Relational uncertainty,partner interference,and privacy boundary turbulence

Abstract

This study explored what relational aspects contribute to spouses managing private, infertility-related information differently. We examined associations among relational uncertainty and partner interference and three manifestations of boundary turbulence: discrepancies in disclosures, relative discomfort, and perceived spousal acceptability. Using data from 50 infertile couples reporting on 250 social network relationships, we used multi-level modeling to test our predictions that relational uncertainty and partner interference would positively correlate with boundary turbulence. Results were partially consistent with the predictions for relational uncertainty and inconsistent with regard to partner interference. These findings suggest that relational features of a partnership coincide with a couple’s ability to coordinate boundaries about their private information and merits further research on the dyadic component of boundary coordination.

Keywords

boundary coordination boundary turbulence communication privacy management disclosure infertility partner interference relational uncertainty

Petronio’s seminal work on privacy management has served as the foundation for research examining the processes at work as individuals make decisions about concealing and revealing private information. Communication privacy management theory (CPM) (Petronio, 2002) conceptualizes private information as a possession (Petronio, 2002), and uses boundaries as a metaphorical representation of how people share and manage their private information with other people. Research on privacy, such as secret revelations (e.g., W. Afifi & Caughlin, 2006) and health disclosures (e.g., Greene, 2009), focuses largely on how individuals manage ownership of private information (for a review, see T. Afifi, Caughlin, & W. Afifi, 2007). Understanding how dyads, such as relational partners, parent-child dyads, and friends, go about co-owning information is an important feature of privacy management that has received less attention. Relational partners may face instances when they manage their shared information in different ways, known as boundary turbulence (Petronio, 2002). Examining how boundary turbulence is manifested in personal relationships, and how the features of a relationship may coincide with this turbulence, are important aspects of understanding how privacy is managed by relational partners.

As a context for this study, we focused on spouses experiencing infertility and their disclosures of infertility-related information to their friends and family. One couple in six will encounter infertility at some point in their reproductive lifespan. For those partners, reproductive difficulties extend beyond the health issues associated with the condition. Infertility coincides with emotional distress (Greil, 1991), financial strain (Lukse & Vacc, 1999), and feelings of uncertainty about the romantic partnership (Steuber & Solomon, 2008). As individuals manage these stressors, they may desire support from their friends and family to help cope with their condition, but they may also want to protect aspects of their experience by keeping the information confined to the marital partnership. Friends and family may also intrude on a couple’s privacy by asking about plans for children, commenting on a couple’s lack of children, or expressing opinions about the couple’s infertility and treatment choices. Thus, infertile individuals must weigh the benefits of sharing their burdens with others against their desire to avoid disapproval and unwelcomed interactions. Whereas these dueling desires are inherent aspects of an individual’s privacy management (Petronio, 2002), the situation is further complicated when the information is co-owned or shared. In the context of infertility, spouses share ownership over infertility-related information because typically they both have knowledge of their diagnosis, treatment, and prognosis. Under these circumstances, partners have to manage their own disclosures, consider how their spouse may be affected by those revelations, and be subjected to the ramifications of a partner’s disclosure decisions.

The goal of this study is to examine manifestations of boundary turbulence, the asynchronous coordination of privacy rules (Petronio, 2002), within spouses’ disclosures about their infertility to friends and families, and to explore relational characteristics that may promote boundary turbulence. First, we turn to communication privacy management theory to understand privacy boundaries, rule coordination, and behaviors. Next, we use the relational turbulence model to identify relational qualities that may correspond with markers of boundary turbulence. We then test our conceptual model on data collected from a community sample of infertile couples. Finally, we discuss our findings and implications for future study.

A communication privacy management perspective on infertility disclosure

Communication privacy management theory (CPM) is a grounded theory focusing on the management of private information (Petronio, 2002). According to the theory, the need to simultaneously regulate revealing and concealing information forces people to turn to a management system – a system that consists of rule foundations and boundaries. Privacy rules can become especially relevant when the information is co-owned. Information can become co-owned either because of shared experiences or one person disclosing that information to another, with the outcome being that the information is no longer the sole property of one person. Instead, co-owners of private information have shared responsibility for defining and maintaining privacy boundaries.

Petronio (2002) maintained that if the parameters, or rules, surrounding private information are clear, then co-owners are more likely to manage the information effectively. For co-owners to enact mutually approved privacy rules, CPM asserts that individuals need to coordinate linkages, control, and permeability (Petronio, 2002). Linkages connect or convert one boundary type into another type of boundary. Co-owners need to coordinate the linkages or alliances they may form with people outside of their privacy boundary, such as a mother, friend, or co-worker who may become another co-owner of the information (Petronio & Durham, 2008). Control entails determining combine it with the above paragraph who actually owns the information. Coordination occurs when privacy borders are clearly defined with regard to ownership. The third coordination process with regard to privacy rules is permeability, which refers to “the amount of access or openness within a privacy boundary” (Petronio & Durham, 2008, p. 315). Co-owners need to decide the thickness or thinness of their boundary walls, meaning they need to coordinate rules for how open or protected they would like their private information to be.

According to Petronio (2002, p. 117), “when people are unable to collectively develop, execute, or enact rules guiding permeability, ownership, and linkages, the coordinating efforts of privacy management are confounded” and boundary turbulence occurs. Therefore, incongruent expectations and misconstrued parameters can result in conflicting privacy management behaviors, and when boundary coordination fails, boundary turbulence occurs. Petronio (2002) identified several factors that could result in boundary turbulence. First, people could intentionally violate privacy rules and divulge the information. Individuals could also mistakenly breach rules and further spread access to private information. Privacy dilemmas, such as when a doctor discovers that a patient is knowingly spreading a disease and has to decide if he or she should intervene, can also lead to conflict over communication boundaries. Individuals may also have conflicting ideas about how information should be managed, or be thrown into predicaments surrounding boundary definitions. Finally, people might be unclear about boundaries, and the vagueness could contribute to information leaking unintentionally. The potential for boundary turbulence presents a continuous challenge to individuals as they manage private information.

For couples coping with infertility, boundary turbulence can emerge when spouses differ in their perceptions or behaviors related to sharing infertility-related information with people outside the marriage. These differences could stem from a variety of causes, such as different boundary orientations, conflicting perspectives on who owns the information, or competing preferences for privacy management. Perhaps the most tangible manifestation of boundary turbulence is when spouses manage their private information in asynchronous ways. In infertile dyads, this form of boundary turbulence is reflected in a discrepancy between the amounts of information spouses share with a third party. In the vernacular of CPM, disclosure discrepancies may reflect coordination issues surrounding linkages. Boundary turbulence can also be activated when couples are differentially comfortable with disclosures to outsiders. For example, a wife may feel comfortable sharing details about an infertility treatment to her co-worker, whereas her husband may feel embarrassed about the details that have been disclosed; in this case, boundary turbulence is manifest in the husband’s relative discomfort with his wife’s disclosures. Within the parameters of CPM, relative discomfort can stem from privacy rules about ownership, specifically related to issues of control. Finally, boundary turbulence may surface when people perceive that their relationship partner disapproves of their disclosures to others. Within the infertility domain, an example would be a husband perceiving that his wife disapproved of his conversations with others about their infertility. This example can represent issues of permeability, or when co-owners have conflicting ideas about the extent of openness within a privacy boundary. All of these conditions – discrepancies in disclosive behavior, relative discomfort about private disclosures, and perceptions of disapproval – suggest a lack of coordination in the creation and enactment of boundaries for co-owned information and may be indicators of boundary turbulence within a relationship.

Research has examined privacy management related to family planning. Park (2002) noted that couples who decided to be voluntarily childless experienced difficulties sharing that decision because they anticipated a negative response from friends and family. Bute and Vik (2010) found that women coping with infertility reported varying levels of openness depending on shifts in treatment, responses from confidants, and the development of new perspectives on their reproductive disability. Qualitative research has also demonstrated the range of emotional and communicative reactions infertile women have in response to direct requests about their childless state or infertile status (Bute, 2009). Although family and friends have the capability to increase personal and marital well-being during stressful times (Willits & Crider, 1986), privacy management issues such as disapproval from social network members or direct requests for information (e.g., Bute, 2009) can also stress individuals. This body of work implies that infertile couples face challenges coordinating relationships with others; however, we know of no work that has focused specifically on the challenges of boundary turbulence between partners. Accordingly, a first aim of this study is to explore the presence of boundary turbulence between couples who are coping with reproductive disabilities.

RQ1: To what extent is boundary turbulence, in the form of (a) discrepancies between partners’ disclosures of infertility-related information to people outside the marriage, (b) husbands’ and wives’ relative discomfort with their spouse’s infertility-related disclosures to people outside their marriage, and (c) people’s perceptions that their disclosures to individuals outside the marriage are unacceptable to their spouses, apparent within couples managing infertility-related information?

CPM highlights the tensions and challenges involved in deciding whether to disclose information to others, and also the complexities that emerge when people share or co-own information. One limitation of CPM, however, is its relative inattention to specifying the conditions under which co-owners of information will disagree about communication boundaries and violate expectations. Next, we turn to the relational turbulence model to identify relationship qualities likely to influence the experience of boundary turbulence.

Factors affecting boundary coordination between spouses

A number of contextual variables may affect whether couples effectively coordinate boundaries around shared infertility-related information. For example, reoccurring treatment experiences may correlate positively with boundary coordination, because spouses may be able to anticipate responses and rely on past circumstances to guide their information management decisions. The type of diagnosis may also factor into boundary coordination; for example, if the source of infertility is a specific condition within one spouse, that individual may be more protective of that private medical information than the non-infertile spouse. Beyond these contextual features, aspects of the relationship between spouses can affect their ability to establish, recognize, and meet information-sharing expectations. The relational turbulence model (RTM) (Solomon & Knobloch, 2004) highlights relationship qualities that can complicate communication between partners. In the following sections, we review the main assumptions of RTM and deduce hypotheses linking relational characteristics to the coordination of boundaries for infertility information.

Relational uncertainty

The relational turbulence model positions relational uncertainty as one of the underlying mechanisms contributing to heightened reactivity to events occurring within romantic relationships. Relational uncertainty refers to the degree of confidence individuals have in their perceptions of involvement within a relationship, and it encompasses questions people have about their own involvement, a partner’s involvement, and the nature of the dyadic unit (Berger & Bradac, 1982; Knobloch & Solomon, 1999, 2002a). Although the theory is grounded in the assumption that relational uncertainty increases when partners transition between levels of involvement (see Solomon, Weber, & Steuber, 2010), most of the empirical research has documented the correlates of relational uncertainty within romantic dyads. Research has shown that uncertainty about a relationship coincides with polarized cognitive reactions, such as a tendency to view relationship irritations as more serious (Solomon & Knobloch, 2004; Theiss & Solomon, 2006b), and the inability to draw clear relationship inferences (Knobloch & Solomon, 2005). Relational uncertainty also corresponds with more extreme emotional reactions to surprising events (Knobloch & Solomon, 2002b, 2003), jealousy threats (W. Afifi & Reichert, 1996; Knobloch, Solomon, & Cruz, 2001), and changes in sexual intimacy (Theiss & Solomon, 2007). The prevalence of relational uncertainty has also been tied to avoiding conversations about the relationship (Baxter & Wilmot, 1985), avoiding particular topics or withholding private information (W. Afifi & Burgoon, 1998; W. Afifi & Guerrero, 2000; Knobloch & Carpenter-Theune, 2004), and refraining from discussions about surprising relationship events (Knobloch & Solomon, 2002b). This body of research characterizes a person experiencing relational uncertainty as someone who has strong cognitive and emotional experiences, but is also reluctant to communicate with a partner about those experiences.

Steuber and Solomon (2008) demonstrated how relational uncertainty surfaces within infertile couples. In particular, that qualitative study concluded that the difficulty inherent in balancing the treatment process with maintaining a marriage often contributes to doubts about the marital bond. The discourse examined in that study also showed that the management of blame, either directed toward the self or the partner, instigates sentiments of relational uncertainty. To the extent that these issues are inherent aspects of infertility, couples may have to make decisions about disclosure to social network members in a relational context characterized by doubts and ambiguities. Engaging in this negotiation process becomes problematic in couples coping with relational uncertainty, because partners characterized by that ambiguity often avoid conversations about relationships (Baxter & Wilmot, 1985) and withhold private information (e.g., Knobloch & Carpenter-Theune, 2004). If couples coping with relational uncertainty are less likely to communicate openly, opportunities for spouses to negotiate privacy management boundaries may be limited. Accordingly, we expect that relational uncertainty will undermine boundary coordination between spouses, resulting in boundary turbulence. Specifically, we predict:

H1: Husbands’ and wives’ relational uncertainty is (a) positively associated with the size of the discrepancy between partners’ disclosures of infertility-related information to people outside the marriage; (b) positively associated with husbands’ and wives’ relative discomfort with their spouse’s infertility-related disclosures to people outside the marriage; and (c) negatively associated with people’s perceptions that their disclosures to individuals outside the marriage are acceptable to their spouses.

Interference and facilitation from a partner

The second component highlighted by the relational turbulence model focuses on difficulties related to developing and sustaining behavioral interdependence. Interdependence requires that partners allow each other to take part in their daily routines (Kelley et al., 1983). A partner’s integration into day-to-day activities is experienced as facilitation from a partner when it allows routines to be more effective or increases the likelihood that goals will be achieved (Berscheid, 1983; Knobloch & Solomon, 2004). When a partner’s influence disrupts the performance of a behavioral routine, the result is the experience of interference from a partner (Berscheid, 1983; Knobloch & Solomon, 2004). As in the case of relational uncertainty, interference from a partner is positively correlated with the perceived severity of relational irritations (Solomon & Knobloch, 2004; Theiss & Solomon, 2006b) and the intensity of emotional jealousy (Theiss & Solomon, 2006a). Conversely, facilitation from partners is negatively associated with perceptions that a relationship is unsettled or chaotic (see Solomon, Weber, & Steuber, 2010) and the experience of negative emotions (Knobloch, Miller, & Carpenter, 2007).

As with relational uncertainty, Steuber and Solomon (2008) identified interference from a partner as an issue permeating discourse about infertility. In particular, people coping with infertility highlighted tensions arising when their partner wanted to respond to infertility differently than they did. Partner interference was also represented in discourse about frustrations when the partner prioritized infertility over the normalcy of the relationship or, conversely, failed to appreciate the gravity of the condition. Although speculative, we wonder if the inability to form a shared agenda for infertility treatment or the marriage extends to the coordination of rules for sharing infertility-related information. Conversely, couples who facilitate each other’s everyday goals and activities may support and promote each other’s information management preferences. More generally, the research reviewed previously highlighted how experiences of everyday goal interference from a partner can heighten experiences of negative emotion and polarize negative appraisals of a partner’s behavior, whereas facilitation from a partner corresponds with less turbulent relationship experiences. By extension, we predict that markers of boundary turbulence correspond with experiences of interference and facilitation from a partner:

H2: Husbands’ and wives’ partner interference is (a) positively associated with the size of the discrepancy between partners’ disclosures of infertility-related information to people outside their marriage; (b) positively associated with husbands’ and wives’ relative discomfort with their spouse’s infertility-related disclosures to people outside their marriage; and (c) negatively associated with their perceptions that their disclosures to people outside their marriage are acceptable to their spouses.

H3: Husbands’ and wives’ partner facilitation is (a) negatively associated with the size of the discrepancy between partners’ disclosures of infertility-related information to people outside their marriage; (b) negatively associated with husbands’ and wives’ relative discomfort with their spouse’s infertility-related disclosures to people outside their marriage; and (c) positively associated with their perceptions that their disclosures to people outside their marriage are acceptable to their spouses.

To review, we drew upon CPM to conceptualize boundary turbulence, and we posited a research question to explore the extent to which boundary turbulence is manifest in behaviors and perceptions related to spouses sharing infertility-related information with people outside the marriage. Then, we used RTM to deduce hypotheses linking relational uncertainty and interference and facilitation from a partner to manifestations of boundary turbulence. Next, we report a study designed to provide insight into our research question and test our hypotheses.

Method

We required a research design that allowed us to juxtapose spouses’ accounts of their disclosures about their infertility to third parties, their feelings about those disclosures, and their perceptions of each other. To this end, we examined spouses’ disclosures about particular topics to specific individuals within the couple’s social network. To ensure that individuals were reporting on significant people in their lives, subjects narrowed their focus to a subset of their social network, the social convoy. Network members may simply function within a person’s social world, whereas a convoy is a subset of a social network that includes individuals who are significant in terms of social support (Kahn & Antonucci, 1980). In total, each partner reported on relationships with and disclosures to the same five social convoy members. This method sheds light on how couples compared in their privacy management behaviors with respect to specific members of their support convoy.

Subjects

Subjects were 50 heterosexual married couples (100 individuals), the majority of whom responded to advertisements posted in the northeast region of the website Craigslist (n = 18 couples). The remaining couples were recruited from a subject list from a previous study (Steuber & Solomon, 2010; n = 4 couples), from a northeast chapter of a national infertility organization (n = 11 couples), from doctors’ offices (n = 7 couples), from a university Listserv announcing the study (n = 8 couples), from flyers posted in public meeting places (n = 1 couple), and from radio announcements (n = 1 couple). To qualify for our study, couples had to have been coping with infertility for five years or fewer (M = 2.64, SD = 1.19). The majority of our sample was coping with primary infertility (i.e., no previous children; 78 individuals). Couples reported treating their infertility in various ways, from home remedies such as charting menstrual cycles to medically enhanced procedures such as ovulation drugs or in vitro fertilization. In our sample, 85% of couples reported using at least one medically enhanced procedure (M = 5.23, SD = 4.70).

Subjects identified as predominantly Caucasian (90%; Hispanic or Latino, 6%; Black or African American, 5%; Asian, 1%; Afghan, 1%; and “other,” 2%). Individuals ranged in age from 24 years to 47 years (M = 34.4, SD = 4.97). Couples had been married between one and 14 years (M = 5.57 years, SD = 3.15), and most individuals reported being in their first marriage (91%; second marriage, 7%; third or subsequent marriage, 2%).

Procedures

Infertile couples who contacted the researcher and qualified for the study were scheduled for a meeting at a location convenient and comfortable for them. Many meetings took place at local coffee shops or libraries, and the appointments lasted between 45 minutes and 120 minutes. Although both partners were present for the duration of the meeting, they only interacted with each other to select five social convoy members about whom they individually answered questions. Couples were compensated $60 for participation.

The procedures employed were developed for this study. First, spouses individually identified all the current or recent members of their social convoy. Once each spouse finished his or her list of convoy members, they worked together to identify any overlapping members (e.g., both mentioned her mother) so that these individuals would not be represented twice in the combined list of names. In addition, any individuals known to only one spouse were removed. From the remaining combined list, a total of five social convoy members were randomly selected. To avoid drawing a set of convoy members that was more representative of one spouse’s list, names were drawn in a manner that ensured representation of members identified by the husband, wife, and both.

The individual convoys that couples self-recalled ranged from 2 to 32 individuals (M = 10.00, SD = 5.43). The convoys were narrowed down to the five convoy members for each couple, resulting in a total of 250 convoy members (five members per 50 couples). Of the 250 convoy members, 74 were nominated by wives (29.6%), 69 were nominated by husbands (27.4%), and 107 were identified by both the husband and wife (42.9%). Individuals independently reported knowing these convoy members for between 0.15 and 47.00 years (M = 14.50, SD = 11.90). Of those 250 convoy members, 138 were female (55.5%). Convoy members represented a variety of roles in the couple’s life. Many of the convoy members were friends (n = 101, 40.4%), parents (n = 72, 28.8%), and siblings (n = 41, 16.4%). The remaining convoy members were classified as counselors/clergy (n =11, 4.0%), co-workers (n = 9, 3.6%), doctors/nurses (n = 5, 2.0%), aunts/uncles (n = 4, 1.6%), grandparents (n = 3, 1.2%), cousins (n = 2, 0.8%), a neighbor (n = 1, 0.4%), and a nephew (n = 1, 0.4%).

After a couple identified a common set of network members, spouses individually completed questionnaires that asked about qualities of the marital relationship. In addition, each spouse answered questions with regard to each of the five social convoy members. To provide a concrete focus for questions about communication with the convey members, subjects were asked to consider five specific topics for disclosure: treatment plans, treatment outcomes, long-term treatment perspectives, emotional responses to infertility, and marital issues related to infertility. These topics were selected because they capture a variety of private issues that are salient to people coping with infertility (e.g., Greil, 1991; Steuber & Solomon, 2010). For each topic and convoy member, subjects indicated how much they had disclosed to the individual, how comfortable they were with their disclosures to that person, how comfortable they were with their spouse’s disclosure to this person, and how they thought their spouse felt about their own disclosure to this convoy member. Thus, our method gathered assessments from each spouse with regard to specific topics and specific targets, which we could then juxtapose to gain insight into boundary coordination and turbulence.

Measures

The design of the study yielded variables indexing both qualities of the marital relationships and boundary turbulence in spouses’ disclosure of infertility-related information.

Marital characteristics

Subjects provided responses to measures about marital satisfaction, partner interference and facilitation, and relational uncertainty. These scales were subjected to a confirmatory factor analysis with items representing each scale loaded on separate factors and all factors allowed to covary. The data did not fit the measurement model. Modification indices suggested that the relational uncertainty items were contributing to the lack of fit because of deviations from normality apparent in those items. With items indexing relational uncertainty removed, the measurement model for the three remaining measures fit the data (χ ²/df = 1.70; CFI = .94; RMSEA = .09). The following paragraphs discuss each marital characteristic in detail.

Items from the Relationship Assessment Scale (RAS) (Hendrick, Dicke, & Hendrick, 1998) were used to measure marital satisfaction. Subjects answered items on a scale from one to five (five being most satisfied), documenting relational needs, regrets, and expectations (e.g., how well does your partner meet your needs, to what extent has your relationship met your general expectations?). The four items (α = .88) were averaged to form a composite score (M = 4.40, SD = 0.67). Marital satisfaction was not used in substantive analyses, but it was included to provide descriptive information about the sample. In sum, people in the sample were predominantly satisfied with their marriages.

To report partner interference and partner facilitation, respondents used a five-point scale (1 = strongly disagree, 5 = strongly agree) to answer a series of questions addressing the partner’s effects on everyday routines and plans (Knobloch & Solomon, 2004). The interference scale (e.g., in the past two weeks, my partner has disrupted my daily routine) consisted of four items (α = .80, M = 2.12, SD = 0.91). Similarly, the facilitation scale (e.g., in the past two weeks, my partner helped me to do the things I need to do each day) consisted of four items (α = .84, M = 3.94, SD = 0.82).

Abbreviated versions of Knobloch and Solomon’s (1999) scales were used to assess relational uncertainty (see Knobloch, Miller, Bond, & Mannone, 2007). Subjects responded using a six-point scale (1 = completely or almost completely uncertain, 6 = completely or almost completely certain) to items describing how they felt about their marriage over the past two weeks. Items were rescored so that higher values reflected more uncertainty. Self-uncertainty reflected the presence of doubts about the respondent’s own involvement in the relationship, partner uncertainty represented the respondent’s doubts about the spouse’s involvement in the relationship, and relationship uncertainty addressed questions about the future of the relationship in general. Reliabilities for the three facets of relational uncertainty were acceptable in past studies (self-uncertainty, α = .84; partner uncertainty, α = .90; relationship uncertainty, α = .85) (Knobloch et al., 2007) and the current study (self-uncertainty, α = .86; partner uncertainty, α = .93; relationship uncertainty, α = .88). However, a confirmatory factor analysis with self, partner, and relationship uncertainty items loaded onto three factors which were allowed to covary, suggested that the data did not fit the measurement model well (χ ²/df = 2.30; CFI = .89; RMSEA = .12). With this caveat, the measures were retained to allow a test of the corresponding hypothesis.

The three measures of relational uncertainty were highly correlated (r = .74–.91). For reasons of parsimony, then, we focused our analyses on relationship uncertainty (M = 1.48, SD = .73, α = .88). This decision was guided by the conceptualization of self- and partner uncertainty as antecedent to relationship uncertainty (e.g., Berger & Bradac, 1982; Knobloch & Solomon, 1999), and structural equation models of empirical data which position relationship uncertainty as the proximal predictor of exogenous variables (e.g., Knobloch & Carpenter-Theune, 2004; Knobloch, Miller, & Carpenter, 2007; McLaren, Solomon, & Priem, 2010). When results differ substantively for self- and partner uncertainty, we note those differences in an endnote.

Communication variables

Recall that subjects were asked to report on aspects of communication about five different topics: treatment plans, treatment outcomes, treatment perspectives, emotional aspects related to infertility, and marital aspects related to infertility. Specifically, they were asked to report, on a five-point scale, how much information they revealed to each social network member regarding each topic (1 = no information, 5 = a great deal of information). In addition, subjects were asked to report how comfortable (1 = not at all comfortable, 5 = very comfortable) they were “with the information they revealed for each topic”, how comfortable they were “with the information they think their spouse revealed about each topic”, and how comfortable they “think their spouse was with the information they revealed for each topic”.

For each topic and convoy member, subjects responded to single items, which were aggregated across topics when computing the outcome variables. In this measure, the individual items for each topic are indicators of a latent construct representing communication with the social network member about infertility issues in general, and having multiple topics is akin to having multiple items within a scale. Items across topics had high intraclass correlation (disclosure, ICC = .95; comfort with own disclosure, ICC = .95; comfort with spouse disclosure, ICC = .95; and perceived spouse comfort, ICC = .93) indicating high resemblance in responses to targets (see Table 1 for descriptive statistics).

Table 1.

Descriptive information for infertility-related topics

Topics	Disclosure	Comfort with own disclosures	Comfort with spouse’s disclosures	Perceived spouse acceptance of own disclosures
Treatment plans	2.86 (1.57)	4.33 (1.10)	4.23 (1.19)	4.16 (1.17)
Treatment outcomes	2.89 (1.67)	4.35 (1.10)	4.25 (1.20)	4.23 (1.13)
Treatment perspectives	2.70 (1.58)	4.35 (1.07)	4.25 (1.18)	4.26 (1.10)
Emotional responses	2.53 (1.48)	4.24 (1.14)	4.14 (1.24)	4.15 (1.16)
Marital issues	2.01 (1.32)	4.23 (1.15)	4.12 (1.25)	4.13 (1.19)

By integrating the responses spouses provided with respect to specific targets, we were able to index three aspects of boundary turbulence. First, recall that H1a and H2a predicted that relational uncertainty and partner interference were positively associated with the size of the discrepancy between partners’ disclosures to social convoy members, and H3a proposed that facilitation was negatively associated with the size of discrepancy. The dependent variable, disclosure discrepancy, was computed using the item “How much information have you shared with this person?”. The variable was computed by taking the absolute value of the difference between the wife’s disclosure and the husband’s disclosure on each of the five topics (ICC = .90). The five scores were then averaged to get one score for each social convoy member (M = 1.48, SD = 1.10).

To test the associations between relational uncertainty, partner interference, and facilitation and how matched or mismatched spouses were in their comfort with disclosures, we created the variable relative discomfort (H1b, H2b, H3b). For the husband’s relative discomfort with his wife’s disclosure, the husband’s response to “How comfortable are you with the amount of information your spouse has shared?” was subtracted from the wife’s response to “How comfortable are you with the amount of information you’ve shared with this person?”. This discrepancy was calculated across all five topics (ICC = .94) and then averaged (M = 0.16, SD = 1.27). The parallel computation was used to operationalize wives’ relative discomfort with husbands’ disclosures (ICC = .93, M = 0.05, SD = 1.31).

Finally, to test the associations between the perceived acceptability of disclosure and relational uncertainty, partner interference, and facilitation (H1c, H2c, H3c), the response to the item “How comfortable do you think your spouse is with the amount of information you’ve shared with this person?” was used to compute the dependent variable. Perceived acceptability of disclosure across all five topics (husbands’ ICC = .93; wives’ ICC = .94) were averaged to form a composite variable for husbands (M = 4.18, SD = 1.01) and wives (M = 4.19, SD = 1.03).

Analysis

When using dyadic data, adjustments have to be made for the non-independence that is inherent in the dataset. In this study, non-independence exists at two levels. First, the husband and wife data were jointly considered to form the measures of boundary turbulence as detailed previously. Second, the measures of disclosure discrepancy, relative discomfort with spouse’s disclosure, and perceived acceptability of disclosure are measured with respect to convoy members who share membership within the respondents’ social networks. Those convoy members are in a cluster, such that they have more commonality among one another than with those not in the cluster (Slater, Snyder, & Hayes, 2006). To address the non-independence in the data, multilevel modeling was used. One of the benefits of multilevel modeling is that it can handle nested, or clustered, data (Byrk & Raudenbush, 2002), and this nesting controls for the non-independence inherent in the levels of analysis.

To test the hypotheses predicting discrepancies in disclosure, relative discomfort, and perceived acceptability (H1–H3), two-level multilevel models were estimated. Two levels were used for these models because the hypotheses are predicting an outcome based on both husbands’ and wives’ behavior with respect to a particular social network member. Accordingly, social convoy members were at Level 1, and the five social convoy members addressed by each couple were nested within marital dyads, which represented Level 2 (see Figure 1).

Figure 1.

Diagram of two-level model for H1–H3

For a two-level model, the intraclass correlation calculates the proportion of variance in the dependent variable that can be attributed to each level in the model. A ρ closer to 1 indicates that most of the variance in the dependent variable is based on Level 2 differences; in this analysis, differences between couples. A ρ closer to 0 indicates that most of the variance in the dependent variable is based on Level 1 differences, which reflect differences between social convoy members in this analysis (Kreft & de Leeuw, 2002; Snijders & Bosker, 2003). For the discrepancy in disclosure model, ρ = .03, suggesting that most of the variance reflects differences among social network members. For the model estimating wife’s relative discomfort with husband’s disclosures (ρ = .39), husband’s relative discomfort with wife’s disclosures (ρ = .37), wife’s perception of husband’s comfort (ρ = .47), and husband’s perception of wife’s comfort (ρ = .44), scores suggested that variance in these dependent variables reflects both within and between couple effects.

In constructing the models, we included the social convoy member’s sex as a Level 1 predictor. Research has shown that both men and women prefer and seek support from women, suggesting that sex of the target may impact decisions about revealing information to members of their support convoy (Kunkel & Burleson, 1999). At Level 2, we represented characteristics of the marital relationship. In total, we had six variables at this level: relationship uncertainty, partner interference, and facilitation, each of which was reported by both husbands and wives. Not surprisingly, many of these variables were intercorrelated (r = .32–.74), such that we were concerned that evaluating them simultaneously would mask substantive associations. To provide a parsimonious test of our hypotheses, while also taking into consideration covariation in each person’s self-reports, we ran one set of models using the husbands’ variables, and a separate set of models using the wives’ perceptions of the relationship in the analyses. In all cases, the measures of relationship uncertainty, partner interference, and partner facilitation were entered on the model intercept. Thus, the coefficients for these variables estimate differences in the dependent variables as a function of these relational characteristics. The relationship variables were grand mean-centered (i.e., the observed scores were centered around the sample mean for the variable).

In the following equations, the subscript i refers to social convoy member and the subscript j refers to the couple. The model below shows the husbands’ variables as the Level 2 predictors, and an identical model evaluated wives’ variables as Level 2 predictors.

Model 1: Husbands’ perceptions of relationship qualities as predictors of boundary turbulence:

Level 1 equation:

Y_{i j} = β_{0 j} + β_{1 j} (social convoy member sex) + r_{i j}

Level 2 equation:

\begin{array}{l} β_{0 j} + γ_{00} + γ_{01} (husbands’ relationship uncertainty) \\ + γ_{02} (husbands’ partner interference) \\ + γ_{03} (husbands’ partner faciliation)+ u_{0 j} \end{array}

β_{1 j} = γ_{10} + u_{1 j}

In the above model, β_0j is the mean disclosure to target i, and r_ij is the random “target effect” or the deviation of target ij’s mean from the couple’s mean. γ₀₀ is the grand mean, and u _0j is a random “couple effect” or the deviation of couple j’s mean from the grand mean. In total, H1–H3 involved running 10 models, reflecting five different dependent variables (i.e., disclosure discrepancy, husband and wife relative discomfort, and husband and wife perceived acceptability) by both the husbands’ and wives’ predictor variables.

Results

Manifestations of boundary turbulence (RQ1)

To shed light on boundary turbulence within couples who are managing private, infertility-related information, we probed our three measures of boundary turbulence: disclosure discrepancies, relative discomfort with disclosure, and perceived acceptability of disclosures.

In reference to disclosures to social convoy members, recall that subjects were asked to rate on a five-point scale how much they revealed to their convoy member about five different topics related to infertility. We aggregated the disclosure scores across topics to provide a general index of disclosure (husbands’ ICC = .95, wives’ ICC = .95). A paired sample t-test revealed a significant difference in husband and wife disclosures, t (249) = 29.74, p<.001, with wives (M = 2.73, SD = 1.10) disclosing more than husbands (M = 2.46, SD = 1.30). The disclosure discrepancy variable, which was computed as the aggregate of discrepancy scores for each topic, ranged theoretically from 0 (no discrepancy) to a 4 (highest discrepancy). Our sample averaged a discrepancy of 1.48 (SD = 1.10). For those couples whose wives disclosed more than the husbands, wives had an average disclosure score of 3.80 (SD = 1.05, range = 1.20–5.00) and husbands had an average score of 1.97 (SD = 1.09). In the instances where husbands revealed more than the wives, husbands had a mean disclosure score of 3.60 (SD = 1.09, range 1.00–4.60) and wives had an average score of 1.96 (SD = 1.10). In the situations in which couples were disclosing the same amounts to their network members, the mean disclosure was 3.20 (SD = 1.05, range = 1.00–4.20). These findings suggest that when discrepancies exist, one partner is typically being especially private; in dyads with partners disclosing at similar levels, those partners are typically both moderately disclosive.

We also examined husbands’ and wives’ comfort with disclosures. As described previously, we assessed husbands’ relative discomfort with their wife’s disclosures by subtracting wives’ comfort with their own disclosures to a convoy member from their husbands’ comfort level regarding her disclosure about that topic to that target; the wives’ relative discomfort variable was computed in a similar fashion. The scores were then aggregated across the five topics for each couple. The possible range for the variable extended from −4 to +4, with positive scores representing discomfort. Although husbands and wives, on average, manifested minimal relative discomfort with each other’s disclosures (husbands’ M = 0.16, SD = 1.28; wives’ M = 0.05, SD = 1.31), the scores for both husbands and wives spanned the full range from −4 to +4. For husbands’ relative discomfort with their wife’s disclosure, 6.1% of husbands reported a discrepancy score below −2, 23% reported a score between −2 and 0, 64.8% documented a score between 0 and 2, and 6.1% reported a score higher than 2. With regard to wives’ relative discomfort with their husband’s disclosures, 7.7% of wives reported a discrepancy score below −2, 58.3% reported a score between −2 and 0, 27.5% documented a score between 0 and 2, and 6.5% of wives reported a score higher than 2. This look at the variable distribution indicates that the modal category for males was relative discomfort (i.e., a score between 0 and 2), whereas the modal category for females was relative comfort (i.e., a score between 0 and −2).

To address boundary turbulence operationalized as a spouse’s perceived disapproval of disclosures, recall that we used the question “how comfortable do you think your spouse is with your disclosure to this social convoy member?” regarding the five infertility-related topics, and we averaged those scores across topics. On average, both husbands (M = 4.18, SD = 1.01) and wives (M = 4.19, SD = 1.03) perceived that their spouses were comfortable with their disclosures to social convoy members; however, the full range of scores (i.e., from 1.00 to 5.00) were represented for both husbands and wives. For husbands’ perceived acceptability from wives, 4% of the husbands in our study reported a score below 2, 27.6% documented a score between 2 and 4, and 68.4% reported scores higher than 4. For wives’ perceived acceptability from husbands, 4.4% documented scores below 2, 24.6% reported scores between 2 and 4, and 71% reported scores over 4. For this variable, the distributions for husbands and wives were markedly similar, and perceived approval for disclosures is the clear norm.

Disclosure discrepancies (H1a, H2a, H3a)

To test the association between predictor variables and the discrepancy between partners’ tendencies to disclose infertility-related information to social convoy members (H1a), a two-level linear model was estimated with disclosure discrepancy as the outcome variable. Results from this model (see Table 2) indicated that there were more discrepancies between partners when disclosing to male social convoy members. There were no significant associations between disclosure discrepancies and husbands’ or wives’ relationship uncertainty, partner interference, and facilitation. Accordingly, there was no support found for H1a, H2a, or H3a.¹

Table 2.

Results for the hierarchical linear models (H1–H3)

	Disclosure	Relative Discomfort		Perceived Acceptability
	Discrepancies	Husband	Wife	Husband	Wife
Husbands’ Predictor Variables
Intercept	1.95^**	0.18	0.27	4.68^**	4.27^**
Convoy member sex	−0.32^**	−0.30^*	−0.03	−0.35^**	−0.06
Relationship uncertainty^{a ^b}	0.11	0.04	−0.08	−0.38^**	−0.15
Partner interference^a	−0.15	−0.05	0.00	0.07	0.04
Partner facilitation^a	−0.10	0.04	0.12	0.11	0.05
Wives’ Predictor Variables
Intercept	1.92^**	−0.17	0.28	4.69^**	4.23^**
Convoy member sex	−0.31^**	−0.29^**	−0.04	−0.35^*	−0.05
Relationship uncertainty^ab	0.13	0.00	0.14	−0.56^**	−0.25
Partner interference^a	0.08	−0.12	−0.02	0.16	−0.04
Partner facilitation^a	−0.08	0.02	0.06	−0.12	−0.03

Note: Cell entries are model coefficients.

^a Variables on the slope at Level 2,

^b Relationship uncertainty is serving as a proximal predictor for relational uncertainty.

*p<.05,

^** p<.01.

Relative discomfort with disclosures (H1b, H2b, H3b)

Recall that husbands’ relative discomfort with wives’ disclosures is represented by the discrepancy between the wife’s reported comfort with her own sharing and the husband’s reported comfort with the wife’s sharing. Wives’ relative discomfort with husbands’ disclosures was computed similarly. Results from the two-level models (see Table 2) suggest that individuals experienced more relative discomfort when disclosing to male social convoy members. There were no significant findings for the association between relative discomfort and relationship uncertainty (H1b), partner interference (H2b), and facilitation (H3b).²

Perceived acceptability of disclosures (H1c, H2c, H3c)

Results for models testing perceived acceptability suggest that husbands perceived their spouses as more accepting of their disclosures to male social convoy members (see Table 2). Husbands also perceived less acceptability from their wives when they themselves were experiencing relational uncertainty or when their wives were experiencing relational uncertainty (H1c).³ There were no significant associations between perceived acceptability and partner interference or facilitation (H2c, H3c). Thus, partial support was found for hypotheses predicting associations between the predictor variables and perceived acceptability.

Discussion

Petronio (2002) emphasized how the complexity of co-owning information sometimes results in disrupted coordination, otherwise known as boundary turbulence. Several factors can contribute to boundary turbulence, including an intentional leak, a mistaken violation of rules, privacy dilemmas about who has the right to know information, and unclear boundaries. Beyond identifying these threats to coordinated information management, the theory does not specify the relational conditions that contribute to boundary turbulence. Hence, we drew from the relational turbulence model to extend CPM by identifying marital characteristics that might contribute to breaches in boundary coordination.

In this project, we considered three manifestations of boundary turbulence and examined how they relate to relational qualities featured in RTM. One contributor to boundary turbulence can be discrepant rules for linkages (Petronio, 2002). Accordingly, we explored discrepancies between spouses’ actual disclosures, which we measured by comparing the independent self-reports of wives’ and husbands’ disclosures to the same targets about particular infertility-related topics. In addressing our research question (RQ1), we found that wives reveal significantly more than husbands, and the highest levels of disclosure in our sample occurred when wives disclosed to female convoy members. We also observed that couples disclosing at discrepant levels tended to include one partner who was especially non-disclosive. In addition, our multilevel modeling analyses suggest that couples experienced more discrepancies in disclosures when they were revealing information to male social convoy members. These findings provide insight into the characteristics of disclosure discrepancies that occur between spouses coping with infertility: namely, discrepancies are likely to occur when one partner is particularly private and when disclosure targets are male.

Our model testing predictors of disclosure discrepancies yielded no significant findings for relational uncertainty, partner interference, or facilitation. Following CPM, we took discrepancies in reported disclosures to be a sign of boundary turbulence, and our descriptive analysis identified that such discrepancies do occur between partners coping with infertility. Contrary to our expectations, however, discrepancies in disclosures were not related to the relationship qualities we examined. We wonder if the assumption that disclosure discrepancies mark problematic boundary turbulence is incorrect. Perhaps spouses create mutually agreed rules that allow each of them different parameters for disclosure. In fact, the couple might collectively manage information by designating one partner as the interface with social networks, which aligns with our evidence that discrepant dyads typically included one non-disclosive spouse. To the extent that discrepancies in disclosive behavior reflect carefully negotiated and mutually recognized boundary rules, we would not expect these discrepancies to correspond with relational uncertainty or interference and facilitation from a partner.

Boundary turbulence was also operationalized by considering relative discomfort with spousal disclosures, which aligns with CPM’s idea of discomfort with co-owner privacy rules for control (Petronio, 2002). We found that relative discomfort was generally low in our sample, and husbands and wives did not experience significantly different levels of relative discomfort (RQ1). Although couples reported minimal relative discomfort in general, our sample encompassed the full range of our relative discomfort measure. In fact, approximately 15% of couples reported a discrepancy score either below −2 or above 2 (range = −4 – +4), indicating that a representative portion of our sample have couples with one partner reporting being completely or mostly comfortable with their own disclosures, whereas their spouse is not at all comfortable or barely comfortable with those disclosures. In other words, most couples experienced little relative discomfort; however, a small portion of our sample had high discrepancies between their own and their partner’s level of comfort in disclosures.

Analyses testing the associations between relative discomfort and marital characteristics found no significant associations. Although we found no significant associations between relationship uncertainty and relative discomfort, we found significant associations when self- and partner uncertainty were independently substituted in the model for relationship uncertainty. Specifically, positive associations were found between husbands’ self-uncertainty and husbands’ relative discomfort, as well as between both wives’ self-uncertainty and wives’ partner uncertainty and husbands’ relative discomfort. Although speculative, these findings suggest boundary turbulence in the form of relative discomfort may be more closely linked to individual uncertainties about the partnership, rather than ambiguity about the future of the relationship more broadly.

A third marker of boundary turbulence is finding disclosures to friends and family outside the marital dyad unacceptable; this marker corresponds with privacy rules for permeability (Petronio, 2002). Aligning with evidence that our sample had low levels of relative discomfort, individuals also perceived their partners to be largely approving of their disclosures to other parties (RQ1). At the same time, our sample also represented the entire range of our perceived acceptability measure. In fact, approximately 30% of couples reported perceiving that their spouse was not completely accepting of their disclosures to social convoy members.

Results from our multilevel models predicting perceived acceptability indicated that husbands perceived lower acceptance from their wives about their disclosures when either themselves or their wives were experiencing relational uncertainty. These results may reflect the pessimism bias documented by Knobloch, Bond, Miller, and Mannone (2007), such that relational uncertainty corresponds with more negative assessments of affiliation and involvement in a spouse’s messages. Why might this pattern emerge only for husbands’, and not wives’, perceptions of the acceptability of their disclosures? Although speculative, we note that husbands’ disclosures to the social network about sensitive, marital topics may be more counternormative than wives’ disclosures about such issues. In turn, when a relationship is characterized by uncertainty or doubts, this counternormative behavior is subject to greater scrutiny. We see exploring how relational uncertainty may promote more stylized and role-constrained communication behavior to be an interesting direction for future research.

One consistent finding was the lack of support found for the impact of partner interference or facilitation on boundary turbulence. We had reasoned that coordinating privacy boundaries was a facet of interdependence that would reflect the levels of coordination or disruption present in the relationship with respect to other activities; instead, a couple’s ability to manage everyday routines was unrelated to our measures of boundary coordination. Empirical evidence has documented a positive link between interference and direct communication about irritations within relationships (Theiss & Solomon, 2006a). Perhaps couples who experience interferences in their everyday routines are more likely to discuss discrepancies in disclosures about their infertility, and thereby limit boundary turbulence. Conversely, couples with facilitative patterns of interdependence might assume coordination in their disclosures to others, without engaging in conversations that would limit boundary turbulence. If born out in future research, these patterns highlight how experiences of interference, while disruptive in the short term, might prompt communication behaviors that promote coordinated behavior and limit privacy boundary violations. Of course, because we did not assess communication about privacy boundaries in this study, our suggestions on this point remain speculative.

Implications for CPM

This study was one of the first to our knowledge to empirically test boundary turbulence, which presented the task of operationalizing this phenomenon. Drawing from the conceptual description of boundary turbulence and the coordination processes surrounding linkages, control, and permeability (Petronio, 2002), we measured misaligned disclosure behaviors, discrepancies in comfort with disclosure behaviors, and perceptions of approval related to disclosure behaviors. The lack of findings associated with the first two representations of boundary turbulence may be a reflection of poor operationalization, or it could point to a need to reconsider the conception of boundary turbulence. We interpret a core assumption of CPM to be that boundary turbulence is a result of unsynchronized privacy rules; however, our findings suggest that partners who disclose at different levels or are differentially comfortable with disclosure may not necessarily be at odds with each other. At a broad level, we assumed that coordination implied using similar rules for disclosure (for some exceptions, see Petronio, 2002, p. 128). Specific to the marital context and co-owned information, we drew from the idea that “couples expect that disclosure will take place and that coordination of the privacy rules concerning ownership and permeability will be synchronized” (Petronio, 2002, p. 142). Our findings, which do not associate relational uncertainty and partner interference with disclosure discrepancies and relative discomfort, suggest that perhaps synchronization of rules is not necessary to avoid boundary turbulence. Although speculative, our findings suggest that boundary turbulence may be avoided through coordinating agreed rules that are different for each spouse, rather than synchronized ones. This point, which may be implied in previous writing about CPM, merits further attention teasing out the extent to which boundary turbulence lies within the mutuality of rules for disclosure or the similarity of standards for disclosure by the co-owners of information. In contrast to our two markers of boundary turbulence that tested for synchronization in disclosure (disclosure discrepancies) and synchronization in comfort (relative discomfort), our measure of perceived approval or disapproval associated with disclosures may capture the subjective experience of boundary turbulence that is more relevant to understanding couple functioning. In particular, our findings suggest that relational ambiguities may coincide with people perceiving that they violated a privacy rule, but not necessarily with an actual violation. These findings may also imply that concern about boundaries, rather than actual discrepant behaviors, represents perceptions of boundary turbulence in ways that impact the functions of a couple. Although replication is needed, these results challenge the way we define boundary turbulence both conceptually and empirically.

Relatedly, we see a need to reconsider the impact of boundary turbulence on relational and personal well-being. Much of the discussion related to boundary turbulence focuses on the problematic components stemming from not coordinating privacy rules; however, these rule breakdowns can be “instructional even if they are embarrassing, awkward, tricky, unpleasant, unnerving, and distressing” (Petronio, 2002, p. 204). Similarly, couples may learn how to manage their different preferences for privacy rules in socially healthy ways. To the extent that spouses are differently motivated to talk about private experiences, and have different levels of comfort with those disclosures, opportunities for the more disclosive partner to talk with family or friends may help that spouse meet communication needs without burdening the less disclosive partner. In this manner, confiding in a friend may fill a void a person could feel because of his or her partner’s reluctance to talk about the topic. We are encouraged that exploring the benefits of differential disclosure, in addition to the turbulence it might create, can offer a more complete understanding of information management within relational systems.

At a broader level, we also wonder about the function of privacy in situations of co-ownership. Within CPM, private information is viewed as a possession and a person who owns this information is entitled to make decisions about who they think should and should not know this information (see Petronio & Durham, 2008). According to CPM, disclosure of an infertility diagnosis to family and friends means that those recipients are now co-owners of that information as well; however, while those recipients may feel that the information is private, they may not think that the information belongs to them. More recent literature on privacy management suggests the idea that recipients, especially in family situations, may be told information and viewed as guardians of those private details (Petronio, 2010). This refinement suggests that couples may share information with family and friends, but view others as guardians rather than co-owners, with perhaps a limited vote on the rules and regulations surrounding that private information. Put more simply, the notion that private information is a possession may function most dominantly at the original owner’s level, and then future recipients of that information may perceive themselves to be guardians of private information that is owned by someone else.

From a broader perspective, we also wonder about the relationship between privacy and sensitivity. In this study, we focused on infertility, because it is a topic that some couples report withholding from friends and family, but also because it encompasses sensitive topics such as reproductive activities and personal limitations. Our results indicated no significant patterns for spouses disclosing information at different levels, but we did find evidence that relational uncertainty heightened a husband’s perceptions that his disclosures were not acceptable. We wonder if sensitive, stigmatizing, or identity-damaging information is the domain where attitudes about disclosure matter, irrespective of whether that information has been widely shared or kept under wraps. Put more simply, just because everyone knows a couple is infertile does not make their experience any less private, and does not mean that the dynamics of privacy management and coordination are not salient for them. We wonder if private information that is perceived by people as sensitive and perhaps stigmatizing intensifies the conceptual framework offered by CPM.

As a final point, we note the implications of our findings for couples coping with and counselors treating infertility. First, husbands in our sample perceived less acceptability regarding their disclosures to network members when either they themselves or their wives were feeling ambiguous about the relationship. Although these observations require replication, they underscore how marital characteristics can coincide with communication outside the partnership and suggest that conversations surrounding relationship functioning should address both marital and social network communication. Another interesting finding is that there were no significant associations between relational characteristics and actual discrepancies in disclosures. In the cases where couples do have conversations about plans for privacy management, it may be helpful to note that privacy rules may not be the same for each spouse within the relationship. In fact, couples may benefit from specifying rules for each individual partner based on comfort levels of each spouse. In addition, rules may function more efficiently when they are unique to each salient social network member, rather than applied globally to all people external to the marriage.

Limitations and future directions

The results associated with relational uncertainty are qualified by the relatively low level of uncertainty within this sample. One explanation for the limited relational uncertainty in this sample is that subjects who self-select themselves to take part in a research study are often the same partners who are experiencing marital stability. An alternative possibility is that the relational uncertainty measures we employed are not tapping the appropriate sources of ambiguity within a marriage. Recent research (Knobloch, 2008) examined the substance of relational uncertainty within the more established context of marriage to see if the sources of uncertainty were different from dating relationships – the original dyadic context the theory addressed. Of the 12 content areas found to contribute to relational uncertainty in marriage, only one addressed a commitment to the relationship, the rest focused on children, finances, family members, and other relational issues. Knobloch (2008) asserted that, whereas the content of ambiguity within dating relationships was focused on the establishment and sustenance of the relationship, most of the sources of ambiguity within a marriage may actually stem from the external pressures on the relationship. These findings suggest that measuring relational uncertainty in different contexts may require unique measures tailored to the dyadic context being investigated.

Results from this study may substantiate the idea that using a common scale across different relationship domains does not appropriately capture relational uncertainty. In a qualitative study of discourse within online forums dedicated to infertility support, relational uncertainty was prominent in themes involving invalidation of the marriage and feelings of blame connected with infertility (Steuber & Solomon, 2008). Feelings of ambiguity about their relationship permeated through the discourse for each topic. Despite the presence of relational uncertainty themes in the qualitative study, quantitative results from this study, which used the original measure focused on commitment to the relationship, suggest that infertile couples in this sample are coping with low amounts of uncertainty. Adjusting the relational uncertainty measure to the marital context is one strategy for better indexing the phenomena observed by Steuber and Solomon (2008).

To close, this project investigated how relational characteristics correspond with manifestations of boundary turbulence between infertile spouses who co-own private information about their reproductive disability. Although this study is situated within the infertility context, the findings may provide insight into how relational uncertainty and partner interference impact disclosure between spouses more generally. Future studies should continue to uncover relational characteristics that may contribute to boundary turbulence, as well as whether it is best to capture boundary turbulence based on behavioral discrepancies or discrepancies in comfort levels of privacy management. Finally, forthcoming research on CPM and marriage should unravel how turbulence functions within marriage by examining not only how it may disrupt a partnership, but ways in which it may provide some benefit.

Footnotes

Acknowledgements

This research was a portion of the first author’s dissertation conducted under the direction of the second author. A version of this manuscript was presented at the 2010 National Communication Association Conference in San Francisco, CA. The authors would like to thank the couples who participated in this research study.

The author(s) declared no conflicts of interest with respect to the authorship and/or publication of this article.

This project was supported by Funds from Hellene Runtagh and The Don and Carol Miller Graduate Funds through the College of Liberal Arts at The Pennsylvania State University.

Notes

References

Afifi

T. D.

Caughlin

Afifi

W. A.

(2007). Exploring the dark side (and light side) of avoidance and secrets. In Spitzberg

Cupach

(Eds.), The darkside of interpersonal relationships (3rd ed.). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.

Afifi

W. A.

Burgoon

J. K.

(1998). “We never talk about that”: A comparison of cross-sex friendships and dating relationships on uncertainty and topic avoidance. Personal Relationships, 5, 255–272.

Afifi

W. A.

Caughlin

(2006). A close look at revealing secrets and some consequences that follow. Communication Research, 33, 467–488.

Afifi

W. A.

Guerrero

L. K.

(2000). Motivations underlying topic avoidance in close relationships. In Petronio

(Ed.), Balancing the secrets of private disclosures (pp. 165–180). Mahwah, NJ: Lawrence Erlbaum Associates.

Afifi

W. A.

Reichert

(1996). Understanding the role of uncertainty in jealousy experience and expression. Communication Reports, 9, 94–103.

Baxter

L. A.

Wilmot

W. W.

(1985). Taboo topics in close relationships. Journal of Social and Personal Relationships, 2, 253–269.

Berger

C. R.

Bradac

J. J.

(1982). Language and social knowledge: Uncertainty in interpersonal relations. London: Edward Arnold.

Berscheid

(1983). Emotion. In Kelley

H. H.

Berscheid

Christensen

Harvey

Huston

T. L.

Levinger

(Eds.), Close relationships (pp. 110–168). San Francisco, CA: Freeman.

Bute

(2009). “Nobody thinks twice about asking”: Women with a fertility problem and requests for information. Health Communication, 24, 752–763.

10.

Bute

Vik

T. A.

(2010). Privacy management as unfinished business: Shifting boundaries in the context of infertility. Communication Studies, 61, 1–20.

11.

Byrk

A. S.

Raudenbush

S. W.

(2002). Hierarchical linear models: Applications and data analysis methods. Thousand Oaks: Sage.

12.

Greene

(2009). An integrated model of health disclosure decision-making. In Afifi

T.D.

Afifi

W.A.

(Eds.), Uncertainty, information management, and disclosure decisions (pp. 226–254). New York: Routledge.

13.

Greil

A. L.

(1991). Not yet pregnant: Infertile couples in contemporary America. New Brunswick, NJ: Rutgers University Press.

14.

Hendrick

S. S.

Dicke

Hendrick

(1998). The relationship assessment scale. Journal of Social and Personal Relationships, 15, 137–142.

15.

Kelley

H. H.

Berscheid

Christensen

Harvey

J. H.

Huston

T.L.

Levinger

McClintock

Peplau

L. A.

Peterson

D. R.

(1983). Analyzing close relationships. In Kelley

H. H.

Berscheid

Christensen

Harvey

Huston

T. L.

Levinger

McClintock

Peplau

L. A.

Peterson

D. R.

(Eds.), Close relationships (pp. 20–67). New York: Freeman.

16.

Kahn

R. L.

Antonucci

T. C.

(1980). Convoys over the life course: Attachments, roles, and social support. In Baltes

P. B.

Brim

O. G.

, (Eds.), Life-Span development and behavior. New York, NY: Academic Press.

17.

Knobloch

L. K.

(2008). The content of relational uncertainty within a marriage. Journal of Social and Personal Relationships, 25, 467–495.

18.

Knobloch

L. K.

Carpenter-Theune

K. E.

(2004). Topic avoidance in developing romantic relationships: Associations with intimacy and relational uncertainty. Communication Research, 31, 173–205.

19.

Knobloch

L. K.

Miller

L. E.

Bond

B. J.

Mannone

S. E.

(2007). Relational uncertainty and message processing in marriage. Communication Monographs, 74, 154–180.

20.

Knobloch

L. K.

Miller

L. E.

Carpenter

K. E.

(2007). Using the relational turbulence model to understand negative emotion within courtship. Personal Relationships, 14, 91–112.

21.

Knobloch

L. K.

Solomon

D. H.

(1999). Measuring the sources and content of relational uncertainty. Communication Studies, 50, 261–278.

22.

Knobloch

L. K.

Solomon

D. H.

(2002a). Information-seeking beyond initial interaction: Managing relational uncertainty within close relationships. Human Communication Research, 28, 243–257.

23.

Knobloch

L. K.

Solomon

D. H.

(2002b). Intimacy and the magnitude and experience of episodic relational uncertainty within romantic relationships. Personal Relationships, 9, 457–478.

24.

Knobloch

L. K.

Solomon

D. H.

(2003). Manifestations of relationship conceptualizations in conversation. Human Communication Research, 29, 482–515.

25.

Knobloch

L. K.

Solomon

D. H.

(2004). Interference and facilitation from partners in the development of interdependence within romantic relationships. Personal Relationships, 11, 115–130.

26.

Knobloch

L. K.

Solomon

D. H.

(2005). Relational uncertainty and relational information processing. Questions without answers? Communication Research, 32, 349–388.

27.

Knobloch

L. K.

Solomon

D. H.

Cruz

(2001). The role of relationship development and attachment in the experience of romantic jealousy. Personal Relationships, 8, 205.

28.

Kreft

de Leeuw

(2002). Introducing multi-level modeling. Thousand Oaks, CA: Sage.

29.

Kunkel

A. W.

Burleson

B. R.

(1999). Assessing explanations for sex differences in emotional support. Human Communication Research, 25, 307–340.

30.

Lukse

M. P.

Vacc

N. A.

(1999). Grief, depression, and coping in women undergoing infertility treatment. Obstetrics and Gynecology, 93, 245–251.

31.

McLaren

R. M.

Solomon

D. H.

Priem

(2010). Explaining variation in contemporaneous responses to hurt in premarital romantic relationships: A relational turbulence model perspective. Communication Research, 35, 339–357.

32.

Park

(2002). Stigma management among the voluntary childless. Sociological Perspectives, 45, 21–45.

33.

Petronio

(2002). Boundaries of privacy: Dialectics of disclosure. Albany, NY: SUNY Press.

34.

Petronio

(2010). Communication privacy management theory: What do we know about family privacy regulation? Journal of Family Theory and Review, 2, 175–196.

35.

Petronio

Durham

W. T.

(2008). Communication privacy management theory. Significance for interpersonal communication. In Baxter

Braithwaite

D. O.

(Eds.), Engaging theories in interpersonal communication (pp. 309–322). Thousand Oaks, CA: Sage.

36.

Slater

Snyder

Hayes

A. F.

(2006). Thinking and modeling at multiple levels: The potential contribution of multilevel modeling to communication theory and research. Human Communication Research, 32, 375–384.

37.

Snijders

Bosker

(2003). Multilevel Analysis: An introduction to basic and advanced multilevel modeling. Thousand Oaks, CA: Sage.

38.

Solomon

D. H.

Knobloch

L. K.

(2004). A model of relational turbulence: The role of intimacy, relational uncertainty, and interference from partners in appraisals of irritations. Journal of Social and Personal Relationships, 21, 795–816.

39.

Solomon

D. H.

Weber

Steuber

K. R.

(2010). Turbulence in relational transitions. In Smith

Wilson

(Eds.), New directions in interpersonal research (pp. 115–134). Thousand Oaks, CA: Sage.

40.

Steuber

K. R.

Solomon

D. H.

(2008). Relational uncertainty, partner interference, and infertility: A qualitative study of discourse within online forums. Journal of Social and Personal Relationships, 25, 831–855.

41.

Steuber

K. R.

Solomon

D. H.

(2010). Managing information within social networks during infertility. In Miller-Day

(Ed.), Family communication, connections, and health transitions (pp. 297–322). New York: Peter Lang.

42.

Theiss

Solomon

D. H.

(2006a). Coupling longitudinal data and multilevel modeling to examine the antecedents and consequences of jealousy experiences in romantic relationships: A test of the relational turbulence model. Human Communication Research, 32, 469–503.

43.

Theiss

Solomon

D. H.

(2006b). A relational turbulence model of communication about irritations in romantic relationships. Communication Research, 33, 391–418.

44.

Theiss

Solomon

D. H.

(2007). Communication and the emotional, cognitive, and relational consequences of first sexual encounters in heterosexual dyads. Communication Quarterly, 55, 179–206.

45.

Willits

Crider

(1986). Well-being at midlife. Agricultural Experiment Station Bulletin 863. University Park, Pennsylvania State University.