A Typology of Interactional Patterns Between Youth and Their Stepfathers: Associations with Family Relationship Quality and Youth Well-Being

Abstract

Stepfamilies are an increasingly common family form, many of which are headed by a resident mother and stepfather. Stepfather-child relationships exert notable influence on stepfamily stability and individual well-being. Although various stepfather roles have been observed, more research is warranted by which stepfather-child interactions are explored holistically and across a variety of life domains (e.g., recreational, personal, academic, and disciplinary). Thus, the primary purpose of the current study is to explore varying interactional patterns between youth and their stepfathers. A latent class analysis is conducted using a representative sample of 1,183 youth (53% female; mean age = 15.64 years, SD = 1.70 years; 62% non-Hispanic White) residing in mother-stepfather families from Wave I of the National Longitudinal Study of Adolescent to Adult Health. Latent-class enumeration processes support a four-class solution, with latent classes representing inactive, academically oriented, casually connected, and versatile and involved patterns of youth-stepparent interaction. Notable differences and similarities are evident across patterns with respect to family relationship quality, youth well-being, and sociodemographic characteristics. Differences are most stark between the inactive and versatile and involved patterns. Ultimately, the results showcase notable variation in youth-stepparent interactional patterns, and one size does not necessarily fit all stepfamilies. Family practitioners should be mindful of variation in youth-stepparent interactional patterns and assist stepfamilies in seeking out stepparent-child dynamics that are most compatible with the needs and dynamics of the larger family system.

Stepfamilies are one of the fastest growing family forms in the United States. Indeed, nearly one-third of youth are estimated to live with a biological parent and stepparent at some point before reaching legal adulthood (Bumpass, Raley, & Sweet, 1995; Pew Research Center, 2011). Stepfamilies form when at least one partner in a new committed relationship brings a child or children from a previous relationship (Ganong & Coleman, 2017). Consequently, stepfamilies embody notable complexity as they merge together new and existing dyadic relationships that often transcend single households.

The complexities of the stepfamily transition, as well as the transitions that precede it, can create significant demands for family members. Research has shown that members of stepfamilies often grapple with stepparent role ambiguity, coparenting conflict, stepcouple disagreements about parenting, shifts in economic and social resources, and opposing expectations among family members (Ganong & Coleman, 2017; Jensen & Shafer, 2013; Jensen, Shafer, & Larson, 2014; Papernow, 2013). Relationships between youth and stepparents can be particularly ambiguous, resulting in conflict, strain, or disengagement. Because the stepparent-child relationship is heralded as the crux of stepfamily stability and individual well-being (Ganong, Coleman, Fine, & Martin, 1999; Ganong, Coleman, & Jamison, 2011; Pryor, 2014), scholars and practitioners continue exploring its development, variation, and impact.

Although various roles that stepparents assume amid and following the transition to stepfamily life have been observed, largely in clinical and qualitative contexts, more research is warranted by which stepparent-child interactions are explored holistically and across a variety of specific life domains (e.g., recreational, personal, academic, and disciplinary). Moreover, associations between emergent youth-stepparent interactional patterns and family and individual characteristics should be assessed in an effort to understand the contexts in which specific interactional patterns are likely to emerge, as well as the potential implications of those patterns. Thus, the primary purpose of the current study is to explore varying interactional patterns between youth and their stepfathers. Rather than examining the frequency of single forms of interactions and assessing their distinct influence on youth and family outcomes, a holistic analysis of various forms or types of youth-stepfather interactions could illuminate common interactional patterns, the totality of which might have meaningful implications for youth and family outcomes. Thus, the primary purpose of the current study is coupled with efforts to assess the construct validity of the interactional patterns identified with respect to the quality of family relationships, youth well-being, and socio-demographic characteristics.

Interactional Patterns and Stepparent Roles in Stepfamilies

As noted, investigations and descriptions of interactional patterns and stepparent roles in stepfamilies, particularly from a typological perspective, have emerged largely in clinical and qualitative literatures. Some studies have focused on the development of stepfamily relationships over time, highlighting variation in the types and timing of stepparent-child relationship quality and development (Ganong, Coleman, & Jamison, 2011; Kinniburgh-White, Cartwright, & Seymour 2010; Papernow, 2013). Others have focused on typologies of stepfamily-level dynamics and communication (Schrodt, 2006), and the quality of the coparenting relationship between resident and nonresident biological parents (Pryor, 2014).

More relevant to the current study is research exploring features of the stepparent-child relationship and the types of stepparenting employed among stepfathers and stepmothers. In one qualitative study of Isreali stepfamilies, Erera-Weatherly (1996) identified five stepparent styles. The “birth parent style” is marked by stepparents’ view that parenthood and stepparenthood are the same, and that children and stepchildren should be treated equitably. The “super good step-mom” style reflects stepmothers’ attempts to excel as stepparents in an effort to win the affections of their stepchildren. The “detached” style reflects stepparents’ minimal involvement in the lives of their stepchildren. The “uncertain” style embodies stepparents’ ambivalence and confusion with respect to their role as a stepparent. Lastly, the “friendship” style reflects stepparents’ belief that they ought to befriend and accept their stepchildren, rather than exert authority.

In another qualitative study, Weaver and Coleman (2005) identified a variety of role identities among nonresident stepmothers (i.e., stepmothers whose stepchildren do not share primary residence). Stepmother identities included being a responsive and caring adult, friend, provider of emotional support, mentor, facilitator of other relationships, and involved outsider, among others. Relatedly, Crohn (2006) explored styles of stepmothering from the stepchildren’s perspective, identifying the following styles: “disengaged,” “peer-like,” “an older close friend,” “a type of kin,” and “like another mother” (Crohn, 2006).

There exists a relative dearth of quantitative typological analyses of interactional patterns, or relationship/parenting types, involving youth and their stepparents; however, there are at least two notable exceptions. For one, Fine and Kurdek (1992) examined the role of varying combinations of the support and control dimensions of parenting (Baumrind, 1971; Maccoby & Martin, 1983) in the context of stepfamily life. Crosbie-Burnett and Giles-Sims (1994) built on this work and proposed revised labels for some parenting styles to better reflect the unique dynamics of stepfamily relationships in general and the stepparent role in particular; they examined the role of the following stepparenting styles: authoritative (i.e., high control and high support), authoritarian (i.e., high control and low support), supportive (revised label; i.e., low control and high support), and disengaged (revised label; i.e., low control and low support).

Taken together, past research highlights wide variation with respect to the roles stepparents assume in the lives of their stepchildren. Undergirding this variation could be the affinity-seeking interactions stepparents employ. Affinity-seeking strategies consist of active processes, both verbal and nonverbal, by which individuals aim to build and strengthen relationships with others (Daly & Kreiser, 1994). Ganong and colleagues (1999) found that stepparents tend to adopt one of three affinity-seeking patterns: “continuous affinity-seekers,” or stepparents who continually strive to build affinity with their stepchildren; “early affinity-seekers,” or stepparents who strive to garner their stepchildren’s approval early on but then cease their efforts once the new romantic partnership is secured; and “nonseekers,” or stepparents who make relatively few attempts to generate affinity with their stepchildren.

Among stepparents who actively engage with their stepchildren, there are a variety of domains in which affinity-seeking can take place (e.g., recreational, personal, academic, and disciplinary). Affinity-seeking efforts, in part, can be indicated by the type and quantity of interactions in which stepparents and youth engage (although the precise motivations on the part of the stepparent might not be apparent depending on how interactions are measured). Moreover, singular interactions, or types of interactions, might not occur in isolation. More likely, a variety of types and quantities of interactions occur between youth and their stepparents. For example, some stepparents might engage with their stepchildren in several academically oriented tasks or conversations, whereas other stepparents might engage with their stepchildren in sports-related activities. Other stepparents might engage in both sets of activities or interactions, whereas other stepparents might be disengaged entirely. Importantly, our understanding of common and specific interactional patterns between youth and their stepparents is limited.

It is important to note that stepfamily dynamics can vary with respect to the socio-demographic characteristics of stepfamilies and individuals. Nearly 80% of all stepfamilies in the United States are reared by a mother and stepfather (Kreider & Ellis, 2011). Thus, youth are more likely to reside with a stepfather than a stepmother, and the current study focuses on youths’ interactions with resident stepfathers. Past research has commonly highlighted the “gendered” nature by which stepfamily life is experienced (Ganong & Coleman, 2017). Ideologies and social scripts embedded within Western cultures, such as those pertaining to expectations regarding fathering and mothering, can influence how men and women experience a transition into a stepparent role. Whereas mothers are often (unrealistically) expected to focus exclusively or primarily on parenting, providing emotional support, and cultivating or maintaining positive family relationships; fathers are often expected to provide instrumental or financial support and assume a disciplinarian role in the family (Ganong & Coleman, 2017). These gendered societal and cultural pressures exerted on new stepparents can shape stepfamily dynamics. Thus, explorations of interactional patterns between youth and their stepparents should attend to stepfathers and stepmothers distinctly.

In addition to stepparent gender, there is some evidence that male youth tend to adapt to stepfamily life more quickly and favorably than their female counterparts, particularly in mother-stepfather families (Jensen & Howard, 2015). Older youth, such as those in their adolescent years, are more likely than their younger counterparts to disengage from or resist the influence of their stepparents as they seek to further their autonomy or maintain family processes that existed prior to the formation of the stepfamily (Ganong & Coleman, 2017). Racial/ethnic minority groups, such as African American and Hispanic families, also have been observed by clinicians to exhibit unique strengths that aid families during the transition to stepfamily life (Papernow, 2013). Families with greater socio-economic resources, such as education and income, might experience the stepfamily transition with greater ease, or at the least, experience fewer stressors in other domains of family life. Stepfamily processes might also be influenced by the number of precursory family transitions, the number of individuals residing in the household, and the duration of the stepfamily (Papernow, 2013; Sweeney, 2010). Thus, the current study included a focus on youth and family socio-demographic characteristics as possible correlates of emergent youth-stepparent interactional patterns.

Family and Individual Correlates of Youth-Stepparent Interactional Patterns

Consistent with a family systems perspective (Cox & Paley, 1997), the type and quantities of interactions between youth and their stepparents are likely associated with the quality of stepfamily relationships, particularly the stepparent-child relationship. Rather than viewing interactional patterns between youth and their stepparents as a direct indicator of relationship quality, it seems more appropriate to conceptualize interactional patterns as a precursor to relationship quality (although interactional patterns and relationship quality are likely interdependent and bidirectional in their influence). In other words, consistent with an affinity-seeking perspective, what youth and stepparents do together, and the domains in which they share their lives, could influence the quality of the stepparent-child relationship. For example, past research has identified a positive linear association between stepfather involvement and stepfather-child relationship quality (Jensen & Pace, 2016). Moreover, relationship-maintaining behaviors and interactions between youth and their stepparents have been linked to the development of stronger stepparent-child relationships (Ganong et al., 2011).

Interactional patterns between youth and their stepparents also are likely associated with other family relationships (Coleman, Ganong, & Russell, 2013; Jensen & Harris, 2017a; King, Thorsen, & Amato, 2014). For example, situations in which stepparents and youth arrive at a mutually satisfying interactional pattern could foster higher quality parent-child relationships, as youth would be less likely to feel resentful toward their parent following the formation of the stepfamily. Moreover, youth who maintain good communication and positive relationships with their resident parents are more likely to engage new stepparents and form closer stepparent-child relationships (Jensen & Shafer, 2013).

Interactional patterns between youth and their stepparents could also be associated with the nonresident parent-child relationship. In stepfamilies, youth can either augment, reduce, retain, or substitute portions of their parental network when acquiring a new stepparent (Ganong & Coleman, 2017; Gross, 1987). Such shifts in youths’ parental network and sources of interactions could have implications for the quality of the nonresident parent-child relationship. In addition, youth who are disconnected from a nonresident parent might be better positioned or more willing to accommodate the development of richer interactional patterns with new stepparents.

Because some resident parents serve as gatekeepers between their children and new romantic partners, the quality of the stepcouple relationship might increase in the context of interactional patterns between youth and their stepparents that are perceived by parents as acceptable or desirable (Ganong & Coleman, 2017; Papernow, 2013). Conversely, youth-stepparent interactional patterns perceived as undesirable by parents might stir conflict in the stepcouple relationship. In the context of a highly conflictual stepcouple relationship, youth might also be reluctant to engage with their resident stepparent (Jensen & Shafer, 2013).

Given existing research, it is unclear how specific interactional patterns between youth and their stepparents would be associated with youth well-being directly. More common are studies in which stepparent-child relationship quality (i.e., closeness, relationship satisfaction) is examined as a correlate of youth adjustment. In this context, higher quality stepparent-child relationships have been associated with higher levels of youth psychological, behavioral, academic, and physical well-being (Jensen & Harris, 2017b; Jensen, Lippold, Mills-Koonce, & Fosco, 2016; King, 2006). In addition, supportive behaviors on the part of the stepparent have been associated with lower levels of internalizing and externalizing among youth (Crosbie-Burnett & Giles-Sims, 1994; Papernow, 2013). Thus, it is reasonable to posit an association between youth-stepparent interactional patterns and youth well-being, such that patterns marked by greater involvement across various life domains could confer on youth positive adjustment outcomes. Perhaps even more reasonable is to posit that associations between youth-stepparent interactional patterns and youth well-being are bidirectional. For example, youth with significant adjustment problems might disengage from their stepparents or discourage stepparents’ efforts to interact and build affinity (e.g., King, Amato, & Lindstrom, 2015).

Another important consideration is that stepparents, on average, tend to engage less with children relative to biological parents. This is often the result of stepparents’ “role ambiguity, confusion about boundaries, rejection by adolescents, lack of biological and psychological ties, and/or lack of legal parent rights and responsibilities” (Crosbie-Burnett & Giles-Sims, 1994, p. 395). Past research has also shown that youth and family outcomes are optimized when stepparents avoid exerting discipline or control, especially early on in stepfamily development and in stepfamilies with adolescent youth (Crosbie-Burnett & Giles-Sims, 1994; Papernow, 2013). Thus, variation in interactional patterns between youth and their stepparents is a warranted area of ongoing and holistic investigation, along with a greater understanding of its association with other family and individual outcomes or characteristics. Moreover, the interactional patterns between youth and their stepparents that might emerge should not be subjected to the same moral and value judgments that are often imposed on parents with biological or life-time ties to their children (Crosbie-Burnett & Giles-Sims, 1994). As shown in previous research, stepparents can assume one of many different roles in the lives of their stepchildren, and one size does not necessarily fit all stepfamilies. The current study centers on the following two research questions:

1. What specific patterns of youth-stepfather interactions exist in a representative sample of youth residing with a biological mother and stepfather?

2. In what ways are disparate patterns of youth-stepfather interactions associated with stepfamily relationship quality, youth well-being, and socio-demographic characteristics?

Methods

Data and Sample

Data for the current study came from the Wave I, in-home youth interviews and parent questionnaires from the National Longitudinal Study of Adolescent to Adult Health (Add Health). Respondents for in-home interviews were randomly selected from a nationally representative in-school sampling frame of adolescents, which served as the initial source of data for the Add Health Study. In-home interviews with youth were conducted using laptop computers and included questions about youths’ relationships with family and peers, features of the neighborhood and other contexts, health behaviors, and various well-being indicators. Parent data at Wave I were collected using interviewer-assisted, op-scanned questionnaires that were issued primarily to resident mothers. The questionnaires included items about parents’ romantic relationships and socioeconomic characteristics, among other features pertaining to their children and larger social contexts.

A total of 20,745 adolescents in grades 7 through 12 during the 1994 to 1995 school year comprised the Wave I sample, which served as a nationally representative sample of youth in those grades during those years. The analytical sample for the current study was limited to youth who were residing primarily with a biological mother and stepfather during Wave I. Because stepfamily dynamics are influenced by the pathways to stepfamily life (e.g., parental divorce or parental death; Ganong & Coleman, 2017), the sample was also limited to include youth who had living nonresident fathers. The final analytical sample included 1,183 youth (53% female, mean age = 15.64 years, SD = 1.70 years). Roughly 62% of youth identified as non-Hispanic White, 19% as non-Hispanic Black, 14% as Hispanic, and 4% as Asian, Native American, or another racial/ethnic identity. Nearly 74% of parents reported being married to the stepparent (as opposed to unmarried cohabitation [12%] or missing responses [14%]). Average household income was $50,016, and average stepfamily duration (i.e., time the stepparent has resided in the household) was 6.73 years with notable variation (SD = 4.11 years). The modal level of mothers’ education was the completion of some college (37%); the modal level of stepfathers’ education was the completion of high school (34%). Levels of nonresident father involvement varied—in the past 12 months, 52% of youth indicated they did not stay overnight at all with their nonresident fathers; 29% stayed over night once/twice or several times; 7% stayed overnight about once a month; 12% stayed overnight about once a week or more.

Measures

Youth-Stepfather Interactions

Information about interactions between youth and their stepfathers was measured with 11 binary items, representing recreational, personal, academic, and disciplinary domains. Specifically, youth were asked to indicate whether they had engaged with their stepfathers in any of the following activities during the past four weeks (1 = Yes, 0 = No): (a) gone shopping, (b) played a sport, (c) gone to a religious service or church-related event, (d) talked about someone they were dating or a party they went to, (e) gone to a movie/play/museum/or concert/sports event, (f) had a talk about a personal problem they were having, (g) had a serious argument about their behavior, (h) talked about their school work or grades, (i) worked on a project for school, and (j) talked about things they were doing in school. Another item allowed youth to specify if they had engaged in none of these activities during the past four weeks. In this context, youth-stepfather interaction was used as a broad term to reflect an area of shared activity or some form of dyadic engagement. Notably, the interaction items did not capture information about the depth, quality, or nature of dyadic engagement between youth and their stepfathers, but simply whether the activity or form of engagement had occurred in the past four weeks as perceived by youth. The results of the current study, and the patterns of interactions that emerged from the analysis, should be interpreted accordingly.

Family Relationship Quality

In an effort to generate evidence for the construct validity of the latent classes derived from the youth-stepfather interaction items, measures of relationship quality across the following four relationships were used: stepfather-child, mother-child, nonresident father-child, and stepcouple. Stepfather-child relationship quality was a five-item scale (α = .87), with items asking youth how close they felt to their stepfathers, how much they thought their stepfathers cared about them, whether their stepfather was warm and loving, how satisfied they were with communication, and how satisfied they were with the relationship overall. Response options ranged from 1 (“strongly disagree”/”not at all”) to 5 (“strongly agree”/”very much”). Thus, higher values were indicative of higher relationship quality. Mother-child relationship quality was measured with the same five items as the stepfather-child relationship quality scale, only items were worded to describe the mother-child relationship (α = .86). Nonresident father-child involvement (α = .83) was measured with three items, two of which asked youth to indicate how often in the last 12 months they stayed overnight with their biological father, and talked to him in person, on the phone, or received a letter from him. Response options for these two items ranged from 0 (“not at all”) to 5 (“more than once a week”). The remaining item asked youth to indicate how close they felt to their biological father, with response options ranging from 1 (“not close at all”) to 5 (“extremely close”). Higher values were indicative of a more involved relationship. Stepcouple relationship quality was measured with one item, which asked mothers to indicate their degree of relational happiness with their partner (rating from 1 = “completely unhappy” to 10 = “completely happy”).

Youth-Mother Interactions

Emergent latent classes were also contrasted in terms of youths’ interactions with their resident mothers. Using the same 10 interaction items used to measure youth-stepfather interactions, a count variable was created representing the number of interactions that youth endorsed having with their resident mothers in the past four weeks.

Youth Well-Being

Measures of youth depression, delinquency, and self-esteem were used to assess youth well-being. Youth depression was measured with a 9-item version of the Center for Epidemiologic Studies Depression Scale (α = .80; Radloff, 1977), allowing youth to indicate how often they experienced several depressive symptoms in the past week (e.g., felt depressed, felt sad, felt too tired to do things). Response options ranged from 0 (“never or rarely”) to 3 (“most or all of the time”). Thus, higher values were indicative of higher levels of depression. Youth delinquency was an 8-item scale (α = .74), allowing youth to indicate how often in the past year they engaged in various delinquent behaviors (e.g., deliberately damaged the property of another, stole items, hurt someone in a fight, threatened others). Response options ranged from 0 (“never”) to 3 (“5 or more times”); higher values were indicative of higher levels of delinquency. Youth self-esteem was a 6-item scale (α = .85), allowing youth to indicate the extent to which they agreed or disagreed with statements such as “you feel loved and wanted,” “you feel socially accepted,” “you have a lot of good qualities,” and “you like yourself just the way you are.” Response options for these items ranged from 1 (“strongly agree”) to 5 (“strongly disagree”), and were reverse coded such that higher values were indicative of higher levels of self-esteem.

Socio-Demographic Characteristics

Additional socio-demographic items were used for latent-class comparisons. These included youth sex (female [1], male [0]), youth age (continuous measure in years), youth racial/ethnic identity (binary variables representing non-Hispanic White, non-Hispanic Black, Hispanic, and Asian/Native American/Other), mother’s education (less than high school [1], completed high school/GED [2], some college [3], college degree or more [4]), stepfather’s education (coded the same as mother’s education), household income (continuous measure in thousand-dollar units), parental marital status (married [1], unmarried cohabiting [0]), mother’s past romantic relationships in the last 18 years (continuous measure), stepfamily duration (continuous measure in years), and household composition (continuous measure representing the number of household residents).

Analysis

To begin, the proportion of responses across all youth-stepfather interaction items was investigated. Then, because interaction items were binary, latent class analysis was used to explore the presence of unobserved subgroups that clustered around similar item-response patterns (Collins & Lanza, 2010). A variety of latent-class solutions (i.e., varying sets of specified classes) were assessed in an effort to identify the optimal or best-fitting solution. To evaluate relative fit across solutions, the following criteria or indices were examined: Bayesian Information Criterion (BIC), bootstrap likelihood ratio tests (BS LRT), entropy, and mean posterior probabilities (Nagin & Odgers, 2010; Nylan-Gibson & Masyn, 2016). These indices signal to the analyst which number of extracted classes appears to fit the data best, and highlights the amount of precision and accuracy with which a chosen solution partitions respondents into separate classes. Per methodological recommendations, large sets of random starts were specified for all analyses to avoid model estimations derived from local log-likelihood maxima, which can produce misleading or anomalous findings (Muthén & Muthén, 2012). Following the identification of the optimal or best-fitting solution, the three-step procedure was used to estimate covariate-mean differences across latent classes, both in terms of raw means and standardized Z scores (i.e., sample mean set to 0 with a standard deviation of 1; Bauer & Shanahan, 2007). The three-step procedure is an effective method for handling classification uncertainty, as it simultaneously handles the process of extracting latent classes, assigning respondents to their most likely class, and assessing covariate differences across classes (Asparouhov & Muthén, 2014). These analyses served as a form of construct validation with respect to the latent classes. All analyses were conducted in Mplus 7.4. Sampling weights were utilized to generate representative model parameters.

Results

Youth-Stepfather Interaction

Table 1 displays the proportion (weighted) of item endorsements across each youth-stepfather interaction item. The most frequently endorsed item was talking to stepfathers about school work or grades. Nearly 42% of youth indicated they had engaged in this type of interaction in the past four weeks. Next, roughly 38% of youth indicated they had talked with their stepfathers about “other things” they were doing in school. Thirty percent of youth indicated they had talked to their stepfathers about someone they were dating or a party they attended. In terms of the least frequently endorsed items, only 10% of youth indicated they had worked with their stepfathers on a project for school. In addition, only 19% of youth indicated they had engaged in none of the 10 listed interactions in the past four weeks. See Table 1 for more details.

Table 1.

Endorsement of Youth-Stepfather Interaction Items in Full Sample and Across Latent Classes

	Proportion of “Yes” responses in full sample	Latent Class
		1	2	3	4
		Inactive	Academically Oriented	Casually Connected	Versatile and Involved
Probability of membership		0.19	0.38	0.31	0.12
Conditional probability of a “Yes” response
Gone shopping	24.83%	0.00	0.17	0.34	0.60
Played a sport	23.87%	0.00	0.20	0.25	0.65
Gone to a religious service or church-related event	19.42%	0.00	0.15	0.25	0.46
Talked about someone you’re dating or a party you went to	29.93%	0.00	0.36	0.28	0.60
Gone to a movie, play, museum, or concert, or sports event	17.06%	0.00	0.15	0.15	0.50
Had a talk about a personal problem you were having	18.05%	0.00	0.13	0.18	0.56
Had a serious argument about your behavior	25.89%	0.00	0.31	0.28	0.44
Talked about your school work or grades	41.94%	0.00	0.79	0.01	0.87
Worked on a project for school	10.34%	0.00	0.09	0.03	0.44
Talked about other things you’re doing in school	38.42%	0.00	0.67	0.03	0.89
None	18.92%	1.00	0.00	0.00	0.00

Note: Conditional probabilities valued at .50 or higher are bolded to facilitate interpretation. Item-endorsement estimates were derived from weighted data.

Latent Class Enumeration and Solution

Table 2 displays results associated with the fit of each evaluated model and the class enumeration process. Beginning with only one class specified, BIC values steadily decreased until the four-class specification, following which BIC values began to increase. Moreover, non-significant BS LRTs indicated that each k set of classes might be preferable over the k – 1 set of classes (up to a total of 6 classes). Notably, the five-class solution yielded a class with only 21 youth, representing less than 1.8% of the full sample. This could indicate an over-extracted solution (Petras & Masyn, 2010). Taken together, the four-class solution was selected as optimal.

Table 2.

Model Fit and Class Enumeration

Classes	BIC	BS LRT (p-value)	Entropy	smallest n	Mean posterior probabilities
					1	2	3	4	5	6
1	13972.17
2	12557.65	0.00	1.00	224	1.00	1.00
3	12180.08	0.00	0.86	224	0.94	1.00	0.92
4	12117.25	0.00	0.81	142	1.00	0.87	0.91	0.84
5	12133.31	0.00	0.86	21	0.90	0.88	0.89	0.83	1.00
6	12168.95	0.00	0.80	21	0.80	0.71	1.00	0.80	0.87	0.89

Note: BIC = Bayesian Information Criterion; LRT = Likelihood ratio test; aLRT = adjusted LRT; BS LRT = Bootstrap LRT. Bold items correspond with the solution selected as optimal.

The four-class solution yielded an entropy value of .81 and mean posterior probability values ranging from .84 to 1.00, indicating acceptable class separation (Asparouhov & Muthén, 2014). Table 1 displays the probabilities of latent-class membership, as well as conditional item-endorsement probabilities associated with each latent class. Class 1 (19%), the inactive pattern, was marked by zero-probability endorsements across all interaction items; youth in this class had a 100% probability of endorsing the item indicating they had engaged in none of the listed activities with their stepfather in the past four weeks. Class 2 (38%), the academically oriented pattern, was marked by relatively high probabilities of youth endorsing items indicating they had discussed school-related activities and performance with their stepfathers. Class 3 (31%), the casually connected pattern, was marked by low-to-moderate probabilities of youth endorsing the interaction items pertaining largely to recreational, personal, and disciplinary domains. Class 4 (12%), the versatile and involved pattern, was marked by moderate-to-high probabilities of youth endorsing each of the interaction items. To aid in the interpretation of the four-class solution, Figure 1 displays the conditional probabilities of item endorsement visually.

Figure 1.

Latent-Class Solution

Note: Values along the y-axis indicate the probability of a “Yes” response to each youth-stepfather interaction item, conditional on latent class membership.

Latent Class Differences

Family Relationship Quality

Table 3 displays substantive differences across each latent class or pattern in terms of raw means as well as standardized Z scores (i.e., standard deviation units). Youth and mothers in the versatile and involved pattern reported the highest levels of stepfather-child relationship quality (Z = .72), mother-child relationship quality (Z = .33), nonresident father-child involvement (Z = .32), and stepcouple relationship quality (Z = .29), with most levels being significantly different from the other three classes (see Table 3 for specific between-class mean difference tests). The quality of stepfather-child (Z = −.87), mother-child (Z = −.33), and stepcouple (Z = −.35) relationships was relatively lowest within the inactive pattern. The level of nonresident father-child involvement, however, was lowest in the academically oriented pattern (Z = −.11). It is worth noting also that stepfather-child, mother-child, and stepcouple relationship quality was near or above average in the academically oriented pattern, but near or below average in the casually connected pattern. Moreover, nonresident father-child involvement was above average in the inactive and casually connected patterns.

Table 3.

Latent-Class Differences

Covariate	Full Sample (N = 1,183)		1 Inactive (19%) (n = 224)			2 Academically Oriented (38%) (n = 446)			3 Casually Connected (31%) (n = 371)			4 Versatile and Involved (12%) (n = 142)			Class differences, p ≤ .05
	M	mean Z score	M	SE	mean Z score	M	SE	mean Z score	M	SE	mean Z score	M	SE	mean Z score
Relationship quality
Stepfather-child relationship qualitya	3.83	0.00	3.04	(0.08)	−0.87	4.04	(0.06)	0.23	3.71	(0.08)	−0.14	4.49	(0.05)	0.72	4 > 1, 2, 3; 3 > 1; 2 > 1, 3
Mother-child relationship qualitya	4.41	0.00	4.20	(0.07)	−0.33	4.53	(0.03)	0.19	4.37	(0.06)	−0.06	4.62	(0.05)	0.33	4 > 1, 3; 3 > 1; 2 > 1, 3
Nonresident father-child involvementb	2.17	0.00	2.44	(0.14)	0.19	2.03	(0.10)	−0.11	2.34	(0.11)	0.12	2.62	(0.20)	0.32	4 > 2; 3 > 2; 1 > 2
Stepcouple relationship qualityc	8.50	0.00	7.92	(0.20)	−0.35	8.55	(0.11)	0.03	8.41	(0.16)	−0.05	8.99	(0.16)	0.29	4 > 1, 2, 3; 3 >1; 2 > 1
Youth-mother interactions (count)	4.30	0.00	3.25	(0.15)	−0.54	4.57	(0.11)	0.14	3.31	(0.14)	−0.51	6.60	(0.18)	1.18	4 > 1, 2, 3; 2 > 1, 3
Youth well-being
Youth depressiond	0.70	0.00	0.74	(0.03)	0.08	0.66	(0.04)	−0.07	0.72	(0.04)	0.05	0.70	(0.11)	0.00
Youth delinquencyd	0.19	0.00	0.21	(0.03)	0.08	0.14	(0.02)	−0.14	0.21	(0.03)	0.07	0.18	(0.04)	−0.02	1, 3 > 2
Youth self-esteema	4.08	0.00	3.92	(0.05)	−0.26	4.11	(0.05)	0.06	4.00	(0.05)	−0.13	4.23	(0.08)	0.25	4, 2 > 1; 4 > 3
Socio-demographic characteristics
Youth is female	0.53		0.50	(0.05)		0.56	(0.04)		0.49	(0.04)		0.45	(0.06)
Youth age	15.64	0.00	15.62	(0.18)	−0.01	15.55	(0.18)	−0.06	15.48	(0.19)	−0.09	14.87	(0.26)	−0.45	1, 2, 3 > 4
Youth racial/ethnic identity
White	0.62		0.74	(0.05)		0.79	(0.03)		0.70	(0.04)		0.67	(0.08)		2 > 3
Black	0.19		0.12	(0.03)		0.09	(0.02)		0.16	(0.03)		0.20	(0.07)		3 > 2
Hispanic	0.14		0.12	(0.03)		0.09	(0.02)		0.10	(0.03)		0.07	(0.03)
Asian/Native American/Other	0.04		0.01	(0.01)		0.04	(0.02)		0.03	(0.01)		0.04	(0.03)
Mother education	2.57	0.00	2.47	(0.13)	−0.11	2.61	(0.07)	0.04	2.51	(0.08)	−0.06	2.71	(0.14)	0.15
Stepfather education	2.59	0.00	2.41	(0.10)	−0.18	2.73	(0.08)	0.15	2.39	(0.09)	−0.21	2.74	(0.14)	0.16	4, 2 > 1, 3
Household income (in thousands)	50.02	0.00	50.27	(4.37)	0.01	46.64	(2.90)	−0.07	42.97	(2.86)	−0.15	75.38	(11.65)	0.55	4 > 1, 2, 3
Married (vs. cohabiting)	0.87		0.88	(0.03)		0.90	(0.03)		0.98	(0.02)		0.85	(0.06)
Mother’s past relationships	2.05	0.00	2.08	(0.10)	0.03	2.14	(0.05)	0.12	2.09	(0.05)	0.05	2.06	(0.09)	0.02
Stepfamily duration	6.73	0.00	6.15	(0.40)	−0.14	6.73	(0.33)	0.00	7.04	(0.34)	0.08	5.70	(0.56)	−0.25	3 > 4
Household composition	3.77	0.00	3.53	(0.15)	−0.16	3.59	(0.12)	−0.12	3.79	(0.11)	0.01	3.76	(0.26)	−0.01

Note: Means and mean differences were estimated using the 3-step procedure. Estimates were derived from weighted data and standard errors were adjusted for clustering. Means represent class-specific proportions for binary/dummy variables. Mean Z scores are only presented for continuous items.

^aRange: 1 to 5.

^bRange: .33 to 5.

^cRange: 1 to 10.

^dRange: 0 to 3.

Youth-Mother Interactions

Turning to youths’ interactions with their resident mothers, latent classes significantly differed. Youth reported the highest count of interactions with their mothers in the versatile and involved pattern—more than a standard deviation above the sample mean (Z = 1.18). Youth in the academically oriented pattern yielded the next highest level of youth-mother interactions (Z = .14). The inactive and casually connected patterns yielded similarly low levels of youth-mother interactions (Z = −.54 and −.51, respectively).

Youth Well-Being

In terms of youth depression, the latent classes did not significantly differ. Youth delinquency was highest among youth in the inactive (Z = .08) and casually connected (Z = .07) patterns—a significant difference from youth in the academically oriented pattern, who reported the lowest levels of delinquency (Z = −.14); youth in the versatile and involved pattern reported average levels of delinquency. Levels of youth self-esteem, on the other hand, were highest within the versatile and involved pattern (Z = .25), followed by the academically oriented pattern (Z = .06). Levels of self-esteem were lowest among youth in the inactive pattern (Z = −.26), followed by youth in the casually connected pattern (Z = −.13).

Socio-Demographic Characteristics

Table 3 also displays socio-demographic differences across latent classes. Socioeconomic differences were particularly apparent. For one, the versatile and involved pattern was marked by an especially above-average level of household income ($75,384; Z = .55), which was significantly higher than the other three classes. Moreover, levels of both mothers’ and stepfathers’ education were highest in the versatile and involved pattern (Z = .15 and .16, respectively), followed by the academically oriented pattern (Z = .04 and .15, respectively). Parental education was below average in the inactive and casually connected patterns. In terms of race, youth identifying as non-Hispanic White were over-represented in the academically oriented pattern relative to the casually connected pattern. Youth identifying as non-Hispanic Black, on the other hand, were under-represented in the academically oriented pattern. With respect to stepfamily duration, the versatile and involved pattern was marked by the lowest average duration (5.70 years; Z = −.25), whereas the casually connected pattern was marked by the longest average duration (7.04 years; Z = .08). Lastly, youth in the versatile and involved pattern (14.87 years; Z = −.45) were younger than the sample-average and significantly younger than youth in the other three patterns.

Discussion

The primary purpose of the current study was to explore holistically the variation in interactional patterns between youth and their stepfathers. The construct validity of emergent patterns was also investigated with respect to several family and individual characteristics, although links between latent-class membership and covariates are correlational, and should not be interpreted as causal (interpretations below speculate directionality, but are not intended to assert causality). Foremost, the results showcase notable variation in the types and quantities of interactions in which youth and their stepfathers might engage.

The most prevalent interactional pattern among stepfamilies in the sample is the academically oriented pattern. Youth and stepfathers in this pattern have a high probability of engaging in conversations related to school work, grades, and other school-related activities. Overall, residential family relationships in this pattern do not appear strained, although youth do report some relative distance with their nonresident fathers. Thus, stepfathers in this pattern might step in to support youth in their academic efforts and attempt to build affinity. These stepfathers appear well suited for this role, as they possess above-average levels of education. Youth in this pattern also report below-average levels of delinquency, which could further facilitate positive interactions centered on school-related activities. The dynamics of the academically oriented pattern appear to reflect, at least in part, the “friendship” and “supportive” styles of stepparenting previously identified, as stepparents in these families likely care about their stepchildren while not necessarily attempting to engage in all possible domains of parenting (Crohn, 2006; Crosbie-Burnett & Giles-Sims, 1994; Erera-Weatherly, 1996).

The casually connected pattern was the next most prevalent. In this pattern, youth and their stepfathers have a low-to-moderate probability of engaging in interactions across recreational, personal, and disciplinary domains. Although stepfathers in this pattern do not appear altogether disengaged, they do not appear too involved either. Notably, youth in this pattern report above-average involvement with their nonresident fathers and below-average relationship quality with their stepfathers. Given youths’ connections to their nonresident parents, it might be unclear to stepparents how they can best involve themselves in the lives of their stepchildren. Together, these dynamics could result in stepparent role ambiguity, consistent with the “uncertain” stepparenting style identified in past research (Erera-Weatherly, 1996). Because youth in this pattern report above-average levels of delinquency, stepfathers might also be reluctant to over-engage their stepchildren. Youth in this pattern also report below-average levels of self-esteem, which might inhibit their desire or ability to engage with their stepfather. Overall, stepfathers in this pattern still appear, at least minimally, to interact or engage with their stepchildren.

The inactive pattern was the third most common in this sample. Youth in this pattern report no engagement with their stepfathers across the 10 interactions analyzed. The relatively low levels of family relationship quality in the residence, coupled with the above-average involvement between youth and their nonresident fathers, could indicate that youth have disengaged from their resident family members. These youth might opt for time and interaction with their nonresident fathers, either out of preference or out of an effort to evade the dynamics of resident family relationships, although the level of nonresident father-child involvement is still moderate at best, as indicated by the raw or non-standardized average level of involvement in the inactive pattern. Youth in this pattern also report the most adjustment problems in the form of delinquency and low self-esteem. This could be either a consequence of or an antecedent to a disengaged stepparent and lower-quality family relationships. Overall, the inactive pattern appears consistent with the “detached” or “disengaged” styles of stepparenting previously identified (Crohn, 2006; Crosbie-Burnett & Giles-Sims, 1994; Erera-Weatherly, 1996), and could reflect a “nonseeker” pattern of affinity-seeking among stepparents (Ganong et al., 1999).

The versatile and involved pattern was the least common in this sample. Youth in this pattern had a moderate-to-high probability of engaging with their stepfathers in all 10 of the interactions analyzed. Youth in this pattern also appear well adjusted, particularly with respect to self-esteem, and are younger than average. Thus, stepfathers might be better able to adopt an involved role in the lives of their stepchildren, as research has shown younger stepchildren are more accepting of a new stepparent relative to their older counterparts (Jensen & Howard, 2015). Stepfamilies marked by the versatile and involved pattern of youth-stepparent interaction also appear well-functioning overall, as each parent-child relationship and the stepcouple relationship are above-average in quality. Both mothers and stepfathers are relatively well educated, and household income is substantially higher than average. In general, families in higher socio-economic brackets possess fewer stressors and demands than families who confront poverty and other forms of disadvantage (Patterson, 2002; Peters & Massey, 1983). Thus, the versatile and involved pattern might more naturally emerge among stepfamilies with a rich portfolio of psychosocial and socio-economic resources, along with a cooperative coparental regime across households. Overall, the versatile and involved pattern appears most congruent with the “birth parent style” of stepparenting (Erera-Weatherly, 1996) and the “continuous affinity-seeker” pattern of affinity-seeking among stepparents (Ganong et al., 1999), resulting in the successful augmentation of youths’ parental network to include a new stepparent (Ganong & Coleman, 2017; Gross, 1987).

That non-Hispanic Black youth are relatively well represented in the casually connected and versatile and involved patterns is not overly surprising. Past research suggests that Black stepfathers engage in more interactions with their stepchildren relative to White stepfathers, on average, particularly in the areas of play, religion, and moral education (Stewart, 2007). That non-Hispanic White youth are overrepresented in the academically oriented pattern could reflect educational aspiration differences across racial/ethnic groups (Kao & Tienda, 1998), although the influence of racial/ethnic identity, culture, socio-economic status, and other characteristics can be difficult to disentangle.

Similarities across interactional patterns are equally informative. The quality of family relationships and youth well-being did not dramatically differ across some of the interactional patterns, particularly between the academically oriented and casually connected patterns. Perhaps not surprisingly, the starkest differences with respect to family and individual characteristics were between the inactive and versatile and involved patterns. Levels of youth-stepfather interaction also appear to be commensurate with levels of youth-mother interaction, suggesting that these patterns might reflect broader family-system dynamics rather than being dyad-specific.

It is worth noting that the emergent patterns in this study might also map onto previously identified patterns of relationship development between youth and their stepparents (Ganong et al., 2011; Kinniburgh-White et al., 2010). The core challenge of mapping extant developmental patterns onto the cross-sectional interactional patterns highlighted in this study is that the latter is not focused on continuity or change in relationship quality over time. Each of the four interactional patterns identified in the current study could reflect features of a developmental process outlined by previous researchers, ranging from accepting as a parent or continuous positive regard to rejecting or continuous struggle (Ganong et al., 2011; Kinniburgh-White et al., 2010), but again, the focal patterns in this study focus more on clusters of interaction types.

Limitations and Future Research

The conclusions associated with this study should be interpreted in the context of some limitations. For one, person-oriented methods, including latent class analysis, are not without criticism. At the forefront of such criticism is uncertainty about the ontological nature of emergent latent classes (i.e., true and distinct subpopulations versus variation in one population). Consequently, rather than making any assertions about the ontology of the latent classes identified, interpretations should center on latent classes being representative of possible variation in interactional patterns in the larger population of mother-stepfather families (Petras & Masyn, 2010). Another limitation is that the current study does not include all possible interactions in which youth and their stepfathers might engage. Perhaps it is not truly possible to know, measure, and analyze all such interactions. One strength is that the interaction variables included in the analysis encompass several life domains, and thus capture information about a variety of interaction types and quantities.

The latent class analysis conducted was also cross-sectional. Thus, the interactional patterns identified serve as momentary snapshots, representing interactions between youth and their stepfathers during the previous four weeks. Youth-stepparent interactions are likely dynamic, and could change or develop over time (Ganong et al., 2011; Kinniburgh-White et al., 2010; Papernow, 2013); however, it is a strength that respondents reported information about a four-week period of time, rather than a shorter timeframe. Moreover, the validation analyses conducted were also cross-sectional, and the results are unable to speak to causal associations between interactional patterns and family and individual characteristics. Thus, the validation analyses should be viewed strictly as correlational, with the possibility that associations are bidirectional and circular. Another limitation is that the data for the current study came from a representative sample of youth who were adolescents just before the turn of the century—a time in which shared or joint custody arrangements were less common than they are now. Indeed, the youth analyzed in this study held primary residence with their mother and stepfather, with low average levels of nonresident-father involvement. Technological advancements, changes in legal custody arrangements, and other cultural shifts have undoubtedly shaped the landscape of family life and interaction with implications for stepfamily functioning, nonresident father involvement, and youth-stepparent relationships.

To move this area of research forward, researchers should expand the scope of interactions analyzed between youth and their stepparents, and incorporate more and newer information about digitally or electronically based forms of interaction. Researchers should also examine interactional patterns between youth and their stepmothers. As noted earlier, stepfamily dynamics can vary as a function of stepparent sex. There is also a need for more research that allows for insights to emerge relating to family processes enacted by stepfamilies headed by gay and lesbian couples. The results provided by this study also point to the value in assessing interactional patterns between youth and their stepparents over longer periods of time. Latent transition analysis (Collins & Lanza, 2010), for example, could be used to examine how membership within one interactional pattern is associated with membership in that same pattern or another pattern at a subsequent point in time. This line of research would be consistent with a developmental perspective by tracking continuity or change in youth-stepparent interactional patterns—features of some previously conducted qualitative research (Gonang et al., 2011; Kinniburgh-White et al., 2010). Longitudinal analyses could also be used to bolster the exploration of antecedents or outcomes of various interactional patterns.

Practical Implications

Despite some limitations, the current study contributes to the existing literature and generates insights for family-oriented practice. In general, family practitioners and scholars should continue attending to family structures that deviate from the nuclear-family model—such deviations are becoming the norm, not the exception (Cherlin, 2010). More specific to the results of the current study, practitioners should be mindful of the variation in interactional patterns that can emerge between youth and their stepfathers. Although an interactional pattern like the versatile and involved pattern might seem optimal, it was the least frequently occurring pattern among stepfamilies in the sample. Thus, it seems unreasonable to expect that all stepfathers will or should engage with youth in ways that reflect the versatile and involved pattern. Many stepfamilies might benefit from other interactional patterns more compatible with the needs and dynamics of the larger family system. Moreover, emergent interactional patterns between youth and their stepfathers are likely influenced by a host of factors, including the age of children in the family, the family’s socio-economic resources and culture, and the duration of the stepfamily or stepcouple relationship, among others. Thus, youth-stepfather interactional patterns should be interpreted in context, and practitioners should withhold making value judgments given the complexity of stepfamily life and diverse experiences and preferences stepfamily members might have.

Ultimately, practitioners should assist stepfamilies in seeking out the interactional patterns between youth and their stepparents that will work best for their situation. Such intervention might be most salient during the middle stages of the stepfamily cycle, in which stepfamily members often begin airing differences and negotiating relationships and roles (Papernow, 2013). Better yet, and consistent with a prevention perspective (Biglan, 2016), family life educators could assist newly forming stepfamilies in understanding the reasonable variation that can exist with respect to the stepparent role and the interactions that take place between stepparents and youth. The results of this study point to some potentially common and appropriate variants, with disparate levels and forms of stepparent involvement in the lives of youth. Associations between youth-stepfather interactional patterns and youth adjustment and family relationship quality seem to favor contexts, when feasible, in which youth and stepfathers establish some form of interaction, whether infrequent and casual, centered on school-related activities, or wide-reaching and versatile.

Practitioners should skillfully attend to dynamics in each stepfamily relationship that might facilitate or obstruct efforts to promote mutually beneficial patterns of interaction between youth and their stepfathers. Such attention should include a focus on nonresident father involvement. Interestingly, the versatile and involved pattern was associated with the highest levels of nonresident father-child involvement, indicating that there is room for both a resident stepfather and nonresident father to engage with youth. In addition, practitioners should attend to youth adjustment in stepfamilies, and explore how youth behavior exerts influence on the formation of youth-stepfather interactional patterns, as well as how youth behavior arises in response to interactional patterns.