Abstract
Research links open communication and self-disclosure to a host of beneficial outcomes in romantic relationships, including better relationship quality, relationship satisfaction, feelings of closeness and commitment, and relationship longevity. However, learning some information about a partner may evoke negative consequences. In two studies, we examined the types of information people may wish to avoid learning about their past or current romantic partners (Study 1), and individual differences in avoidance of information about a current romantic partner (Study 2). Content-coding of data from Study 1 revealed that a majority (51%) of respondents wished to avoid information related to their partner’s history of sexual behavior and/or infidelity. In Study 2, participants who were older and who scored higher on measures of anxious and avoidant attachment were more likely to want to avoid information about their partner. Furthermore, replicating and extending Study 1, participants most wished to avoid information about their partners’ prior sexual behavior and/or infidelity, as well as their partners’ prejudices. We interpret our findings in terms of a threat-management framework for information avoidance, suggesting that threats to the self and relationship may influence the decision to avoid information about a romantic partner.
Disclosure and open communication can offer a foundation for intimate relationship success. Disclosing information to one’s romantic partner is linked to feelings of intimacy (Laurenceau et al., 1998), trust, love, and commitment (Sprecher & Hendrick, 2004), and improved relationship quality (Tan et al., 2012). Nevertheless, people may wish to avoid discussing some topics with their partners (e.g., past sexual partners) to mitigate conflict and maintain the relationship (Roloff & Ifert, 2000). The current study examined an understudied factor that might undermine disclosure and sharing in intimate relationships—the desire to avoid information about one’s romantic partner. We specifically explored (1) the types of information people would rather not know about their intimate partners, and (2) individual differences associated with the desire to avoid information about one’s intimate partner.
Information avoidance represents any “behavior designed to prevent or delay the acquisition of available but potentially unwanted information” (Sweeny et al., 2010, p. 341). Research suggests that people may avoid learning information that threatens a cherished world- or self-view (Dwyer et al., 2015; Howell et al., 2017), disrupts positive feelings (Kelly, 1996; Thompson et al., 2002), or obligates unwanted/onerous behavior (Howell & Shepperd, 2012). Other research suggests that people will avoid learning information if it is uncontrollable—that is, if they cannot change the outcome of that information (Howell & Shepperd, 2012; Melnyk & Shepperd, 2012). Despite a rich theoretical background, to our knowledge no study has explicitly examined information avoidance in intimate relationships.
What relationship information do people avoid?
Although research on avoidance in relationships is sparse, several theoretical perspectives suggest that people may be inclined to avoid certain classes or categories of information. We enumerate on two somewhat-overlapping classes of information people are inclined to avoid: information that could be threatening to (1) the self and (2) the relationship.
First, people are likely to avoid information that could threaten the self, particularly information that threatens how they think, feel, or behave (Sweeny et al., 2010). For instance, people may be inclined to avoid information about their romantic partner if that information might undermine their view of themselves (e.g., someone who evaluates themselves as attractive avoiding learning that their partner finds them unattractive) or might cause negative emotions (e.g., avoiding learning information that might spur jealousy, guilt, or anger). Related to the notion that people want to avoid information that might threaten the self, people may be more likely to avoid information about their partner that is inconsistent with their beliefs about themselves or their partners. Research on cognitive dissonance and selective exposure (see Hart et al., 2009) suggests that people may also avoid information about their partner that evokes feelings of being wrong about their partners. For instance, people may purposely avoid information that suggests they made a wrong decision in becoming involved with a partner (Geher et al., 2005; Murray & Holmes, 1997). Indeed, research on information seeking styles indicates that people may be more likely to learn information about past partners if they are seeking reconciliation (Dillow et al., 2008). As an extension of this, people may avoid learning information about a past partner versus a current partner because it could undermine their belief about selecting suitable partners in the present and obstruct the process of “moving on” (Clayton et al., 2013; Marshall, 2012).
Second, people may be more likely to avoid information that could threaten the relationship or produce conflict. Research on romantic relationships suggests that partners often try to avoid or minimize conflict in their relationship by avoiding discussions that could stimulate an argument, especially one that would expose incompatibility and extend conflict into other areas of the relationship (Deutsch, 1973). For instance, people sometimes withhold complaints about their partners’ behavior in an effort to avoid relationship conflict (Afifi & Guerrero, 2000; Caughlin & Afifi, 2004; Cloven & Roloff, 1994; Roloff & Cloven, 1990). Relatedly, research on “taboo” discussion topics in romantic relationships suggests that the most avoided discussions in relationships are those that lead to conflict or might cause or relationship dissolution (Baxter & Wilmott, 1985). Thus, we suspect that people are likely to avoid learning information that could elicit conflict in their relationship (e.g., learning whether their partner has sexually fantasized about someone else) than to avoid more benign information that is unlikely to damage the relationship (e.g., learning about their partner’s favorite food).
Who avoids relationship information?
People differ in the propensity to avoid information, with some people displaying a greater tendency than other people to avoid information (Howell et al., 2018; Howell & Shepperd, 2016). Although we know of no studies linking individual differences to information avoidance in romantic relationships, research linking individual differences to information sharing in romantic relationships implicate five individual differences in information sharing: attachment style, gender, partner gender, age, and commitment level of the relationship. Given the overlapping themes between information sharing and information avoidance, these same individual differences may correspond with information avoidance in romantic relationships. We examine each in turn.
Attachment style
Research documents three universal attachment “styles” between children and their primary caregiver: anxious, avoidant, and secure attachment (Bowlby, 1969/1982). People high in anxious attachment cling to their primary caregiver in new situations and show outward signs of distress (i.e., tantrums) when their primary caregiver leaves them alone. Children high in avoidant attachment show indifference and little external reaction to their primary caregiver when their caregiver leaves them alone. Both the anxious and avoidant styles indicate a lack of felt security with the primary caregiver. Conversely, securely attached children use their caregivers for emotion regulation and as a secure base for exploration. They show distress at separation, but calming when reunited with their caregiver (Ainsworth et al., 1978).
Because bonding with intimate partners exhibits similarities to bonding between parent and child, researchers have investigated these three styles of attachment in the context of adult romantic relationships (e.g., Cassidy, 2000; Hazan & Shaver, 1987; Shaver et al., 1988). Measures of adult attachment style generally describe attachment as existing along two continuums: anxiety (i.e., a fear of abandonment or betrayal) and avoidance (i.e., mistrust of a partner, uneasiness with emotional closeness). More securely attached adults tend to be confident in their partner’s availability and are comfortable with emotional intimacy, scoring low on both continuums (Cassidy, 2000).
Although we know of no research examining attachment styles and information avoidance, we reasoned that attachment would relate to information avoidance based on work demonstrating that attachment styles represent different mental models of the self and others (Mikulincer & Shaver, 2007). Attachment anxiety is associated with a negative mental representation of the self and a positive mental representation of others, whereas attachment avoidance is associated with a positive mental representation of the self and a negative mental representation of others (Mikulincer & Shaver, 2007). As a result, people who are anxiously attached tend to prioritize the needs of other people above their own, whereas people who are avoidantly attached tend to prioritize themselves and their needs over those of others, and emotionally distance themselves from situations that may otherwise promote bonding (e.g., deep conversations with their partners; Mikulincer & Shaver, 2007). By contrast, people who are securely attached have positive mental models of the self and others, and thus are more likely to trust and bond with others.
In the context of information avoidance, higher attachment anxiety may be linked with a strong desire to learn all available information about their partners in an effort to learn how to maintain their partner’s needs and interest, and ultimately prevent abandonment. Higher attachment avoidance, on the other hand, may be linked with wishes to avoid learning information about their partner as a strategy to avoid intimacy and emotional closeness. Although more securely attached individuals may feel comfortable learning most information about their partners, they may be more likely to understand their partner’s boundaries. Indeed, research finds that secure attachment corresponds with greater self-disclosure (Tan et al., 2012). Thus, securely attached partners may be open to learning the majority of information about their partners.
Gender
Research examining gender differences in managing conflict within romantic relationships reveals that women are more likely than are men to engage in relationship maintenance, including actively engaging in discussions about relationship problems (Rusbult et al., 1986). In other domains, however, women appear more inclined than men to avoid information, such as information about health (Emanuel et al., 2015). These findings lead us to speculate that women in general are more likely than are men to avoid information, and therefore more likely than men to avoid information about their romantic partner. Another study, however, found that gender was unrelated to willingness to learn whether one’s partner was unfaithful (Howell & Shepperd, 2016). In sum, the literature has mixed findings relating to gender differences in information avoidance and thus we offer no predictions.
Partner gender
In general, women are more likely than are men to disclose information (Tardy & Smithson, 2018). It is possible, then, that partners of women are more accustomed to disclosure from their partner, and thus less likely to avoid information about their female partner.
In addition, patterns of avoidance might differ between same-gender and mixed-gender relationships. From an opportunity perspective, people exclusively interested in same-gender relationships may have fewer romantic alternatives than people interested mixed-gender relationships, simply because adults primarily desiring same-gender romantic relationships comprise only 4.5% of adults in the US (Gallup, 2017). The availability of partners may influence commitment to a relationship. Indeed, the Investment Model asserts that people feel committed to a relationship to the extent that fewer or worse partner alternatives are available (Rusbult et al., 1986). With a limited number of alternative romantic partners, people exclusively interested in same-gender relationships may be more likely than people interested in mixed-gender relationships (Vencill & Wiljamaa, 2016) to display information avoidance to avoid confronting news that could lead to the dissolution of the relationship. Conversely, same-gender couples may experience greater ease in conversation and intimacy (Umberson & Kroeger, 2016; Umberson et al., 2015a, 2015b). Accordingly, people in same-gender relationships may be more open to learning information about their partner. Given the competing findings, we made no predictions about whether same-gender and mixed-gender couples would display different levels of avoidance.
Relationship status
People in more formalized romantic relationships (e.g., cohabitating, married) have more relationship investment at stake than do people in less formalized relationships (e.g., casual dating; Kelley & Thibaut, 1978; Sprecher, 2001; Weigel, 2008). Just as the availability of partners may influence commitment to a relationship, investment into a relationship may influence commitment to that relationship. According to the Investment Model, people feel committed to the extent that they have invested resources in the relationship (Rusbult et al., 1986). This investment likely translates into less willingness to dissolve the relationship. As a consequence, some information (e.g., information about a partner’s infidelities) may have negative utility and may serve only to make people feel bad or sully their image of their partner (Ayres, 1983; Stafford & Canary, 1991). People in more formalized relations thus may be more inclined to avoid unwanted information about their relationship partner. By comparison, people who are in less formal relationships—perhaps because the relationship is still developing—may be disinclined to avoid information about their partner (Fox & Warber, 2014), particularly information that might suggest that their partner is a poor choice for a continued relationship. Consistently, we expected that people with a more formal relationship status (i.e., married or living as married) would generally be more likely to avoid information about their romantic partner than people in less formal relationships (i.e., dating).
Age
Older adults may display greater avoidance than younger adults for at least three reasons. First, older adults, much like people seeking same-gender partners, may feel unwilling to leave their romantic relationships because of a smaller pool of eligible romantic partners (Brown & Shinohara, 2013). Second, with time, people become more invested in and committed to their partner (Sternberg, 1986) and thus less willing to end their relationship. Other things being equal, the relationships of older adults have likely endured longer than the relationships of younger adults, leading to greater relationship commitment. Third, older adults may be more likely than younger adults to have better emotional regulation for disengagement from situations that may be unpleasant or offensive (Charles & Carstensen, 2008). Indeed, research suggests that effective emotion regulation may promote information avoidance (van ‘t Riet & Ruiter, 2013). We thus predicted that older people, compared with younger people, may be more inclined to avoid unwelcomed information about their relationship partner.
The present studies
In two studies, we examined information avoidance in the context of romantic relationships. In Study 1, we examined the nature of information people wish to avoid about a past or current partner. We predicted that people would display the greatest avoidance of information that is personally threatening (i.e., is inconsistent with one’s self-view or that threatens how one wishes to think, feel or behave) and information that might produce relationship conflict.
In Study 2, we examined the extent to which individual differences (i.e., attachment, demographics) moderate the desire to avoid information in a current romantic relationship. Consistent with research on threat and relationship disclosure, we predicted that participants who are not securely attached to their partner (i.e., people higher in attachment anxiety and/or avoidance), people in a more formal relationship status (i.e., married or living as married), and people who are older would report the greatest desire to avoid information about their romantic partners. We also explored the role of gender, partner gender, and their interaction in information avoidance, but offered no predictions. Last, in Study 2, we explored whether people are less likely to avoid information about their partner’s past behavior versus current behavior and whether avoidance differed by the categories of information identified in Study 1.
Study 1: Method
Participants and procedure
We recruited 142 participants (Ages 18–68 years, M = 26.98 years, SD = 8.30) using online social networking websites (e.g., Facebook, Craigslist) and email. Table 1 presents demographics of our sample. Participants were involved in at least one romantic relationship prior to taking the survey, and most (69%) indicated that they were currently in a romantic relationship. Participants who were in a relationship responded to questions about their current partner; participants not in a relationship responded to questions about their most recent past partner.
Study 1 demographic characteristics (N = 142).
After consenting to participate and completing a demographic questionnaire, participants responded to the question, “Is there some information that you would rather not know about your partner?” (yes/no). Participants then responded to an open-ended question, “Of all the information you don’t want to know about your partner, what piece of information ranks at the top (i.e., what information do you least want to know about your partner)? Please describe the information in the space below.” Participants then completed some additional measures that served as a pilot for methodology development for Study 2, but that we do not discuss here. All data and questionnaires are available online at: https://osf.io/6zfys/.
Study 1: Results
Across the sample, 34.8% (40 of 115 responding) of participants indicated that they wanted to avoid some piece of information about their partner. Proportionally more participants reported wanting to avoid information if they were responding about a past partner (57.6%: 19 of 33 participants) than if they were responding about current partners (25.6%: 21 of 82 participants), X 2(1, 115) = 10.60, p = .001, φ = .30. In a separate, open-ended question, all participants had the opportunity to report the piece of information that they would most wish to avoid about their partner, and 84.3% (97 of 115) of participants provided a response. Of the 97 people who provided a response, 64.9% (63 of 97) described a specific piece of information (the remainder provided responses such as “nothing”). Consistent with the earlier avoidance item, proportionally more participants writing about a past partner provided a text response indicating a piece of information they would rather not know (83.3%; 25 of 30) than did people writing about a current partner (56.7%; 38 of 67), X 2(1, 97) = 6.45, p = .011, φ = .26. See Table 2.
Avoidance of information about a current versus past partner in Study 1.
Figure 1 shows three word-clouds representing the open-ended responses about the type of information that people reported they would most want to avoid about their partner. The larger the word, the more frequently it appeared in participant responses. The figure depicts the top pieces of information reported by participants toward all partners (top panel), current relationship partners (bottom left), and past relationship partners (bottom right). Two raters content coded participant responses to the open-ended item. After reading the content, the raters working together identified three primary categories: 1) sexual behavior/infidelity, 2) relationship concerns, and 3) an “other” category. Table 3 provides example statements for each category. Each of the coders independently coded each statement in the relevant categories. When the coders disagreed, they discussed the coding until they reached agreement. Although we did not calculate inter-rate reliability at the time of coding, disagreement was rare.

Word clouds of top piece of information people wished to avoid about a partner in general (top panel), and a past (right bottom panel) or current partner (left bottom panel).
Study 1: Examples of the information people most want to avoid and the frequency of avoidance (n = 63).
Note. Percentages are for total sample, not of each category. Within columns the bolded percentages sum to 100%; the non-bolded percentages do not because some responses could be coded into multiple categories. Categories are in bold, subcategories are in regular typeface indented once, subcategories within these subcategories are indented twice.
The top half of Table 3 shows the major categories of information participants wished to avoid learning based on the content coding of responses. Within some of these categories were subcategories. Sexual thoughts and behavior included responses that were about sexual thoughts in general (e.g., fantasies); sexual behavior prior to, during, or since the relationship (e.g., other partners); infidelity (e.g., kissing another person). The coders also identified and coded for four sub-categories of sexual thoughts and behavior: infidelity specifically (which had three subcategories: sexual infidelity, emotional infidelity, and a desire for another), sexuality (e.g., that their partner was attracted to a gender different than their own), sexual history (e.g., their partner’s number of past partners), and other sexual thoughts and behavior (e.g., masturbation).
Relationship concerns included responses that involved how a partner thought about the participant or their relationship. The coders identified and coded for three sub-categories: other relationships, partner’s negative perceptions, and other relationship concerns. The other relationships sub-category primarily involved wanting to avoid knowing how their partner saw them or their mutual relationship in comparison to other romantic partners or other romantic relationships. The partner’s negative perceptions category primarily involved learning that the partner viewed the participant negatively (e.g., seeing them as unattractive). The other relationship concerns only had one item, concerning pregnancy, which did not fit with the other subcategories.
The miscellaneous category comprised topics that were infrequent and did not fit into the other two primary categories or their subcategories. Grouping the topics beyond a miscellaneous category was not feasible because no two responses were similar. Responses in this category included “drug use/criminal activity during our relationship” and “how many times she looked at my Facebook profile when I wasn’t blocking her.”
The bottom half of Table 3 shows the frequency with which each of the content coding categories appeared both overall, and for current and past partners. The results were generally consistent across past and current partners. Examining the frequency with which each of the categories appears revealed that the majority of the responses were sexual behavior/infidelity (51.8%; e.g. “details about their past sexual experiences”)—particularly infidelity (e.g., “if he cheated on me”). The next most common theme was relationship concerns (33.7%; e.g. “that she loved her ex more than me”).
Study 1: Discussion
The results of Study 1 indicate that participants wished to avoid many types of information about current and past romantic partners. The information that participants most frequently wished to avoid was information about sexual thoughts, behavior, or infidelity. Additionally, information avoidance was proportionate greater among partners who rated a past partner than participants who rated a current partner. These findings provide an initial indication about the types information people wish to avoid and “who” avoids information about their romantic partner—specifically people reporting on current partners avoided information about their partners less than people reporting on past partners.
Study 2
Study 2 expanded on Study 1 in four ways. First, most participants in Study 1 (86%) were in mixed-gender relationships. In Study 2, we explicitly recruited more people in same-gender relationships. Second, in Study 1, we asked participants to report the top piece of information they wished to avoid about their romantic partner. Although this question provided insight into the types of information people might avoid, it provided little information about the extent to which people would avoid a variety of types of information. In Study 2, we examined the extent to which people wished avoid a variety of different pieces of information about their partner. Third, in Study 1, we examined both people in current and past relationships. In Study 2, we exclusively examined people who were currently in a relationship and asked them questions about their current relationship. We chose to devote our resources to focus exclusively on current romantic partners because we believed information avoidance is more meaningful and can have more direct consequences for ongoing relationships conflict people (Caughlin & Afifi, 2004; Deutsch, 1973). Finally, in Study 2 we provide a more in-depth investigation of the relationship between individual differences and information avoidance.
Study 2: Method
Participants
We recruited 305 adults who reported on a past or current romantic relationship recruited via Facebook, email, and Amazon.com’s Mechanical Turk. We paid participants recruited through Mechanical Turk USD$0.51. Although 121 (39.7%) of the participants responded regarding a past relationship, we did not include them within the analyses because the temporal nature of the items to which they responded did not align with the purpose of the current study. Nonetheless, we provide their data on the OSF link. For the scope of the current study examining information avoidance in ongoing romantic relationships, we analyzed data from the 184 adults in a romantic relationship. Demographic information about the sample appears in Table 4.
Study 2: Frequencies and descriptive statistics on participants’ demographic characteristics (N = 184).
Measures
After consenting to participate, participants read a list of 61 pieces of information they might wish to avoid about their current partner. We generated the list based on the content coding from Study 1 and from an ad hoc focus group of adults with relationship experience. Participants indicated the extent to which they wanted to know each piece of information about their current partner on a scale ranging from 1 (definitely want to know) to 7 (definitely do not want to know). For each item, participants could also select “I already know this information.” We recoded these values as missing, because these participants could no longer seek or avoid that piece of information.
Supplemental Appendix A (see OSF link) contains a full list of the 61 pieces of information we evaluated and descriptive information for each item. Of note, the desire to avoid information on most items was relatively low (i.e., most means were below the scale midpoint), consistent with past literature that people generally want to know information about their partners. We categorized the items based on Study 1 content-coding and discussion among the authors: 1) sexual behavior/infidelity (e.g., “My partner was sexually promiscuous before our relationship”), 2) relationship concerns (e.g., “My partner thinks poorly of me now”), 3) illegal or troublesome behavior (e.g., “My partner used illegal drugs during our relationship”), 4) partner’s prejudice (e.g., “My partner believes that his/her racial group is superior to others”), and 5) health concerns (e.g., “My partner currently has a mental illness”). Within each category, we grouped the items into past and present information. Past information items were items that could have happened prior to the current relationship or early in the current relationship but had ceased since (e.g., “My partner was sexually promiscuous before our relationship” and “My partner masturbated earlier during our relationship but does not now”). Present information items were items described as current or ongoing (e.g., “My partner is currently having sexual relations with someone outside our relationship”).
Next, participants completed the Relationship Structures questionnaire, which included 10-items assessing adult attachment style (ECR-RS; Fraley et al., 2000). The ECR-RS includes subscales assessing anxious attachment and avoidant attachment styles, with higher values indicating greater anxious or avoidant attachment. The five-item attachment anxiety dimension represents the extent to which people tend to worry about availability or responsiveness of their partner (e.g., I’m afraid this person will abandon me; α = .90; M = 2.46, SD = 1.58). The five-item avoidance dimension represents the extent to which people are uncomfortable opening up to or depending on their partner (e.g., I usually discuss my problems & concerns with this person [reverse-coded]; α = .83 M = 2.15, SD = 1.12). For both dimensions, lower scores indicated more secure attachment. We adhered to best practice recommendations to examine anxious and avoidant dimensions separately despite the fact that they are highly correlated (e.g., Beyderman & Young, 2016; Goodall, 2015; Neave et al., 2016). Finally, participants responded to demographic questions about their age, gender, partner’s gender, race, and relationship status (married, living as married, engaged, dating, other).
Study 2: Analyses & results
Who avoids information?
We examined the extent to which individual differences predicted the desire to avoid information about one’s partner using multilevel modeling. We nested responses to each item and the temporal context in which the information described in the item occurred (past = −.5, present = −.5) (Level 1) within participants (Level 2). At Level 2, we included avoidant attachment and anxious attachment separately (grand-mean centered), age (grand-mean centered), gender (male = −.5, female = .5), relationship status (dating/other = −.5, married/living as married = .5), partner gender (male = −.5, female = .5), and interaction of gender × partner gender to predict avoidance. Avoidant and anxious attachment styles were highly correlated, r(174) = .57, p < .001, so we examined them in separate models. We allowed the intercept of information avoidance and the slope of the past versus present comparison to vary randomly between participants at Level 1. That is, we allowed each participant to have his or her own mean level of avoidance and own slope of the discrepancy between past and present information.
Using a multilevel approach allowed us to estimate the effect of the individual differences on information avoidance broadly even though we had missing data resulting from participants already knowing some information about their partner. Using an alternative approach that required listwise deletion, like MANOVA, would only allow participants who did not know any information about their partner to inform the estimates—thus severely limiting sample size and biasing the estimates.
The first data column in Table 5 shows the results of a multilevel model predicting information avoidance (Level 1) from anxious attachment, age, gender, relationship status, partner’s gender, the interaction of gender and partner-gender, and whether the information represented a past or present occurrence. The second data column shows the same model but examined avoidant attachment rather than anxious attachment in the model. In both analyses, higher scores on our measures of anxious and avoidant attachment style corresponded with greater information avoidance. Participants were also more likely to avoid information if they were older and if the information was about the past rather than present. Gender, relationship status, partner gender, and the interaction of gender and partner gender did not significantly predict avoidance.
Study 2: Multilevel models predicting information avoidance.
Note. Age and attachment are grand-mean centered. We used effects coding on all categorical variables: gender (male = −.5, female, .5), relationship status (dating/other = −.5, married/living as married = .5), partner’s gender (male = −.5, female, .5), interaction of gender × partner’s gender (−.25 to .25), past or current behavior (past = −.5, current = .5). Estimates from the model examining anxious attachment as a predictor appear on the left column. Estimates from the model examining avoidant attachment appear on the right column.
What information is avoided most?
To examine differences in the types of information that participants wanted to avoid, we used participants’ ratings of the information from the five categories to rank order the categories according to the information participants most wanted to avoid to the information participants least wanted to avoid (order: partner prejudice, sex/infidelity, other relationship concerns, illegal/troublesome behaviors, health concerns). Based on this ranking, we created a series of Helmert contrast codes (see Table 6), to examine the extent to which these categories differed.
Helmert contrast codes to examine differences in avoidance by information category.
Note. Categories appear in order from highest avoidance mean (partner prejudice) to lowest avoidance mean (health concerns).
Using these codes, we compared the avoidance of information from a target within our rank of categories to the mean avoidance of the categories above the target category. Finally, we computed interactions between these contrast codes and our individual difference variables to examine whether the types of information people wanted to avoid varied as a function of the significant person-level predictors (i.e., avoidant and anxious attachment, age). We entered the past versus present comparison and the non-significant person-level (i.e., gender, partner gender, gender × partner gender, and relationship status) variables as covariates in the model. None of the interactions with the contrast codes significantly predicted avoidance. As such, we present the simpler model here and include the full interaction results in the supplemental table (see OSF link).
As evident in Table 7, the multilevel model revealed that the level of avoidance reported in a target category was significantly different from the mean level of avoidance reported in the categories above the target category with one exception. The exception was that participants did not differ in their avoidance of information about their partner’s sexual behavior/infidelity and information about their partner’s prejudice (the top two categories).
Study 2: Multilevel model predicting information avoidance (Level 1) from contrast codes for each type of information (Level 1).
Note. Age and attachment are grand-mean centered. All categorical items effect coded: gender (male = −.5, female, .5), relationship status (dating = −.5, married/living as married = .5), partner’s gender (male = −.5, female, .5), interaction of gender × partner’s gender (−.25 to .25), past or current behavior (past = −.5, current = .5). Estimates from the model examining anxious attachment as a predictor appear on the left column. Estimates from the model examining avoidant attachment appear on the right column. Models include covariates for individual differences (Level 2) and whether the information is a past or current behavior (Level 2).
Study 2: Discussion
The results of Study 2 reveal attachment style and age predict information avoidance. Participants reported a greater desire to avoid information about their partner if they were older and if they reported having an anxious or avoidant attachment style. Additionally, the results suggest that participants most wished to avoid information about their partner’s prejudices and sexual behavior/infidelity (similar to Study 1), followed by information about relationship concerns, illegal/troublesome behavior, and health concerns. Finally, Study 2 revealed that participants were more likely to avoid learning information about their partner’s past behavior than information about their partner’s present behaviors.
General discussion
We conducted two studies that examined the type of information people wish to avoid about their current and past romantic partners (Study 1), and whether some people more than others desire to avoid information about a current romantic partner (Study 2). Study 1 revealed that participants reported most wanting to avoid information related to their partner’s past and current sexual behaviors, including infidelity. We replicated this finding in Study 2, showing that participants most wanted to avoid learning about their partner’s past and current sexual behavior and about their partner’s prejudices. The latter finding did not emerge as a theme in Study 1. In Study 1, participants generated specific information they would be most likely to avoid, while in Study 2 we provided specific information. We suspect that participants may not explicitly think to avoid their partners’ prejudices, but when presented as a type of information to avoid, participants may perceive it as a potential threat to the relationship.
The finding that participants reported feeling particularly inclined to avoid information about their partners’ sexual history aligns with research suggesting that people are particularly likely to avoid information that may lead to relationship conflict (Afifi & Guerrero, 2000; Deutsch, 1973). Knowing about a partner’s personal sexual thoughts, behavior (e.g., masturbation), or infidelity may expose the relationship to vulnerability and dissolution (Brase et al., 2014). Furthermore, proportionally, participants reporting about their past romantic partner indicated a greater desire to avoid learning information than people reporting about their current partner. We suspect people may wish to avoid learning new information about a past partner because it may undermine their own belief about their ability to select appropriate future partners(s), and even obstruct the process of “moving on” (Clayton et al., 2013; Marshall, 2012), especially for very recent separations.
In Study 2, participants who were older and who scored higher in anxious or avoidant attachment reported a stronger desire to avoid information about their romantic partner. As we suggested in the introduction, the age effect may reflect the fact that fewer eligible romantic partners are available if the relationship dissolves (Brown & Shinohara, 2013), or that older adults feel more committed to their relationship compared with younger adults (Sternberg, 1986). Furthermore, older adults’ desire to avoid information about their partners may reflect increased emotional regulation as way to disengage from unwanted or onerous information (Charles & Carstensen, 2008; van ‘t Riet & Ruiter, 2013).
Regarding attachment, the greater avoidance we observed among participants higher in anxious or avoidant attachment may stem from the distinct motivations underlying these two styles. Specifically, participants higher in anxious attachment report greater insecurities surrounding their partner’s availability and about potential abandonment (e.g., Cassidy, 2000; Shaver et al., 1988). These participants may wish to avoid learning more information about their partner to prevent exposure to information that may exacerbate their negative mental representation of the self (see Mikulincher & Shaver, 2007). Participants higher in avoidant attachment, conversely, may wish to avoid more information about their partner because disclosure among partners often corresponds with increasing emotional closeness, an aspect of relationships with which people with higher avoidant attachment may be uncomfortable particularly given their negative mental representations of attachment figures (Laurenceau et al., 1998; Mikulincher & Shaver, 2007). Stated otherwise, our results show that participants lower in avoidant or anxious attachment (i.e., more secure attachment) wish to avoid less information about their romantic partners. People with more secure attachment styles are more comfortable with emotional intimacy, which can result from self-disclosure (Tan et al., 2012). Additionally, the felt security that comes with secure attachment may correspond with a sense of being able to endure negative information and move past it.
Other individual differences (i.e., gender, partner’s gender, whether the relationship was same- or different-gender and relationship status) did not predict avoidance of information about a relationship partner. The null effect for relationship status was surprising and may reflect the fact that it represents an imprecise measure of relationship commitment.
Study 2 also revealed that people are more likely to avoid information related to their partner’s past (i.e., past sexual experiences) than their partner’s present thoughts and behaviors. Participants may have avoided their partners past thoughts and behaviors for two reasons. First, they might aim to idealize their partner’s past and avoid information that could disrupt this idealized view (Geher et al., 2005). Second, people may view information about the past to be less useful because it may do little more than evoke regret or feelings of personal failure. Discovering information about a current partner’s past may diminish what makes the current relationship unique or exceptional (Frampton & Fox, 2018; White & Mullen, 1989).
Limitations and future directions
The present work provides insights into information avoidance in romantic relationships. Nevertheless, it is limited in at least two ways. First, in Study 2, many participants already knew some of the information we asked about their partner, limiting our ability to examine whether participants would wish to avoid the information. Although we used multilevel modeling to address issues arising from missing data, it is possible that our results would have differed were all participants naïve about these topics. Second, our data are correlational and thus we cannot establish a causal link between the individual factors examined and information avoidance. Although it is not possible to experimentally manipulate the individual differences we examined, future studies can examine whether changes in some of our individual difference variables (e.g., age, relationship commitment) precede changes in relationship information avoidance.
Despite its limitations, the present work has implications for future research on romantic relationships and information avoidance. First, our findings provide a novel examination of the types of information people wish to avoid in romantic relationships and who prefers to avoid relationship information. To date, most research has focused on communication and healthy disclosure (Gottman, 1993; Laurencau, et al., 1998; Tan et al., 2012) or taboo conversation topics (Kelly, 1996; Roloff & Ifert, 2000)—thus examining communication from the perspective of information providers. Second, as far as we are aware, our study is the first to examine information sharing from the perspective of the recipient; highlighting what information people would rather not know about their partners. It provides a promising initial step in understanding an additional factor that might undermine conversation and disclosure. Future research can bring these lines of inquiry together to examine the consequences of information avoidance for disclosure and communication, and for relationship maintenance.
Future research can also expand the present study to account for the dyadic nature of romantic relationships. In the present study, we collected responses from only one partner in each romantic partnership. Future studies including both partners can examine the dynamic interplay of information avoidance tendencies in a relationship as well as how individual differences in the desire to avoid information influence relationship communication behaviors like disclosure. Additionally, future research might examine how past relationships influence current relationship information avoidance. In our first study, we included participants who responded about their most recent past partner if they were not currently in a relationship along with participants who responded about their current relationship. Using the relational uncertainty perspective (Knobloch & Carpenter-Theune, 2004), future studies might examine the extent to which avoiding unwanted information about a past partner may alter the course and development of intimacy and certainty within future relationships.
Conclusion
The present research is among the first to examine information avoidance in the context of romantic relationships. We found that relationship partners wished to avoid information about sexual behavior/infidelity, information about past partners, and information about a partner’s past thoughts/behaviors. Further, we found that age and attachment style moderated information avoidance in relationships. Our findings offer an important first step in understanding information avoidance in romantic relationships and in so doing expands both the information avoidance and romantic relationships literatures.
Supplemental material
Supplemental Material, Supplementary_Appendix_A - Avoiding information about one’s romantic partner
Supplemental Material, Supplementary_Appendix_A for Avoiding information about one’s romantic partner by Maryam Hussain, Devon M. Price, Amanda N. Gesselman, James A. Shepperd and Jennifer L. Howell in Journal of Social and Personal Relationships
Footnotes
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental material
Supplemental material for this article is available online.
Open research statement
As part of IARR’s encouragement of open research practices, the author(s) have provided the following information: This research was not pre-registered. The data used in the research are available. The data can be obtained at:
or by emailing: mhussain5@ucmerced.edu. The materials used in the research are available. The materials can be obtained at: https://osf.io/6zfys/?view_only=bb87cf7d9eab43ca828b61e8473bfc7b or by emailing: mhussain5@ucmerced.edu.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
