Conditions affecting the association between father identity and father involvement.

To better understand the conditions under which fathers are more or less involved with their children, we tested the moderating influences of interparental relationships on the association between identity and behavior in a sample of 186 married and 93 divorced fathers. Results showed that identity investment and satisfaction were positively associated with involvement in child-related activities, but identity salience was not. Also, the relationship between satisfaction and involvement was stronger when there was less cooperation and less indirect conflict, and the relationship between investment and involvement was stronger when there was less cooperation. These moderating effects were stronger for divorced fathers.

Keywords: father identity, father involvement, divorced fathers, interparental relationships, identity, satisfaction

**********

Research demonstrates that fathers have less involvement with their children than do mothers, and following divorce father involvement decreases (Pleck, 1997). Pasley and Braver (2004) argued that the context of postdivorce families makes father involvement more complicated. In spite of such complications, research shows that just under half of divorced fathers (47%) saw their children 1-3 times per month or more, and 25% saw their children at least weekly (Seltzer, 1991). Although some fathers report less close relationships with their children following divorce (Booth & Amato, 1994), many adult children of divorced parents report maintaining close relationships with their fathers (Ahrons & Tanner, 2003).

Whereas some debate exists regarding the extent of father involvement in married and divorced families, the importance of father involvement is well documented (Marsiglio, Amato, Day, & Lamb, 2000), so scholars have begun to examine how contextual factors affect involvement (e.g., Rane & McBride, 2000; Rettig & Leichtentritt, 2001; Stone & McKenry, 1998). The present study adds to this literature by examining how dimensions of the interparental relationship moderate the link between father identity and involvement in a sample of married and divorced fathers.

THEORETICAL FRAMEWORK

Identity theory (Stryker, 1968) has been used to explain both levels and forms of father involvement, positing that fathers' involvement with children stems from the meanings and importance they assign to being fathers (Ihinger-Tallman, Pasley, & Buehler, 1993; Marsiglio et al., 2000). According to identity theory, it is through social interactions that statuses (e.g., father, husband) and roles (e.g., provider, nurturer, disciplinarian) are given meaning, and behaviors reflecting these statuses and roles are either inhibited or reinforced. The meanings that individuals attach to particular roles result in the creation of identities, and these identities subsequently guide behavior (Burke & Reitzes, 1981). Over time, behaviors associated with particular identities become stable and are invoked across a wider variety of situations, rather than being situation- or context-specific. Some recent research suggests that the relationship between identity and behavior is bidirectional, such that individuals' behaviors also affect their identities (e.g., Cast, 2003). Because of the data used here (collected at a single point in time), we limited our examination to the association between father identity and involvement rather than any causal relationships.

Identity is defined as "internalized sets of role expectations" (Stryker, 1987, p. 90). Thus, father identities are conceptualized as fathers' self-perceptions and expectations regarding how they should enact different roles within the father status. This conceptualization has been measured in a number of ways, including fathers' satisfaction with and competence enacting their father identities, role clarity, and willingness to invest time and resources in their father identities (e.g., Fox & Bruce, 2001; Ihinger-Tallman, Pasley, & Buehler, 1993; Pasley & Minton, 1997; Stone & McKenry, 1998). Here we operationalized father identity as consisting of men's selfperceptions regarding satisfaction and investment in the father status.

Structural identity theorists like Stryker also posit that statuses and their associated identities are arranged hierarchically according to their salience and centrality. Salience is the likelihood that a particular status, role, or identity will be invoked in any given situation in comparison to the likelihood that other statuses, roles, or identities might be invoked (Rane & McBride, 2000; Stryker & Serpe, 1994). Salience is not a part of one's consciousness but reflects only the probability that an identity will be enacted (Rane & McBride), and usually it is measured by asking individuals to name the first thing they would tell someone about themselves (e.g., Minton& Pasley, 1996; Stryker & Serpe). Centrality requires conscious awareness and reflects the importance an individual attaches to an identity; centrality usually is measured by asking individuals to rank different identities according to their importance (Rane & McBride).

Identity theory traditionally has held that an individual's behavior is more likely to reflect identities that are both salient and central. However, scholars have challenged the influence of identity salience. Studies that contain measures of both salience and centrality (e.g., Rane & McBride, 2000; Stryker & Serpe, 1994) generally found that centrality is more influential to behavior than salience. Because our data did not include a measure of centrality, we could not test its relative importance compared with salience. However, we did measure salience and were able to test its association with behavior.

Commitment to an identity also is associated with the enactment of identity-related behaviors. According to identity theory, commitment consists of the relationships that support the enactment of an identity and the costs associated with giving up these relationships if the identity is not enacted (Burke & Reitzes, 1991; Stryker & Serpe, 1994). Theoretically, commitment to father identities should be higher when a greater number of important relationships encourage the enactment of identity-related behaviors. Thus, commitment acts as a moderator such that higher levels of commitment strengthen the relationship between identity and behavior, Although identity theory does not address the potential effect of relationships that discourage the enactment of an identity, theoretically, fathers with more relationships that discourage the enactment of their father identities should be less likely to behave in accordance with their identities. Specific to this study, we assessed how the interparental relationship as an indicator of commitment affects involvement by examining both the level of cooperation and the levels of direct and indirect conflict. If interparental interactions are more cooperative (supportive), then commitment to the father identity should be high and fathers' involvement also should be higher; however, if interparental interactions are more negative (high direct and indirect conflict), then the relationship between identity and behavior should be weakened.

Although Burke and Reitzes (1981) posited that commitment moderates the translation of identity into behavior, no research to date has tested the existence of this moderating effect with fathers or the potential moderating influence of unsupportive relationships. As such, we make a contribution to the literature by examining how the relationship between father identity and fathering behaviors is moderated by commitment through examining interparental interactions. Because research shows that divorced parents often interact in unsupportive ways (e.g., mothers interfering with visitation, former spouses drawing the children into interparental conflict; Braver, Wolchik, Sandler, & Sheets, 1993), we also hypothesize that the moderating influence of these indicators of commitment to an identity would be stronger for divorced fathers.

FATHER INVOLVEMENT

Great variation exists in frequency and types of father involvement and in the expression of fathering behaviors (Pasley, 1994; Pasley & Minton, 1997). This variation is largely explained by the number of factors that potentially influence father involvement, including fathers' personal characteristics (e.g., father's age, ethnicity, education, income), relational factors (e.g., mother's perceptions of the father's abilities and skills as a father), and other outside factors (e.g., mother's employment status, age and sex of the child) (Rane & McBride, 2000). For example, father involvement typically declines over time, with the predominant explanation being child maturation; that is, as children age they require less direct supervision and more indirect forms of involvement (e.g., monitoring, worrying) (Harris, Furstenberg, & Marmer, 1998). However, Parke and Buriel (1998) found that the relationship between involvement and child age was nonlinear and fathers were less involved when their children were infants, most involved when their children were young (preschool) or adolescents, and least involved during middle childhood. Due to the potential influence of such factors, here we included measures of age of oldest child, average age of all biological children, and length of marriage as well as fathers' employment status, education level, and household income as potential control variables.

POSTDIVORCE FATHER INVOLVEMENT

Although many of the generalizations regarding father involvement apply to fathers in both first-married families and divorced families, being part of a postdivorce family system presents additional barriers to involvement. For example, the physical separation of fathers from their children's daily lives means that nonresident fathers are no longer a de facto influence through their presence in the household (Fox & Blanton, 1995). This poses serious difficulties for some nonresident fathers, as parenting skills are obtained primarily through practice (Fox & Blanton). Like married fathers, divorced fathers tend to become less involved with their children over time (Lamb, 2000), in part due to their children's maturation (Pasley & Braver, 2004). However, some divorced fathers completely disengage from their children, dramatically reducing or ending their involvement. Although research shows that interference with visitation by the mother and conflictual interparental relationships are associated with disengagement (Arendell, 1995; Kruk, 1994), some of this decline simply is due to fathers' geographic distance from their children (Dudley, 1991).

Due to the potential influence of these factors specifically for divorced fathers, measures of geographic proximity and length of prior marriage to the child's mother were included and used as control variables. Because no information was available from mothers, mothers as informants are not included in our model.

INFLUENCES ON FATHER INVOLVEMENT AND FATHER IDENTITY

Research has established the link between father identity and father involvement, although there only is limited support for the influence of identity salience specifically. Ihinger-Tallman et al. (1993) found a moderate positive relationship between measures of parenting role identity (role satisfaction, competence, investment, and salience) and father involvement. Minton and Pasley (1996) also found similar relationships between the first three aspects of father identity and father involvement, but found no relationship between identity salience and involvement. Stone and McKenry (1998) also reported that the fathering identities of nonresident fathers were related to father involvement, although they observed that role clarity, rather than the salience of the nurturer role, was the single best predictor of involvement. Unlike others, Bruce and Fox (1999) found that the salience of the father status was associated with involvement and that age of the child and the resident status of the father moderated this relationship. Rane and McBride (2000) found that the centrality of the father status was not associated with father involvement, but centrality of the nurturer role within the father status was.

Most research to date has focused primarily on confirming the link between identity and involvement rather than on determining the factors that affect the relationship or whether this varies for married and divorced fathers. We extend previous work by focusing on the interparental relationship as a key moderating factor and assessing the strength of its influence in married and divorced fathers.

INTERPARENTAL RELATIONSHIPS

Studies show that the quality of the interparental relationship influences father involvement for both married and divorced fathers: cooperative, low-conflict relationships are associated with more father involvement, whereas high-conflict relationships are associated with diminished involvement (Kelly, 2000; Whiteside, 1998). Interparental conflict is recognized as having at least two distinct dimensions: direct (conflict that involves direct interactions between the spouses or former spouses) and indirect (behaviors that are designed to triangulate or indirectly compete with the spouse or former spouse). Each of these dimensions is conceptually distinct from cooperation (Buehler & Trotter, 1990). Noteworthy is Whiteside's (1998) finding that cooperation and conflict co-exist in many relationships, and so long as such interactions lack hostility and anger, conflict will not severely damage the father-child relationship. Thus conflict, in and of itself, does not appear to be harmful in either married or divorced families, yet hostile conflict appears to be detrimental. High levels of hostile conflict are more likely among divorced families; thus, we anticipated that the moderating influence of such conflict also might be stronger among divorced families.

We examined three measures of interparental relationship quality: hostile direct conflict, indirect conflict, and cooperation. We hypothesized that these indicators of relationship quality would moderate the link between identity and involvement, such that self-perceived father identities would be more related to involvement when interparental relationships had less direct and indirect conflict and more cooperation and that these effects would be stronger for divorced fathers.

METHOD

PARTICIPANTS

We used extant data from a survey of 303 married and divorced fathers in central North Carolina (see Minton & Pasley, 1996). The sample was reputational, obtained from volunteers at presentations given to community groups, churches, and undergraduate classes. Questionnaires were mailed to prospective participants and followed up according to Dillman (1978).

The sample is best described as White, middle-income, early- to middle-aged men who are well educated. Married and divorced fathers did not differ significantly on age, ethnicity, level of education, or occupation. Married fathers were more likely to be employed full-time; have higher annual incomes, more children, and younger children; and be married longer to their children's mothers (see Table 1).

Because the sample was a convenience sample, generalizability is limited. Minton and Pasley (1996) demonstrated that (at least for the divorced, nonresident fathers) there were few differences between the sample characteristics and fathers included in the NSFH (Wave 1), except our sample was more likely to be White, more educated, and have higher incomes. They were of similar age, employment status, and number of children. Unfortunately, nationally representative data regarding the demographic characteristics of fathers (and particularly divorced fathers) still are lacking (Hernandez & Brandon, 2002), and so more detailed comparisons are not possible at the present time.

PROCEDURE

Questionnaires were sent to 485 fathers (326 married; 159 divorced), and 303 completed questionnaires were received (200 married; 103 divorced); the response rate was 62%. Only 279 fathers (186 married; 93 divorced) met criteria for inclusion in our analysis: married fathers were in a first marriage to the biological mother of their child and had a child less than 18 years old; divorced fathers had one prior marriage with a child less than 18 from this marriage who resided primarily with the mother. These criteria were used to eliminate the possible confounding effects of (a) prior marriages and (b) nonbiological connection to children, as well as possible effects of formal and informal custody issues that arise under shared physical custody arrangements.

MEASURES

Father Identity. Father identity was measured by asking fathers about their perceived satisfaction in the father status and their identity investment, using an adapted version of the Self-Perceptions of the Parental Role Scale (MacPhee, Benson, & Bullock, 1986), a measure originally designed for use with low-income mothers. Factor analysis (principal components analysis with an oblimin rotation) resulted in two subscales: identity satisfaction (12 items) and identity investment (four items) (see Table 2). The first subscale reflects fathers' self-perceived satisfaction with being a father, and the investment subscale represents their self-perceived efforts toward and investment in being a good father. Responses used a structured alternative format (e.g., "Being a parent is a satisfying experience to some adults," but "For other adults, being a parent is not all that satisfying"--identity satisfaction). Respondents first decide which statement best reflects themselves, then mark whether the statement is really true or sort of true for them, resulting in a four-point scale. A sample item for investment was "Some parents do a lot of reading about how to be a good parent," but "Other parents don't spend much time reading about parenting." Item means were used. Alpha reliabilities for mothers ranged from .72 to .80 (MacPhee et al.); here the alpha was .88 for satisfaction and .74 for investment.

Our measure of father identity salience asked fathers to "Think about meeting people for the very first time" and queried: "If you could only tell them one thing about yourself, what would you say first?" In a second item fathers then were prompted to describe what they would say next. Each of the two responses was coded according to whether it was related to being a father (0 = no; 1 = yes), and the scores from the two items were summed, resulting in a possible range from 0 to 3 (0 = no response related to fathering, 1 = only second response related to fathering, 2 = only first response related to fathering, 3 = both first and second responses related to fathering). Higher scores indicate that the father identity was invoked more readily and therefore more salient. Although order effects were a concern for this measure (i.e., the mere process of responding to a fathering survey might prompt fathers to invoke their father identities more readily than they would in other contexts), the salience measure appeared early in the survey, and the low frequency of responses related to fathering for these items argues against the presence of such effects. Only 9% of participants' responses to the first query and 19% of responses to the second query related to fathering.

Commitment. Identity commitment was assessed using measures of the levels of hostile direct conflict, indirect conflict, and cooperation in the interparental relationship. Hostile direct interparental conflict was measured using nine items adapted from Rands, Levinger, and Mellinger (1981) and Jacobson (1978). Items asked about disagreements between the husband and wife/ex-wife about "something important to her." Items were coded so higher scores indicate more hostile conflict (such as yelling and name-calling). Sample items included "She says or does something to hurt my feelings" and "She tries to reason with me" (reverse coded). Responses reflected how well each response described their wife/ex-wife and ranged from 1 = not at all to 4 = very well. Responses to items were summed and item means used. An alpha of .92 was obtained.

Indirect interparental conflict was measured using two items from Kurdek (1987) that assessed the degree to which the father attempts to engage in competitive or triangulating behaviors. Fathers were asked how well each item described their relationship with their wife/ex-wife. Items were "'How often do you encourage the children to side with you?" and "How often do you say bad things about your [wife's/ex-wife's] character?" Responses, ranging from 1 = never to 5 = always, were summed, and item means were used. Higher scores indicate more indirect conflict in the interparental relationship (alpha = .56 for the two items).

Interparental cooperation was measured using six items from Ahrons's (1981) interparental support scale. Responses ranged from 1 = never to 5 = always and measured the degree to which fathers perceived their wife/ex-wife to be supportive, accommodating, and understanding in parenting interactions. Sample items included "How often do you see yourself as a resource to your [wife/ex-wife] in raising the child/ren?" and "How often does your [wife/ex-wife] provide emotional support in dealing with the child?" Responses to items were summed, item means were used, and higher scores reflect more cooperation in the interparental relationship (alpha = .77).

Father Involvement. Fathers reported the frequency of involvement on 11 child-related activities adapted from the Ahrons (1983) scale. Two original items were omitted (dressing and grooming; taking the children for recreational activities), and three additional items were added (helping with schoolwork, planning and preparing meals, going to the doctor or dentist). The original response set was retained (1 = not at all to 5 = very much), so higher scores indicate more frequent involvement. Other sample activities included discipline, celebrating holidays, and going on vacation. Responses were summed, and item means were used. The alpha reliability on the original measure was .97, and an alpha level of .91 was obtained here.

DATA ANALYSIS

Our goal was to identify the most parsimonious model to explain father involvement from the perspective of identity theory. We first examined bivariate relationships to determine which of the demographic variables to include as controls (see Table 3 for correlations between variables). Thus, in the first block of the regression we included age of oldest child because it was significantly related to father involvement (r = -.26), as were total number of children in the household (r = .37) and father's geographic proximity (r = -.36). Proximity was recoded dichotomously as 0 if fathers lived less than 50 miles from their children (approximately two-thirds of the sample, since this category included all married, resident fathers) and 1 if fathers lived 50 or more miles from their children (approximately one-third of the sample). All identity measures were included in the second block. Our decision to retain identity salience in the preliminary regression was based on its conceptual importance, although it was not significantly associated with involvement in the bivariate analyses (r = .06). All measures of interparental relationship quality also were included in the second block, as was marital status.

Regression analyses with interaction terms were used to test the moderating effects of interparental relationship quality on the association between father identity and father involvement. The initial analysis included two-way interactions between each of the independent variables and marital status to test for differences between married and divorced fathers in each of the independent variables of interest. Because no significant differences were found, a single model was tested that included both married and divorced fathers. A block design was used, with pertinent demographic information entered first; blocks that followed included main effects (self-perceived father identity and interparental relationship quality measures), two-way interaction terms (interparental relationship quality measures x father identity) to test whether interparental relationship quality moderated the relationship between identity and involvement, and three-way interaction terms to test for any significant differences between married and divorced fathers. Variables were centered to correct for possible heterogeneity of variance (Jaccard, Turrisi, & Wan, 1990). Standardized beta coefficients were examined for significance as well as significant F-changes for each block.

RESULTS

We ran the full model to determine which variables to include in the reduced, more parsimonious model (see Table 4 for results of the regression analysis of the full model). Based on these findings, we included two control variables (age of oldest child and geographic proximity) and two identity variables (satisfaction and investment) in the reduced model. The full model further confirmed findings of prior research that salience was not associated with behavior, and it was dropped from the second model. All interparental relationship variables were retained, but only six two-way interactions and two three-way interactions were included.

In the reduced model (see Table 5), age of oldest child ([beta] = -. 13, p < .01) and geographic proximity ([beta] = -.15, p < .01) were significantly related to involvement, F(2,279) = 40.51, p < .01, explaining 23% of the variance in father involvement. As expected, fathers whose children were younger and who lived closer to their children were more involved in child-related activities.

Identity satisfaction ([beta] = .11, p < .05), identity investment (13 = .12, p < .01), indirect conflict ([beta] = -. 12, p < .01), interparental cooperation ([beta] = .36, p < .01), and marital status ([beta] = .20, p < .01) also were significantly related to involvement, explaining an additional 29% of the variance, F-change(8,279) = 26.08, p < .01. Fathers were more involved when they viewed their father identities as more satisfying, when they were more invested in these identities, when their interparental relationships had less indirect conflict and were more cooperative, and when they were married.

Regarding interaction effects, five were found. Both indirect conflict and cooperation moderated the relationship between identity satisfaction and involvement (indirect conflict x satisfaction, [beta] = -.20, p < .01; cooperation x satisfaction, [beta] = -.l2, p < .05); however, cooperation did not moderate the association in the hypothesized direction. Although we hypothesized that high levels of cooperation would be associated with stronger associations between identity and involvement, the positive relationship between identity satisfaction and involvement was stronger when there were lower levels of indirect conflict and lower levels of cooperation. Interparental cooperation also moderated the relationship between identity investment and involvement (cooperation x investment, [beta] = -.23, p < .01). Again, the positive relationship between identity investment and involvement was stronger when there were lower levels of cooperation. Finally, the interactions between direct conflict and satisfaction and indirect conflict and investment approached significance (direct conflict x satisfaction, [beta] = -.10, p = .08; indirect conflict x investment, [beta] = -.09, p = .06). In each case, lower levels of conflict were associated with a stronger positive relationship between identity and involvement, F-change(12,279) 3.04, p < .01.

One three-way interaction was found for marital status by indirect conflict on the relationship between identity satisfaction and involvement ([beta] = .15, p < .05), F-change(3,279) = 3.15,p < .05, [DELTA][R.sup.2] = .01. The moderating effects of indirect conflict were stronger for divorced fathers. Like the full model, the parsimonious model explained more than half of the variance in father involvement (Adjusted [R.sup.2] = .53); copies of the figures plotting the various interactions can be obtained from the first author.

Taken together, these results indicate support for the first hypothesis, as both identity satisfaction and investment were associated with involvement; however, salience was not associated with involvement. The second hypothesis also was partly supported. Moderating effects were found for indirect conflict on the relationship between identity satisfaction and father involvement; however, the moderating effects of cooperation on the relationship between investment and father involvement and identity satisfaction and father involvement were opposite from the hypothesized direction. Finally, the third hypothesis was partly supported, only for "the effect of indirect interparental conflict on the relationship between identity satisfaction and involvement, and this effect was stronger for divorced fathers.

DISCUSSION

According to identity theory (Stryker, 1968), father involvement can be explained through the meanings and importance men assign to their individual father identities. Although the association between identity and behavior can be influenced by many factors, we specifically tested the ideas of Burke and Reitzes (1991) about the influence of identity commitment by investigating the relationship between father identity and father involvement and the ways in which this link is affected by the quality of the interparental relationship (our indicator of commitment).

Our findings that only identity investment and identity satisfaction were associated with father involvement replicate those of others (Minton & Pasley, 1996; Rane & McBride, 2000). Again, salience of the father identity was not associated with involvement. The simple fact that a father mentions his status to people in new encounters apparently does not play out in his self-reported behavior as measured here. Although we agree with Rane and McBride that the centrality of a father's identities likely is more important to explaining his behavior than is salience, our data did not permit us to test this directly.

Our hypothesis that the quality of the interparental relationship as an indicator of commitment to identity would moderate the relationship between father identity and father involvement found partial support. The positive relationship between identity satisfaction and involvement was stronger when there were lower levels of indirect conflict and lower levels of cooperation, and the positive relationship between identity investment and involvement was stronger when there were lower levels of cooperation.

These results partially support the assumption of Burke and Reitzes (1991) that when commitment to an identity is low, behavior is less likely to be associated with that identity. When relationships are actively unsupportive of their identities (as evidenced by high indirect conflict here), fathers are less likely to act in accordance with their identities. However, our findings also suggest the need for additional specificity regarding the influence of supportive and unsupportive relationships. Specifically, we found that when fathers have relationships that support an identity of being involved fathers (i.e., cooperation is high in the interparental relationship), their behavior reflects this support, regardless of their self-perceptions of themselves as fathers. That is, fathers with cooperative interparental relationships remained highly involved with their children, irrespective of whether they personally viewed themselves as being satisfied or invested in their father identities. This likely is due to the fact that our measure of support assesses support for involvement more than support for the father identity per se. Fathers could feel either satisfied or dissatisfied with themselves as fathers, and according to identity theory fathers who were less satisfied with themselves as fathers would be less involved with their children. However, here coparental relationships were classified as supportive if they supported fathers' involvement with their children, regardless of whether fathers were satisfied or dissatisfied with themselves as fathers. Thus, our measure of interparental support could more accurately be described as assessing support for a behavior (involvement) rather than support for an identity (satisfied or dissatisfied). Given the limitations of our measure, then, it is not surprising that lower levels of cooperative interparental behaviors were associated with a stronger link between identity and behavior.

The fact that direct conflict did not moderate the association between identity satisfaction and involvement might be due to our measure of direct conflict, which assessed generalized conflict between the spouses/former spouses, whereas the measures of indirect conflict and cooperation dealt specifically with coparenting conflict (i.e., child-related issues). Previous literature demonstrates that coparenting conflict is more likely to affect parenting than is generalized conflict (e.g., Margolin, Gordis, & John, 2001); therefore, had our measure of direct conflict focused on child-related issues as well, then it is likely that direct conflict might have moderated the links between identity and involvement. Without further testing of this proposition, we cannot be certain whether the differences in moderation we found are due to a lack of influence by direct conflict or to differences in the constructs we assessed (generalized conflict versus coparental conflict).

Our hypothesis that the moderating influence of the interparental relationship would be stronger for divorced than married fathers was partly supported. Specifically, the influence of indirect conflict on the relationship between identity satisfaction and involvement was stronger for divorced than married fathers. Although we found only one significant three-way interaction (likely due to the relatively small size of our sample), again these findings build upon prior research (e.g., Bruce & Fox, 1999; Minton & Pasley, 1996) regarding the influence of fathers' marital and resident status on interparental relationship quality and involvement as well as the link between identity and involvement. However, further studies using larger samples need to be conducted before we can conclude how and whether the moderating influence of the coparental relationship varies according to marital status.

Theoretically, these findings highlight the particular importance of supportive and unsupportive relationships for fathers who, by way of their nonresident status, are removed from the day-to-day interactions that maintain their father identities and that living with their children in the marital household provides. For nonresident fathers, the quality of the interparental relationship might be even more important, since those with highly conflicted interparental relationships experience a lack of support as well as an actively unsupportive relationship that discourages their enactment of their father identity (i.e., their involvement with their children). The primary reason that divorced fathers interact with their former spouses is to maintain involvement. Consistent with prior research (Arendell, 1995; Kruk, 1994), divorced fathers with hostile and uncooperative interparental interactions may attempt to avoid these negative interactions by reducing their involvement, thereby eliminating the need for contact with their former spouses.

Interestingly, the moderating effect of indirect conflict was stronger only for the association between divorced fathers' identity satisfaction and their involvement but not for the association between identity investment and involvement. Here, triangulating behaviors and interparental denigration might be particularly damaging for the link between identity and behavior for divorced fathers, such that under conditions of high indirect conflict, fathers who are more satisfied with their father identities actually are less involved with their children. However, the link between fathers' identity investment and their involvement is not affected. This is consistent with other literature that suggests that there is a direct association between fathers' parenting skills/knowledge and their involvement in parenting (Dickie & Gerber, 1980; McBride, 1990), and it appears from our findings that this relationship is not affected by the quality of the coparental relationship.

Similarly, fathers with cooperative interparental relationships tended to be more involved with their children, irrespective of the level of satisfaction or investment in the father identity. This is somewhat inconsistent with predictions from identity theory and suggests that the direct effects of cooperation are stronger than its moderating effects. However, as noted, this finding is likely due to our measure of cooperation, which did not account for interparental behaviors that support negative self-perceptions (low satisfaction and investment). This presents some interesting methodological considerations, since other research regarding commitment to an identity also typically has assessed support for identities that are viewed positively (e.g., Burke & Reitzes, 1981, 1991). Our findings suggest that support for particular behaviors (e.g., involvement) might influence individual behavior more strongly than does self-perceived identity when that identity is viewed negatively, a proposition which should be tested further.

Further, it was only indirect conflict and cooperation that served as stronger moderators for the link between identity satisfaction and involvement. As noted earlier, this in part might be due to our measure of direct conflict. The findings also might be due to triangulating behaviors being directly under the father's control (fathers reported only their own behaviors), whereas direct conflict assessed his perceptions of his spouse's/former spouse's behaviors. The proposition that one's own behaviors within the coparenting relationship (triangulation or denigration of the spouse) influences and is influenced by the ways in which he enacts his fathering identities (i.e., involvement with his child) is consistent with identity theory. For example, Burke (1991) argued that individuals seek to verify their identities by engaging in interactions and exhibiting behaviors that they see as being consistent with. their self-perceived identities. Although these interactions are dyadic in nature, identity theory tends to focus on the influence of others (via feedback or support), rather than the contributions of the individuals themselves to the interaction. Given our findings here that individual behaviors (particularly triangulating and denigrating) might be more influential to identity enactment than would the lack of support received from others, this is an area worthy of further exploration.

We recognize that other theoretical frameworks can explain fathering (e.g., social exchange [Braver, Wolchik, Sandler, & Sheets, 1993], resource theory [Rettig & Leichtentritt, 2001], role enactment [Stone & McKenry, 1998], interactional-feminist [Arendell, 1995]). Unfortunately, the data we used were collected explicitly to test assumptions from identity theory. Therefore, we have no data with which to determine whether rival theories might better explain father involvement. For example, the data did not include measures of costs, rewards, and comparison levels needed to apply social exchange theory or measures of perceived power and control that might be informed by feminist perspectives.

IMPLICATIONS

Collectively, our results have important implications, as they support the notion that interventions and policies aimed at increasing father involvement should consider contextual factors. Fathers, and particularly divorced fathers, might benefit from relationships that actively support their father identities as well as relationships that are not actively hostile toward them. Overall, fathers with low indirect conflict and high cooperation in their interparental relationships tended to be more involved with their children. Fathers whose interparental relationships were low in indirect conflict tended to act in accordance with their father identities, whereas fathers whose interparental relationships were high in cooperation were more involved with their children regardless of their own self-perceived satisfaction and investment in the father identity. Equally important is that fathers hold positive self-perceptions, because such perceptions might affect behavior. Fathers who had low levels of investment in their father identities and who felt less satisfied with their enactment of their father identities tended to be less involved with their children, regardless of the quality of the interparental relationship. However, the combination of having both positive self-perceptions and positive interparental relationships provided even greater explanatory power for involvement, particularly among divorced fathers.

Enhancing strategies that improve fathers' self-perceived satisfaction with their abilities as fathers and that foster their investment in being fathers should be a focus in intervention. Working with both parents to implement strategies that improve the quality of the interparental relationship is ideal; however, working with even one parent also might foster higher cooperation and reduce direct hostile conflict and triangulating/denigrating behaviors. Especially when working with divorced individuals, the importance of positive interparental interactions should be emphasized since negative interparental interactions following divorce appear to diminish the involvement of even highly invested fathers; similarly, positive interparental interactions appear to promote involvement even for dissatisfied or less invested fathers (see also Geasler & Blaisure, 1998, for a review of divorce education programs).

RECOMMENDATIONS FOR FUTURE RESEARCH

We recognize several limitations of our study that should be addressed in future research. The sample was not representative, and a more representative sample (e.g., greater socioeconomic and ethnic/racial variation) undoubtedly would yield richer findings. If our findings can be replicated regarding the differential influence of the interparental relationship for divorced and married fathers, it is likely that these relationships also could differentially affect the involvement of fathers with varying levels of economic resources and different cultural backgrounds.

Further, although identity theory holds that identity is dynamic and both influences and is influenced by relationships with others and behavior, our findings cannot be interpreted as being causal. It may be that fathers who think more positively about themselves as fathers come to have more positive interparental relationships or to be more involved with their children. The reverse also may be true, that fathers who have more positive interparental relationships or who are more involved with their children develop more positive father identities (although this would seem to be contradicted by our finding regarding cooperation being associated with higher levels of involvement at all levels of satisfaction and investment). Determining the nature and direction of any causal relationship would provide additional understanding of identity development and refinement.

Our findings bolster the proposition that fathering behaviors are differentially associated with different aspects of identity (salience, investment, satisfaction, commitment). Because most research has conceptualized identity unidimensionally, future efforts should further investigate the influence of different identity theory constructs and the ways in which each is related to behavior. Examining the different roles that are enacted as part of the father status (e.g., nurturer, provider, disciplinarian) and their associated identities also may provide greater understanding of how different roles translate into identities and behavior.

Finally, greater attention needs to be paid to other factors that affect the relationship between identity and involvement. For example, exploring fathers' relationships with friends and family members, the remarriage of divorced parents, or perceived parental competence of the spouse/former spouse all would add to our knowledge. We believe that studying moderating effects will reveal the conditions that will allow us to tailor intervention strategies to best meet the needs of those served.

Table 1
Means and Standard Deviations for Demographic Variables
for the Total Sample and by Group

                                            Total
                                          (N = 279)

Variables                              Mean       (SD)

Number of children
  in household                          1.51      (1.16)
Number of biological
  children                              2.15       (.88)
Age of oldest child                    12.86      (6.98)
Average age of
  biological children                   9.83      (6.02)
Months married                        154.36     (81.41)
Father's age                           38.89      (6.89)
Months divorced                         --         --
Geographic proximity
  to child (in miles)                   --         --

                                        N          %

Household income
  [less than or equal to] $19,999     26           9.4
  $20,000-39,999                      91          33.0
  $40,000-59,999                      81          29.3
  e$60,000                            78          28.3

Occupation
  Professional                       141          51.3
  Clerical/sales                      34          12.4
  Service                             30          10.9
  Agricultural                         5           1.8
  Processing                           5           1.8
  Machine/trades                      16           5.8
  Benchwork                           12           4.4
  Structural                          15           5.5
  Miscellaneous                       17           6.2

Employment
  Full-time                          264          95.0
  Other                               14           5.0

Education
  High school or less                 44          15.8
  College                            139          50.0
  Graduate work                        9.5        34.2

Ethnicity
  White                                258        93.1
  Other                                1.9         6.9

                                           Married
                                          (n = 186)

Variables                              Mean       (SD)

Number of children
  in household                          2.07      (0.86)
Number of biological
  children                              2.19      (0.82)
Age of oldest child                    10.85      (6.79)
Average age of
  biological children                   8.95      (5.86)
Months married                        172.15     (81.02)
Father's age                           38.69      (6.72)
Months divorced                         --         --
Geographic proximity
  to child (in miles)                   --         --

                                        n          %

Household income
  [less than or equal to] $19,999       11         6.0
  $20,000-39,999                        58        31.5
  $40,000-59,999                        60        32.6
  e$60,000                              55        29.9

Occupation
  Professional                         103        56.0
  Clerical/sales                        19        10.3
  Service                               19        10.3
  Agricultural                           2         1.1
  Processing                             3         1.6
  Machine/trades                         8         4.3
  Benchwork                              7         3.8
  Structural                            12         6.5
  Miscellaneous                         11         6.0

Employment
  Full-time                            180        97.3
  Other                                  5         2.7

Education
  High school or less                   30        16.1
  College                               96        51.6
  Graduate work                         60        32.3

Ethnicity
  White                                175        94.1
  Other                                 11         5.9

                                           Divorced
                                           (n = 93)

Variables                              Mean       (SD)

Number of children
  in household                           .40      (0.81)
Number of biological
  children                              2.04      (0.95)
Age of oldest child                    14.01      (6.94)
Average age of
  biological children                  11.71      (5.96)
Months married                        118.76     (70.05)
Father's age                           39.33      (7.27)
Months divorced                        66.01     (55.70)
Geographic proximity
  to child (in miles)                 244.90    (620.82)

                                        n          %

Household income
  [less than or equal to] $19,999       15        16.3
  $20,000-39,999                        33        35.9
  $40,000-59,999                        21        22.8
  e$60,000                              23        25.0

Occupation
  Professional                          38        41.8
  Clerical/sales                        15        16.5
  Service                               11        12.1
  Agricultural                           3         3.3
  Processing                             2         2.2
  Machine/trades                         8         8.8
  Benchwork                              5         5.5
  Structural                             3         3.3
  Miscellaneous                          6         6.6

Employment
  Full-time                             84        90.3
  Other                                  9         9.7

Education
  High school or less                   14        15.2
  College                               43        46.7
  Graduate work                         35        38.0

Ethnicity
  White                                 83        91.2
  Other                                  8         8.8

Note.--Denotes that variable data were not collected
for indicated group.

Table 2
Factor Loadings of Identity Items with Two-Factor Solution

                                            Factor 1        Factor 2
Items                                    (Satisfaction)   (Investment)

1.20    Being a parent is a satisfying
        experience to some adults, but
        for other adults, being a
        parent is not all that
        satisfying                            .75             .12

1.21    Some mothers and fathers
        aren't sure they were suited
        to be parents, but parenting
        comes easily and naturally
        to other parents                      .75

1.18    Some mothers and fathers
        think that they are not very
        effective parents, but other
        mothers and fathers think they
        are pretty capable as parents         .73

1.16    For some parents, children
        mostly feel like a burden,
        but for other parents, their
        children are a main source
        of joy in their lives                 .72             .12

1.8     Some adults are more content
        being a parent than they ever
        thought possible, but for
        other adults, being a parent
        hasn't fulfilled them like
        they had hoped it would               .70

1.7     Some people feel they end up
        making too many sacrifices for
        their children, but for other
        parents, there are more
        rewards than sacrifices in
        rearing children                      .67             .15

1.6     Some parents often can't
        figure out what their children
        need or want, but other
        parents seem to have a knack
        for understanding what their
        children need or want                 .64

1.12    Some adults would hesitate to
        have children if they had it
        to do over again, but given
        the choice, other adults
        wouldn't think twice before
        having children                       .60

1.4     Some parents often wish they
        hadn't had children, but other
        parents rarely regret having
        had children                          .60

1.11    Some parents resent the fact
        that having children means
        less time to do the things
        they like, but other parents
        don't mind having less free
        time for themselves                   .58             .13

1.10    Some parents feel that they
        are doing a good job of
        providing for their children's
        needs, but other parents have
        doubts about how well they are
        meeting their children's needs        .57

1.2     Some parents have clear ideas
        about the right and wrong ways
        to rear children, but other
        parents have doubts about the
        way they are bringing up their
        children                              .50

1.9     Some parents don't think too
        much about how to parent; they
        just do it, but other parents
        try to learn as much as they
        can about how to parent                               .82

1.5     Some parents want to learn
        everything possible about
        being a parent, but other
        parents feel that they already
        know all they need to know
        about parenting                       .21             .77

1.1     Some parents do a lot of
        reading about how to be a good
        parent, but other parents
        don't spend much time reading
        about parenting                       .15             .72

1.13    Some parents feel it's a must
        to keep up with the latest
        childrearing advice and
        methods, but other parents
        would rather deal with their
        children on a day-to-day basis
        with what they already know                           .70

Note. Factor loadings < .10 have been omitted.

Table 3
Correlations Between Composite Measures (N = 279)

Variables                     1         2          3          4

1. Number of children
   in household              --       -.02       -.22 **    .27 **
2. Age of oldest child                 --         .10      -.14 **
3. Geographic proximity                           --       -.16 **
4. Identity satisfaction                                     --
5. Identity investment
6. Identity salience
7. Direct conflict
8. Indirect conflict
9. Cooperation
10. Marital status
11. Involvement

Range                      0-6        1-36     0-3500      2-4
Mean                       1.51      11.86      81.63      3.32
SD                         1.16       6.98     375.40       .42
Alpha                        --        --         --        .88

Variables                     5         6          7          8

1. Number of children
   in household             .10      -.06      -.37 **     -.12 *
2. Age of oldest child     -.07      -.14 **    .10         .02
3. Geographic proximity    -.05      -.02       .21 **      .16 **
4. Identity satisfaction    .11       .06      -.27 **     -.28 **
5. Identity investment       --       .23 **   -.16 **     -.07
6. Identity salience                   --       .05        -.05
7. Direct conflict                                --        .32 **
8. Indirect conflict                                         --
9. Cooperation
10. Marital status
11. Involvement

Range                      1-4       0-3       1-4         1-4
Mean                       2.40      0.37      2.19        1.37
SD                          .61       .69       .75         .57
Alpha                       .74        --       .92         .56

Variables                     9        10         11

1. Number of children
   in household             .33 **    .68 **    .37 **
2. Age of oldest child     -.11      -.21 **   -.26 **
3. Geographic proximity    -.32 **   -.31 **   -.36 **
4. Identity satisfaction    .45 **    .24 **    .41 **
5. Identity investment      .18 **    .05       .24 **
6. Identity salience        .10      -.13 *     .06
7. Direct conflict         -.47 **   -.52 **   -.38 **
8. Indirect conflict       -.37 **   -.14 *    -.29 **
9. Cooperation               --       .40 **    .60 **
10. Marital status                     --       .49 **
11. Involvement                                   --

Range                      1-5         --      1-5
Mean                       4.38        --      4.00
SD                          .5         --       .46
Alpha                       .77        --       .91

Note.--Denotes not calculated for this variable.

* p < .05. ** p < .01.

Table 4
Hierarchical Regressions for Father Involvement Outcome Variables
on Identity Measures: Full Model (N = 279)

Variables                                         B       [beta]     p

Block 1: Control variables
  Number of children in household               -.02      -.03      .59
  Age of oldest child                           -.01      -.12 **   .01
  Geographic proximity                          -.28      -.12 *    .04
  F                                            34.86 **
  Adjusted [R.sup.2]                             .27

Block 2: Independent variables
  Identity satisfaction                          .22       .13 *    .02
  Identity investment                            .20       .16 **   .00
  Identity salience                              .02       .02      .78
  Direct conflict                                .03       .03      .58
  Indirect conflict                             -.13      -.11 *    .03
  Cooperation                                    .31       .31 **   .00
  Marital status                                 .35       .22 **   .00
  F-change                                     18.40 **
  Adjusted [R.sup.2]                             .50

Block 3: Moderator variables
  Satisfaction x Direct conflict                -.35      -.16      .06
  Satisfaction x Indirect conflict              -.46      -.17 *    .02
  Satisfaction x Cooperation                    -.51      -.24 *    .02
  Investment x Direct conflict                  -.36      -.21 *    .03
  Investment x Indirect conflict                -.41      -.20 *    .02
  Investment x Cooperation                      -.51      -.32 **   .00
  Salience x Direct conflict                     .14       .12      .16
  Salience x Indirect conflict                   .25       .14      .10
  Salience x Cooperation                         .16       .13      .12
  F-change                                      2.25 *
  Adjusted [R.sup.2]                             .52

Block 4: Three-way interactions
  Marital x Satisfaction x Direct conflict       .22       .06      .43
  Marital x Satisfaction x Indirect conflict     .52       .14      .05
  Marital x Satisfaction x Cooperation           .41       .11      .15
  Marital x Investment x Direct conflict         .35       .14      .12
  Marital x Investment x Indirect conflict       .24       .08      .30
  Marital x Investment x Cooperation             .34       .15      .07
  Marital x Salience x Direct conflict          -.20      -.09      .33
  Marital x Salience x Indirect conflict        -.19      -.08      .37
  Marital x Salience x Cooperation              -.30      -.13      .08
  F-change                                      1.41
  Adjusted [R.sup.2]                             .53

* p < .05. ** p < .01.

Table 5
Hierarchical Regressions for Father Involvement Outcome Variables
on Identity Measures: Parsimonious Model (N = 279)

Variables                                         B       [beta]     p

Block 1: Control variables
  Age of oldest child                           -.O1      -.13 **   .01
  Geographic proximity                          -.36      -.15 **   .00
  F                                            40.51 **
  Adjusted [R.sup.2]                             .23

Block 2: Independent variables
  Identity satisfaction                          .18       .11 *    .03
  Identity investment                            .15       .12 **   .01
  Direct conflict                                .03       .03      .54
  Indirect conflict                             -.14      -.12 **   .01
  Cooperation                                    .37       .36 **   .00
  Marital status                                 .35       .20 **   .00
  F-change                                     26.08 **
  Adjusted [R.sup.2]                             .50

Block 3: Moderator variables
  Satisfaction x Direct conflict                -.22      -.10      .08
  Satisfaction x Indirect conflict              -.52      -.20 **   .00
  Satisfaction x Cooperation                    -.26      -.12 *    .04
  Investment x Direct conflict                  -.12      -.07      .18
  Investment x Indirect conflict                -.19      -.09      .06
  Investment x Cooperation                      -.36      -.23 **   .00
  F-change                                      3.04 **
  Adjusted [R.sup.2]                             .53

Block 4: Three-way interactions
  Marital x Satisfaction x Indirect conflict     .56       .15 *    .02
  Marital x Investment x Cooperation             .12       .05      .39
  F-change                                      3.15 *
  Adjusted [R.sup.2]                             .53

* p < .05. ** p < .01.

REFERENCES

Ahrons, C. (1981). The continuing interparental relationship between divorced spouses. American Journal of Orthopsychiatry, 5, 415-428.

Ahrons, C.R. (1983). Predictors of paternal involvement postdivorce: Mothers' and fathers' perceptions. Journal of Divorce, 6(3), 55-69.

Ahrons, C.R., & Tanner, J.L. (2003). Adult children and their fathers: Relationship changes 20 years after parental divorce. Family Relations, 52, 340-351.

Amato, P. (2000). The consequences of divorce for adults and children. Journal of Marriage and the Family, 62, 1269-1288.

Arendell, T. (1995). Fathers and divorce. Thousand Oaks, CA: Sage.

Booth, A., & Amato, P.R (1994). Parental marital quality, parental divorce, and relations with parents. Journal of Marriage and Family, 56, 21-34.

Braver, S.L., Wolchik, S.A., Sandler, I.N., & Sheets, V.L. (1993). A social exchange model of nonresidential parent involvement. In C.E. Depner & J.H. Bray (Eds.) Nonresidential parenting: New vistas in family living. (pp. 87-108). Thousand Oaks, CA: Sage.

Bruce, C., & Fox, G.L. (1999). Accounting for patterns of father involvement: Age of child, father-child coresidence, and father role salience. Sociological Inquiry, 69, 458-477.

Buehler, C., & Trotter, B. (1990). Nonresidential and residential parents' perceptions of the former spouse relationship and children's social competence following marital separation: Theory and programmed intervention. Family Relations, 39, 395-404.

Burke, P.J. (1991). Identity processes and social stress. American Sociological Review, 56, 836-849.

Burke, P.J., & Reitzes, D.C. (1981). The link between identity and role performance. Social Psychology Quarterly, 44, 83-92.

Burke, P.J., & Reitzes, D.C. (1991). An identity theory approach to commitment. Social Psychology Quarterly, 54, 239-251.

Cast, A. (2003). Identities and behavior. In P.J. Burke, T.J. Owens, R.T. Serpe & P.A. Thoits (Eds.) Advances in identity theory and research (pp. 41-56). New York: Kluwer Academic.

Dickie, J.R., & Gerber, S.C. (1980). Training in social competence: The effect on mothers, fathers, and infants. Child Development, 51, 1248-1251.

Dillman, D.A. (1978). Mail and telephone surveys. New York: John Wiley.

Dudley, J. (1991). Increasing our understanding of divorced fathers who have infrequent contact with their children. Family Relations, 40, 279-285.

Fox, G.L., & Blanton, P.W. (1995). Noncustodial fathers following divorce. Marriage and Family Review, 20, 257-282.

Fox, G.L., & Bruce, C. (2001). Conditional fatherhood: Identity theory and parental investment theory as alternative sources of explanation of fathering. Journal of Marriage and the Family, 63,394-404.

Geasler, M.J., & Blaisure, K.R. (1998). A review of divorce education program materials. Family Relations, 47, 167-175.

Harris, K.M., Furstenberg, F.F., Jr., & Marmer, J.K. (1998). Paternal involvement with adolescents in intact families: The influence of fathers over the life course. Demography, 35, 201-216.

Hernandez, D.J., & Brandon, P.D. (2002). Who are the fathers of today? In C.S. Tamis-LeMonda & N. Cabrera (Eds.) Handbook of father involvement (pp. 33-62). Mahwah, NJ: Lawrence Erlbaum Associates.

Ihinger-Tallman, M., Pasley, K., & Buehler, C. (1993). Developing a middle-range theory of father involvement postdivorce. Journal of Family Issues, 14, 550-571.

Jaccard, J., Turrisi, R., & Wan, C.K. (1990). Interaction effects in multiple regression. Newbury Park, CA: Sage.

Jacobson, D S. (1978). The impact of marital separation/divorce on children: II. Interparent hostility and child adjustment. Journal of Divorce, 2(1), 3-19.

Kelly, J.B. (2000). Children's adjustment in conflicted marriage and divorce: A decade review of research. Journal of the American Academy of Child and Adolescent Psychiatry, 39,963-973.

Kruk, E. (1994). The disengaged noncustodial father: Implications for social work practice with the divorced family. Social Work, 39(1), 15-26.

Kurdek, L. (1987). Children's adjustment to parental divorce: An ecological perspective. In J.P. Vincent (Ed.), Advances in family intervention, assessments, and theory. (Vol. 4, pp. 1-31). Greenwich, CT: JAI Press.

Lamb, M.E. (2000). The history of research on father involvement: An overview. Marriage and Family Review, 29, 23-42.

Maccoby, E.E., & Mnookin, R.H. (1992). Dividing the child: Social and legal dilemmas of custody. Cambridge, MA: Harvard University Press.

MacPhee, D., Benson, J.B., & Bullock, D. (1986, April). Influences on maternal self-perceptions. Paper presented at the biennial International Conference on Infant Studies, Los Angeles, CA.

Margolin, G., Gordis, E.B., & John, R.S. (2001). Coparenting: A link between marital conflict & parenting in two-parent families. Journal of Family Psychology, 15, 3-21.

Marsiglio, W., Amato, P., Day, R.D., & Lamb, M.E. (2000). Scholarship on fatherhood in the 1990s and beyond. Journal of Marriage and the Family, 62, 1173-191.

McBride, B.A. (1990). The effects of a parent education/play group program on father involvement in child rearing. Family Relations, 39, 250-256.

Minton, C., & Pasley, K. (1996). Fathers' parenting role identity and father involvement: A comparison of married and divorced, nonresident fathers. Journal of Family Issues, 17, 26-45.

Parke, R.D., & Buriel, R.B. (1998). Socialization in the family: Ethnic and ecological perspectives. In N. Eisenberg (Ed.), Handbook of child psychology (5th ed., pp. 463-552). New York: John Wiley.

Pasley, K. (1994). Reflections on men's lives following divorce and remarriage. Stepfamilies, 14(4), 12-13.

Pasley, K., & Braver, S.L. (2004). Measuring father involvement in divorced, nonresident fathers. In R. Day & M. Lamb (Eds.), Conceptualizing and measuring father involvement (pp. 217-240). Mahwah, NJ: Lawrence Erlbaum Associates.

Pasley, K., & Minton, C. (1997). Generative fathering after divorce and remarriage: Beyond the "disappearing dad." In A.J. Hawkins & D.C. Dollahite (Eds.), Generative fathering: Beyond deficit perspectives (pp. 118-133). Thousand Oaks, CA: Sage.

Pleck, J.H. (1997). Paternal involvement: Levels, sources and consequences. In Lamb, M.E. (Ed.), The role of the father in child development (3rd ed., pp. 66-103). New York: John Wiley & Sons, Inc.

Rands, M., Levinger, G., & Mellinger, G.D. (1981). Patterns of conflict resolution and marital satisfaction. Journal of Family Issues, 2, 297-321.

Rane, T.R., & McBride, B.A. (2000). Identity theory as a guide to understanding fathers' involvement with their children. Journal of Family Issues, 21,347-366.

Rettig, K.D., & Leichtentritt, R.D. (2001). Understanding non custodial fathers' relationships with children from resource theory perspectives. Journal of Divorce and Remarriage, 35(3/4), 1-22.

Seltzer, J.A. (1991). Relationships between fathers and children who live apart: The father's role after separation. Journal of Marriage and the Family, 53, 79-102.

Stone, G., & McKenry, P. (1998). Nonresidential father involvement: A test of a mid-range theory. Journal of Genetic Psychology, 159, 313-336.

Stryker, S. (1968). Identity salience and role performance: The relevance of symbolic interaction theory for family research. Journal of Marriage and the Family, 30, 558-564.

Stryker, S. (1987). Identity Theory: Developments and Extensions. In K. Yardley and T. Honess (Eds.), Self and identity: Psychosocial perspectives (pp. 89-103). New York: Wiley.

Stryker, S., & Serpe, R.T. (1994). Identity salience and psychological centrality: Equivalent, overlapping, or complementary concepts? Social Psychology Quarterly, 57, 16-36.

Whiteside, M.F. (1998). The parental alliance following divorce: An overview. Journal of Marital and Family Therapy, 24, 3-24.

Correspondence concerning this article should be sent to Kari Henley, P.O. Box 26170, Department of Human Development and Family Studies, University of North Carolina at Greensboro, Greensboro, NC 27402-6170. Electronic mail: [e-mail].

KARI HENLEY

University of North Carolina at Greensboro

KAY PASLEY

Florida State University