewhat when the resident mothers' involvement in school is included in the model (model 3). When looking at the association by grade level of the children, nonresident fathers' involvement is not associated with whether children in grades 1 through 5 get mostly A's. However, nonresident fathers' involvement in their children's schools is associated with increased odds that 6th through 12th graders get mostly A's. The adjusted odds that 6th through 12th graders gets mostly A's increase by 42 percent if their fathers have attended one school activity and by 54 percent if their fathers have attended at least two school activities compared to children who have had contact with their nonresident fathers but whose fathers have not participated in any of the school activities (model 1). These results pertain to all 6th through 12th graders with a nonresident father. Among such children who are living with their mothers (whether in a stepfamily or in a single-parent family), the adjusted odds that they get mostly A's are also greater if their nonresident fathers have participated in school activities compared to if they have not (model 2). However, once mothers' involvement in school is added to the model, the influence of nonresident fathers' involvement is reduced (model 3).
Enjoy school. Children in grades 1 through 12 are more likely to enjoy school if their nonresident fathers are moderately to highly involved in their schools than if their fathers have contact with them but do not participate in any of the activities. Children also appear to enjoy school more even if their fathers participate in only one activity than if they participate in none (significant at the 0.10 level). The relationship between nonresident fathers' involvement in their children's schools and children's enjoyment of school is weaker when the sample is restricted to children who live with their mothers. However, there is still some evidence that children are more likely to enjoy school if their fathers are moderately to highly involved in their schools (significant at the 0.10 level). When looking at the association by children's grade level, there is no association between nonresident fathers' involvement in their children's schools and the odds that children in grades 1 through 5 enjoy school. However, the odds that children in grades 6 through 12 enjoy school are greater among children whose fathers are moderately to highly involved in their schools compared to children who have had contact with their fathers, but whose fathers have participated in none of the activities. This association weakens when mothers' involvement is included in the model.
Participate in extracurricular activities. The information on participation in extracurricular activities for children in grades 1 through 5 was obtained from parents' reports. Parents were not asked about their 6th through 12th graders' participation in extracurricular activities. The information on such participation among the older children was obtained from the youth themselves. The results in table 10 indicate that nonresident fathers' involvement in their children's schools is associated with a greater odds that the children are involved in extracurricular activities, particularly among children in grades 6 through 12. Among older children, the association persists even after mothers' involvement in the schools is included in the model.
These models also indicate that children who have had no contact with their fathers in more than a year are more likely to be involved in extracurricular activities than children who have seen their fathers in the past year but whose fathers participated in none of the school activities. Part of the explanation for this pattern may be that children are spending time with their nonresident fathers instead of participating in extracurricular activities. However, it is not possible to determine with the NHES data whether this supposition is correct or not.
Ever repeated a grade. Nonresident fathers' involvement in their children's schools reduces the odds that children in grades 1 through 12 have ever repeated a grade, even after controlling for mothers' level of involvement and the other factors in the model. The odds that children in grades 1 through 12 have ever repeated a grade are 39 percent less if their nonresident fathers have participated in one activity at school and 48 percent less if their nonresident fathers have participated in at least two activities at their schools, as compared to their fathers' simply having had contact with them but not participating in any of the school activities (model 1). The pattern is similar when the sample is restricted to children living with their mothers (models 2 and 3). There is a weak association between the odds that children in grades 1 through 5 have ever repeated a grade and the involvement of their nonresident fathers (significant at the 0.10 level). Children in grades 6 through 12, however, are significantly less likely to have ever repeated a grade if their nonresident fathers are involved in their schools.
Ever suspended or expelled. The question on whether the children have ever been suspended or expelled from school was only asked of children in grades 6 through 12. Nonresident fathers' involvement in schools decreases the adjusted odds that children have ever been suspended or expelled from school even after controlling for mothers' involvement in school. The adjusted odds that children have ever been suspended or expelled among all children in grades 6 through 12 who have nonresident fathers' (model 1) are 50 percent less if the fathers have participated in only one activity and 59 percent less if the fathers have participated in two or more of the school activities.
Nonresident fathers with contact but no involvement in their children's schools. Another interesting pattern in the models is that children who have had no contact with their nonresident fathers in the last year are somewhat more likely to enjoy school than children who have had contact, but whose fathers have not participated in any of the school activities. This association is strongest among children in grades 6 through 12 and persists even after controlling for mothers' involvement in the schools. It may be that children with some contact with their fathers, but whose fathers are not as involved as the children might wish, face more ongoing psychological strain that also affects their attitude towards school than children who do not expect that their fathers are going to be involved because they have not seen them at all. There is some support for this speculation in the literature (Nord and Zill, 1996). There is other evidence that children who have not seen their fathers at all in the past year are somewhat better off than children who have seen their fathers but whose fathers have had no involvement in their schools. Even after controlling for mothers' involvement in their schools, the adjusted odds that children have ever been suspended or expelled are 29 percent less (significant at 0.10 level) if they have not seen their fathers at all in the past year compared to children who have seen their fathers, but whose fathers have not participated in any of the school activities. Moreover, for two of the outcomes, getting mostly A's and having ever repeated a grade, there is no difference in the outcomes among children whose fathers have had contact with them but did not participate in any of the school activities and children who have not had any contact with their fathers in more than a year or who have never had contact with their fathers.
Taken together, these results may offer a clue as to why existing studies yield mixed results on the importance of nonresident fathers' involvement in children's lives: The measures that are used to assess nonresident fathers' involvement may not be adequate. Often involvement is measured by the simple measure days of contact. The results in these models suggest that it is not contact, per se, that is important, but rather other dimensions of involvement that go along with contact that are beneficial to children's lives./30 Indeed, contact may be a mixed blessing if the contact is enough to tantalize children but not enough to satisfy. Although with the data in hand it is not possible to determine what it is about involvement in schools that is beneficial for children it seems likely that fathers who make the effort to attend school functions may be demonstrating to their children how much they care about them and the importance that they place on education. Their involvement may also reflect their level of commitment to their children. If future studies include more specific measures about what fathers do with their children and the circumstances under which contact occurs (e.g., is it regular, is it disruptive to the children, how do the children feel about it), then more consistency in the results might be found.
Overall, the results in this section provide strong evidence that nonresident fathers' involvement in their children's schools is important to children, particularly to older children.
FOOTNOTES:
[13] The words significant and significantly when used in this report always indicate statistical significance (at 0.05 level, unless noted otherwise).
[14] The NHES:96 asked parents whether their children's schools had general school meetings or parent-teacher conferences since the beginning of the school year. Tabulations of these data for important subgroups such as two-parent, single-mother, and single-father families revealed no systematic differences in their opportunities for involvement. Moreover, when the analyses are restricted to only those children whose schools offer the opportunity for involvement, declines in parental involvement are still observed as children move from elementary to middle to high school. The declines, however, are less pronounced.
[15] The pattern of decline for mothers and fathers is similar for attending a general school meeting or a conference, so these data are not shown. Data are also not shown for single-parents because the pattern of decline for mothers and fathers is similar for all the activities.
[16] Families were said to have received Federal assistance if they had received funds or services from any of the following programs in the past 12 months: Women, Infants, and Children (WIC), Food Stamps, or Aid to Families with Dependent Children (AFDC).
[17] In figure 8a and several other figures, mothers and fathers with moderate levels of involvement in their children's schools are not shown. They were omitted to conserve space and to highlight the difference between parents with high and low levels of involvement.
[18] Kindergartners were excluded from the models for two reasons. First, the items that were used to create the school climate scale were not obtained for children in kindergarten. Second, of the five student outcomes examined, only grade repetition was asked of kindergartners. Separate reduced models that limited the regressors to child and family characteristics were estimated for children in kindergarten through 12th grade and for children in 1st through 12th grade to explore whether the exclusion of kindergartners affected the results. The dependent variables examined were high father involvement, high mother involvement, and grade repetition in two-parent and in single-parent families. Only one result changed when kindergartners were included: in two-parent families, fathers were less likely to be highly involved in their children's schools when the mother was not working.
[19] The logistic regression results in this report are presented as adjusted odds ratios. See the Methodology and Data Reliability section, p. 93 for details on adjusted odds ratios and how to interpret them. In the discussion, the results are expressed as percent changes in the odds. The percent change is calculated as (odds ratio -1)*100, with a negative result indicating a percent decrease and a positive result indicating a percent increase in the odds.
[20] None of the models shown in the following pages include information on whether schools offered general school meetings or had held parent-teacher conferences since the beginning of the school year. In response to a reviewer's comments, additional models were estimated that included this information to determine whether it altered any of the results presented. As expected, parents are more likely to be highly involved in their children's schools if the schools offer them opportunities for involvement compared to if the schools do not. The addition of information on opportunities for involvement, however, does not materially alter the observed associations between the other factors in the models and high parental involvement. The three instances where results are affected are noted in the discussion.
[21] When information on whether schools offered general school meetings and parent-teacher conferences is added to the model, the association between mothers' involvement in their children's schools and maternal education becomes insignificant (p=.11) among children in grades 1-5 living in two-parent families.
[22] A unit increase indicates the amount of additional schooling that fathers have obtained (e.g., no high school degree, a high school graduate, some college or vocational school experience, a college graduate, and graduate or professional school experience).
[23] When information on whether schools offered general school meetings and parent-teacher conferences was added to the model, the association between high mother involvement and frequency with which a parent helps with homework became insignificant.
[24] The correlations between high mother involvement and high father involvement in schools and regularly participating in an ongoing service activity are 0.28 and 0.21, respectively.
[25] Frequency of helping with homework is measured as follows: never, less than once a week, 1 to 2 times per week, 3 to 4 times a week, or 5 or more times a week. Children who did not receive homework were included in the never category. A variable that took the value of 1 when no homework was given and 0 otherwise was added to the model to remove the influence of those who did not receive homework from the estimate of the influence of homework frequency.
[26] A scale of school climate was created by summing the responses to the five statements about children's schools shown in figure 12: teachers maintain good discipline in the classroom; most students and teachers respect each other; the principal and assistant principal maintain good discipline at the school; the school welcomes my family's involvement with the school; and the school makes it easy to be involved there. The response categories were assigned a value from 1 (strongly disagree) to 4 (strongly agree). Thus, the scale ranges from 5 to 20 with higher scores indicating a more positive school climate.
[27] The student outcomes are based on parent reports, and parents tend to provide positive assessments of their children. For example, 38 percent of children in 1st through 12th grade get mostly A's, according to the report of the parents. It is likely that if school records were used to obtain this information instead, the proportion would be lower.
[28] For children in grades 1 through 5, three social capital measures are included in the models: the number of in-home activities parents have shared with their children, the number of out-of-home activities they have shared together, and whether the parents have told their children a story in the past week or shared family history with them in the last month. For children in grades 6 through 12, seven social capital measures are included in the models: expect children will graduate from a 4-year college, confidence that someone in the household can help children with homework, whether have discussed educational plans with children, the number of activities have shared with children in the past week, frequency with which parents help with homework, frequency with which parents attend religious services, and whether parents regularly participate in a community service activity.
[29] See footnote 28 on page 56 for a listing of the social capital measures that are included in the models.
[30] One reviewer noted that the models presented in table 10 do not include information on nonresident fathers' frequency of contact with their children. To address the reviewer's concerns, additional models were estimated that included information on both nonresident fathers' frequency of contact with their children and their level of involvement in their children's schools. Model 2 was re-estimated for three of the five outcomes shown in table 10 with information on frequency of contact added to the existing set of explanatory variables. The outcomes examined were the following: get mostly A's, ever repeated a grade, and ever suspended or expelled. The first two outcomes were examined for children in grades 1 through 12. The third outcome was examined for children in grades 6 through 12.
When both nonresident fathers' frequency of contact and their level of involvement in their children's schools are included in the same model, it is their involvement in school, not their frequency of contact that is important. Nonresident fathers' involvement in their children's schools remains a significant influence on all three outcomes, even after controlling for their frequency of contact. Nonresident fathers' frequency of contact with their children, on the other hand, is not a significant influence on any of the outcomes after controlling for their school involvement.
Findings (previous part)
Summary and Discussion
