Abstract
Young adults’ romantic relationships have continually emerged as a significant area of study in psychological research. The authors examine possible associations between a measure of attitudes toward sibling relationships, a measure of adult attachment styles, and a measure of romantic relationship quality in a sample of 350 Italian university students (68.6% females, 31.4% males; age range 19-30 years, M age = 23.6 years, SD = 3.2). The authors tested the following hypotheses: that the perceived relationship quality between siblings was positively correlated with adult attachment styles and romantic relationship quality, that the frequent use of behaviors associated with adult attachment styles was negatively correlated with romantic relationship quality, and that attitudes toward sibling relationships and adult attachment styles predicted romantic relationship quality. Findings only partly supported the hypotheses, indicating that more positive attitudes toward sibling relationships and lower use of behaviors associated with avoidant attachment style influenced and predicted higher romantic relationship quality.
Keywords
Introduction and Literature Review
Young adults’ romantic relationships have continually emerged as a key topic and area of study in psychological research, with an increasing interest, starting from the beginning of 2000 (e.g., Crouter & Booth, 2006; Fincham & Cui, 2011; Scharf & Mayseless, 2001), in comprehending their antecedents and the processes that contribute to their quality. Moreover, as indicated by Erikson (1968), young adulthood represents a crucial and critical time for the development of romantic ties (e.g., W. A. Collins & Van Dulmen, 2006; Meier & Allen, 2009; Noller, Feeney, & Peterson, 2001; Seiffge-Krenke, 2003). Indeed, the formation of healthful and satisfying romantic relationships contributes to individuals’ psychological well-being and behavioral adjustment, as well as to their social competence (e.g., Arnett, 2004; Davila et al., 2017). In particular, romantic relationships in young adulthood, characterized by identity exploration, self-focus, and instability (e.g., Meier & Allen, 2009; Reifman, 2011), are important for three substantial reasons: (1) the formation and maintenance of romantic ties is one of the critical developmental tasks of this life phase, (2) the formation and maintenance of romantic ties have positive effects on young adults’ well-being and behavioral adjustment; and (3) the formation and maintenance of romantic ties have a predictive role for later relationships and marriage (e.g., Fincham & Cui, 2011). In accordance with this understanding, the literature has strongly emphasized the importance of parent–child and peer relationships for later romantic relationships (e.g., Cohn, Silver, Cowan, Cowan, & Pearson, 1992; Conger, Cui, Chalandra, & Elder, 2000; Crockett & Randall, 2006; Donnellan, Larsen-Rife, & Conger, 2005; Zimmer-Gembeck, 2002). In particular, research findings (e.g., Black, & Schutte, 2006) indicate that subjects reporting a more positive and loving relationship with parents are more trusting and are more comfortable relying on their partners. Furthermore, research findings on the positive and negative emotional and psychological characteristics of friendship—such as support, trust, intimacy, conflict, and so on—indicate that friendship, and more generally peer relationships, play an important role in determining individual social development and adjustment, as well as in determining the quality of later romantic relationships (e.g., Brumbaugh, 2017; Fraley & Davis, 1997; Ponti, Guarnieri, Smorti, & Tani, 2010; Scharf & Mayseless, 2001).
Research has generally stated that experiences in childhood and adolescence are possible predictors of emerging adulthood romantic relationships, but little attention has been paid to the influence of sibling experience on the quality of adult romantic ties (e.g., Brumbaugh, 2017; Doughty, McHale, & Feinberg, 2015; Doughty, Lam, Stanik, & McHale, 2015; Rauer & Volling, 2007; Robertson, Shepherd, & Goedeke, 2014), thus indicating the need for further investigations.
Sibling Relationships, Adult Attachment Styles, and Romantic Relationship Quality
The sibling bond represents a unique, influential tie, and it is the most long-lasting and, statistically, the most enduring relationship to develop throughout the life span (e.g., Cicirelli, 1995). While initial research in the field has focused the most attention on child and adolescent sibling relationships (e.g., Furman & Buhrmester, 1985; Stocker, Dunn, & Plomin, 1989), increasing attention is now being dedicated to adult sibling relationships from a life span perspective (e.g., Donsì, Parrello, Grossi, & Ragozini, 2010; van Volkom, 2006; Whiteman, McHale, & Soli, 2011). In particular, siblings serve as companions, confidants, and role models in childhood and adolescence (e.g., McHale, Updegraff, & Whiteman, 2012), as a source of emotional support and security in the face of parental conflict, and at different points across the life span (e.g., Waite, Shanahan, Calkins, Keane, & O’Brien, 2011).
Adult sibling ties are generally characterized by less conflict, more warmth, less contact, and less proximity (e.g., Conger & Little, 2010), due in part to the specific developmental tasks involved in the transition to emerging adulthood (such as leaving home, completing education, becoming employed, and the ensuing focus on new social relationships outside the family; e.g., Arnett, 2004). The literature also suggests that the sibling bond may continue to be characterized by ambivalent feelings in adulthood (e.g., Myers & Goodboy, 2010; Tani, Guarnieri, & Ingoglia, 2013). Along the same lines, research findings generally indicate that the adult sibling bond is characterized by a decrease in sibling rivalry for parental attention and affection (e.g., Suitor & Pillemer, 2007). However, there are some studies showing that the perception of parental favoritism may continue to influence the siblings’ relationship, specifically affecting warmth, conflict, hostility, and jealousy (e.g., Finzi-Dottan & Cohen, 2011), as well as individuals’ general well-being (e.g., Pillemer, Suitor, Pardo, & Henderson, 2010).
Moreover, research has focused on the effect of experiences within the family and young adult outcomes (e.g., Pauldine, Snyder, Bank, & Owen, 2015; Young & Ehrenbery, 2007), and only a few studies have specifically investigated the link between within-family experiences, especially with parents and siblings, and adult romantic relationships. Rauer and Volling (2007) found a positive correlation between experiencing differential parental treatment, sibling jealousy, attachment styles, and adult romantic relationship quality. Brumbaugh (2017) found that attachment-related feelings are especially likely to transfer from sibling relationships to new social partners, such as romantic partners. Furthermore, Schwarz, Mustafić, and Junker (2015), found that siblings’ attachment functions are transferred to romantic partners who, with age, move to the top of the attachment hierarchy (e.g., Fraley & Tancredy, 2012). Doughty, Lam, et al. (2015) also reported a correlation between the experience of sibling relationships and young adults’ perceived skills in romantic relationships. Finally, although Robertson et al. (2014) did not find a correlation between the quality of sibling relationships and romantic relationship satisfaction, the authors did find evidence of interaction between birth order and romantic relationship length and satisfaction, concluding that sibling attachment might be one of a large set of factors contributing to romantic relationship quality.
Additionally, from a clinical point of view, the international literature suggests that in partner choice and in the development of romantic relationships, an important role may be played by some patterns of the sibling relationships, indicating a sort of transference of these patterns from siblings to romantic partners (e.g., Abend, 1984; Sommantico, 2018a, 2018b; Kriss, Steele, & Steele, 2014; Morley, 2006).
If, on one hand, these contrasting data confirm the hypothesis that siblings may influence attachment function in adulthood (e.g., Feeney & Humphreys, 1996; Trinke & Bartholomew, 1997; Whiteman et al., 2011), on the other, they suggest the necessity of continued studies of sibling relationships, also taking into account within-family experiences, and their role in adult romantic relationship quality.
Regarding this last point, starting with the pioneering research of Hazan and Shaver (1987; Shaver & Hazan, 1988), romantic love has been conceptualized and considered as an attachment process. The authors of other studies from the same time period have demonstrated that attachment patterns found in infant–caregiver relationships (e.g., Ainsworth, 1989; Bowlby, 1988) can be applied in investigations of adult love experiences. Since these seminal studies, several authors have begun investigating romantic attachment (for a review, see Feeney, 2016), hypothesizing that it can be better conceptualized in terms of two fundamental dimensions: anxiety and avoidance (e.g., Brennan, Clark, & Shaver, 1998; Mikulincer & Shaver, 2016). Avoidant attachment style is associated with the minimization of attachment needs due to expectations of refusal/rejection by the romantic partner, which also indicates individuals’ discomfort in close relationships and with interdependence. In contrast, anxious attachment style is associated with a hyperactivation of the attachment system, to obtain the attention of partners felt to be inconsistently available or abandoning. Therefore, while high levels of attachment anxiety are characterized by a fear of abandonment, desire for fusion with the partner, and frustration caused by a perceived lack of closeness, high levels of attachment avoidance are characterized by feeling uncomfortable with closeness and being reluctant to get close to others.
The literature has shown the relevance of these two fundamental dimensions in predicting relationship quality, satisfaction, and commitment (e.g., N. L. Collins & Read, 1990; Kirkpatrick & Davis, 1994; Li & Chan, 2012; Mikulincer, Florian, Cowan, & Cowan, 2002; Sandberg, Bradford, & Brown, 2017). But while there is an enormous amount of research that focuses on adult couple relationship quality (measuring adjustment, satisfaction, outcomes, etc.), and which is subsumed under the general category of “research on marital quality” (e.g., Alexandrov, Cowan, & Cowan, 2005; Bradbury, Fincham, & Beach, 2000; Rossier, Rigozzi, Charvoz, & Bodenmann, 2006), only a few authors have investigated young adults’ romantic relationship quality (e.g., Black & Schutte, 2006; Ponti et al., 2010; Rauer & Volling, 2007; Stackert & Bursik, 2003), thus indicating the need for further investigation.
Finally, research suggests the importance of also taking into consideration the role of structural variables—such as sex, sibling gender constellation, age, number of siblings, birth order, and parents’ marital status—in influencing young adults’ sibling relationships and romantic ties. Indeed, research findings indicate the following: sisters report the most positive sibling relationships, with subjects (especially females) reporting their most significant relationship with a sister (e.g., Milevsky, Smoot, Leh, & Ruppe, 2005; Spitze & Trent, 2006); older siblings report the most positive sibling relationships, as do participants with smaller sibships (e.g., Milevsky et al., 2005; van Volkom, Machiz, & Reich, 2011); and parental separation/divorce is associated with lower relationship satisfaction and commitment (e.g., Rhoades, Stanley, Markman, & Ragan, 2012; Cui & Fincham, 2010).
The Current Study: Hypotheses and Conceptual Model
Based on the above literature review, this exploratory study analyzes the possible associations between a self-report measure of attitudes toward the sibling relationship, a measure of attachment styles in adult romantic relationships, and a measure of adult romantic relationship quality. Furthermore, this study investigates the potential role that structural variables, such as age, sex, birth order, siblings’ gender constellation, and parents’ marital status (married/separated/divorced), may play in this association.
Indeed, as indicated by W. A. Collins and Sroufe (1999), we have assumed that sibling relationships can play an important role as attachment figures, as well as in the formation of the attachment system (e.g., Brumbaugh, 2017; Cassidy, 2016; Doherty & Feeney, 2004; Schwarz et al., 2015; Tibbetts & Scharfe, 2015), also considering that significant past experiences, such as those with siblings, can be predictors of future relationships’ quality and success (e.g., Doughty, McHale, et al., 2015; Doughty, Lam, et al., 2015; Hartsthorne, Salem-Hartsthorne, & Hartsthorne, 2009; Robertson et al., 2014; Stackert & Bursik, 2003).
Thus, we hypothesize the following:
Based on the existing literature (e.g., Doughty, McHale, et al., 2015; Milevsky et al., 2005; Spitze & Trent, 2006; Stocker, Lanthier, & Furman, 1997; van Volkom, et al., 2011), we also hypothesize the following:
Therefore, we have also formulated a conceptual model hypothesizing the following:
Method
Recruitment, Data Collection Procedures, and Participants
Participants were recruited from the University of Naples. All data were collected with self-report questionnaires using an Internet-based survey (e.g., Hewson, Vogel, & Laurent, 2016), according to the following criteria: All participants must be older than 18 years, have at least one sibling, and have been in a stable couple relationship for at least 6 months. Participants gave consent to participate in the study on the first page of the survey, which took approximately 15 to 20 minutes to complete. A basic demographic questionnaire was completed on the next page of the survey, collecting information regarding: participant’s age, participant’s sex, number of siblings, birth order, parents’ marital status (married/separated/divorced), and romantic relationship’s duration. Participants were asked to answer by reporting their relationship with the sibling that has had the greatest impact (positive or negative) on their lives (indicating his or her sex and age), and by reporting their relationship with their current romantic partner.
The next pages of the survey consisted of a presentation of three different instruments, in the following order: Lifespan Sibling Relationship Scale (LSRS; Riggio, 2000); Experiences in Close Relationship–Revised (ECR-R; Fraley, Waller, & Brennan, 2000); Romance Qualities Scale (RQS; Ponti et al., 2010; for a detailed description of the instruments, see the following section). Participation in the study was voluntary and anonymous, and participants were encouraged to answer as truthfully as possible.
The demographic characteristics of study participants are presented in Table 1. The sample consisted of 350 students (68.6% females, 31.4% males), aged from 19 to 30 years (M = 23.6, SD = 3.2). The majority of participants (63.4%) have one sibling (44% were first born; 39.4% were second born, and 16.6% were third born or more). Same-sex siblings’ pairs were 51.7%, while mixed-sex siblings’ pairs were 48.3%. Participants who referred to a sister were 55.1%, and those referring to a brother were 44.9%. Married parents were 88.9%, while those separated or divorced were 11.1%. The majority of participants (66.3%) lived with their sibling/s at the time of the study. Engaged participants were 66.3%, while participants cohabiting with their romantic relationship partner were 33.7%. The average duration of the romantic relationship, ranging from 6 to 60 months, was 23.8 months (SD = 13.4). In our sample, there were no married young adults, and only 4% of the participants declared a same-sex romantic relationship.
Demographic characteristics of study participants (N = 350).
Measures
Lifespan Sibling Relationship Scale
The LSRS is a 48-item self-report instrument (Riggio, 2000; Italian adaptation and validation: Sommantico, Donizzetti, De Rosa, & Parrello, 2017) which assesses the feelings, behaviors, and thoughts related to sibling relationships in childhood and adulthood on six subscales: (1) Adult Affect (Items 1-8; Item 6 “reverse keyed”), (2) Adult Behavior (Items 9-16; Item 13 “reverse keyed”), (3) Adult Cognition (Items 17-24; Item 19 “reverse keyed”), (4) Child Affect (Items 25-32; Items 25, 27, and 28 “reverse keyed”), (5) Child Behavior (Items 33-40; Item 38 “reverse keyed”), and (6) Child Cognition (Items 41-48; Item 42 “reverse keyed”). The affective component is operationalized by assessing emotions toward the sibling and the sibling relationship (e.g., “My sibling’s feelings are very important to me”). The behavioral component is operationalized by assessing the degree of interaction with the sibling (e.g., “My sibling and I ‘hang out’ together”). The cognitive component is operationalized by assessing beliefs about the sibling and the sibling relationship (e.g., “My sibling and I have a lot in common”). Participants are asked to respond according to a Likert-type scale ranging from 1 (strongly disagree) to 5 (strongly agree). Participants are also asked to answer as truthfully as possible, by referring to the sibling that has had the greatest impact (positive or negative) on their lives. Subscale scores are obtained by averaging the items that make up each subscale. Low scores on the six dimensions are indicative of a negative perception of past and current sibling relationships. The total score is obtained by averaging the six subscale scores. Riggio (2000) reported good internal consistency for each subscale (Cronbach’s α ranging from .87 to .91) and for the total score (.96). In the present study, Cronbach’s α ranged from .79 to .90 for each subscale and was .94 for the total score.
Experiences in Close Relationships–Revised
The ECR-R (Fraley et al., 2000; Italian adaptation and validation: Picardi, Bitetti, Pasquini, & Puddu, 2000) is a 36-item self-report questionnaire providing a dimensional evaluation of current attachment style in romantic relationships. The questionnaire produces scores on two subscales, corresponding to the two main orthogonal dimensions of romantic attachment styles: (1) Avoidance (AVO—18 odd-numbered items; e.g., “I prefer not to show a partner how I feel deep down”): difficulties in establishing a deep intimate connection with the partner, and a general feeling of discomfort with intimacy (Items 3, 15, 19, 25, 27, 29, 31, 33, and 35 are “reverse keyed”) and (2) Anxiety (ANX—18 even-numbered items; e.g., “I often worry that my romantic partner doesn’t really love me”): worry about one’s romantic relationship, involvement in the relationship, fear of abandonment, and frequent seeking of the partner’s attention and care (Item 22 is “reverse keyed”). Participants are asked to indicate their feelings regarding their current couple relationship according to a Likert-type scale ranging from 1 (totally false) to 7 (totally true). Subscale scores are obtained by averaging the items that make up each subscale. Low scores on both dimensions are indicative of relatively secure attachment. Especially relevant for clinical use, it is possible to individuate four attachment categories by matching the scores on the two dimensions (or styles; Bartholomew & Horowitz, 1991): Secure, Preoccupied, Fearful, and Dismissing. However, as recommended by the ECR-R authors (Fraley et al., 2000), in this study, we have used the dimensional evaluation in order to more easily observe individual differences, and to compare them with a categorical evaluation. Internal reliability is α = .90 for AVO and for ANX scales, while external validity was shown by correlation with other instruments measuring relationship satisfaction, adult attachment measures, psychopathology, and personality. In the present study, internal reliability was .91 for AVO scale and .84 for ANX scale.
Romance Qualities Scale
The RQS (Ponti et al., 2010) is a 22-item self-report instrument assessing qualitative aspects of romantic relationships. The instrument was created as an equivalent version of the Friendship Qualities Scales (Bukowski, Hoza, & Boivin, 1994). The RQS assesses five fundamental dimensions of romantic relationships: (1) Companionship (Items 1, 4, and 14) refers to the time voluntarily spent together (e.g., “My partner thinks of fun things for us to do together”); (2) Conflict (Items 9, 11, 18, and 19) refers to the frequency of disagreements (e.g., “My partner can bug me or annoy me even though I ask him not to”); (3) Help (Items 3, 5, 10, 12, and 15) refers to mutual aid and to protective attitudes toward injustice and oppression (e.g., “My partner would stick up for me if someone was causing me trouble”); (4) Security (Items 2, 13 “reverse keyed,” 16, 17, and 20) refers to relationship reliability (e.g., “If I have a problem at school, work, or at home, I can talk to my partner about it”); and (5) Closeness (Items 6, 7, 8, 21, and 22) refers to emotional connection and attachment (e.g., “I feel happy when I am with my partner”). Participants are asked to respond, focusing on their current couple relationship, according to a Likert-type scale ranging from 1 (absolutely false) to 5 (absolutely true). Subscale scores are obtained by summing the items that make up each subscale. Low scores on the Companionship, Help, Security, and Closeness dimensions are indicative of a negative perception of romantic relationship quality, while low scores on the Conflict dimension are indicative of a positive perception of romantic relationship quality. The total score is obtained by summing up the six subscale scores. The authors (Ponti et al., 2010) reported good internal consistency for each subscale (Cronbach’s α ranging from .61 to .84). In the present study, Cronbach’s α ranged from .64 to .83 for each subscale and was .88 for the total score.
Data Analysis Plan and Procedure
In order to verify the strength of the association between the three instruments (Hypotheses 1, 2, and 3), we carried out Pearson r correlation analyses. Correlation coefficients between .10 and .29 represent a small association, while coefficients between .30 and .49 represent a medium association, and coefficients of .50 and greater represent a large association or relationship. To thoroughly verify Hypotheses 1 and 3, and to verify Hypotheses 4 and 5, we also carried out analyses of variance (ANOVAs; p < .05). Furthermore, eta-square (η2) was used to measure the effect sizes or the ANOVA, which, in line with Cohen’s (1988) guidelines, should be interpreted as follows: small ≥0.01; medium ≥0.059; large ≥0.138. In order to verify Hypothesis 6, we carried out a hierarchical multiple regression analysis (Method: Enter; p < .05). The analyses of correlation, variance, and regression were calculated by using IBM SPSS.18 software (Coakes & Ong, 2011). Finally, in order to thoroughly verify Hypothesis 6, we used structural equation modeling. Model fit was assessed with unweighted least squares in Lisrel 8.54 (Jöreskog & Sörbom, 2004). Goodness-of-fit of the model was assessed as indicated by a χ2/df ≤ 3, a standardized root mean square residual (SRMR) ≤.08, comparative fit index (CFI) >.90, and a nonnormed fit index (NNFI) >.90. Results of the root mean square error of approximation (RMSEA) are considered to be good if they are <.05 and reasonable if they are <.08. RMSEA < .05. Because the different indices assess different aspects of goodness-of-fit, evaluating multiple fit indexes simultaneously is recommended (Hu & Bentler, 1995; Ullman, 2001). Satisfactory models should show consistently good-fitting results on many different indices.
Results
Means and Standard Deviations
Total scores, means, standard deviations, and Cronbach’s α of the three instruments are presented in Table 2: The mean for the total LSRS score was 163.4 (SD = 26.5); the means for ECR-R scales were 3.3 (SD = 0.4) for the Avoidance scale and 4.1 (SD = 0.6) for the Anxiety scale; and the mean for the total RQS was 86.7 (SD = 10.5).
Descriptive Statistics for the Lifespan Sibling Relationship Scale (LSRS), the Experiences in Close Relationships–Revised (ECR-R,) and the Romance Qualities Scale (RQS) (N = 350).
Note. AA = Adult Affect; AB = Adult Behavior; AC = Adult Cognition; CA = Child Affect; CB = Child Behavior; CC = Child Cognition; LSRS = total LSRS score; AVO = Avoidance; ANX = Anxiety; COM = Companionship; CON = Conflict; HEL = Help; SEC = Security; CLO = Closeness; RQS = total RQS score.
Correlations
Zero-order correlations between the instruments are shown in Table 3.
Correlations between the Lifespan Sibling Relationship Scale (LSRS), the Experiences in Close Relationships–Revised (ECR-R), and the Romance Qualities Scales (RQS) (N = 350).
Note. AA = Adult Affect; AB = Adult Behavior; AC = Adult Cognition; CA = Child Affect; CB = Child Behavior; CC = Child Cognition; LSRS = total LSRS score; AVO = Avoidance; ANX = Anxiety; COM = Companionship; CON = Conflict; HEL = Help; SEC = Security; CLO = Closeness; RQS = total RQS score.
p < .05. **p < .01.
Results indicate small negative correlations (r ranging from −.12 to −.17; p ranging from .05 to .01) or no correlations between the Avoidance and LSRS subscales and the total score, as well as small negative correlations (r ranging from −.11 to −.15; p ranging from .05 to .01) or no correlations between the Anxiety and LRSR subscales and the total score. These findings indicate a minor relationship between attitudes toward past and current sibling relationships and adult attachment styles, thus only partly supporting Hypothesis 1.
Moreover, the results indicate small positive (r ranging from .11 to .29; p ranging from .05 to .01) and negative (r ranging from −.13 to −.17; p = .01) correlations or no correlations between LSRS subscales/total score and RQS subscales/total score. These findings only indicate a minor relationship between attitudes toward past and current sibling relationships and romantic relationship quality, thus only partly supporting Hypothesis 3.
Finally, the results indicate moderate negative correlations (r ranging from −.30 to −.41, p = .01), and a small positive correlation (Conflict subscale: r = .19, p = .01) between Avoidance and RQS subscales/total score. Furthermore, the results indicate small negative (r ranging from −.15 to −.22, p = .01) or positive correlations (Conflict subscale: r = .21, p = .01), and no correlation (Closeness subscale: r = −.06) between the Anxiety and RQS subscales/total score. These results suggest a moderate negative relationship between avoidant adult attachment style and romantic relationship quality, thus only partly supporting Hypothesis 2.
Groups’ Differences
ANOVA showed significant gender differences. Regarding the LSRS, female participants reported significantly higher scores than male participants on the total LSRS score (MF = 165.6 vs MM = 158.7; F(1, 349) = 5.096, p < .05; η2 = .01). In addition, participants reporting their relationship with a sister showed higher scores than participants reporting on brothers on the total LSRS score (MS = 166.0 vs MB = 160.1; F(1, 348) = 4.310, p < .05; η2 = .01). Furthermore, Sister–Sister siblings’ pairs reported significantly higher scores than other siblings’ pairs (Sister–Brother, Brother–Sister, and Brother–Brother) on the LSRS total score (MSS = 166.9, MSB = 163.8, MBS = 164.0, and MBB = 152.1; F(3, 346) = 3.896, p < .01; η2 = .03).
Regarding ECR-R, no gender differences were found.
Regarding RQS, female participants reported higher levels on the Help (MF = 22.1 vs. MM = 21.3; F(1, 349) = 5.750, p < .05; η2 = .02) and Security (MF = 19.6 vs. MM = 18.8; F(1, 349) = 4.375, p < .05; η2 = .01) subscales. Furthermore, Sister–Sister siblings’ pairs reported significantly higher scores than other siblings’ pairs (Sister–Brother, Brother–Sister, and Brother–Brother) on Help (MSS = 22.3, MSB = 21.8, MBS = 21.5, and MBB = 21.0; F(3, 346) = 2.845, p < .05; η2 = .02) and Security (MSS = 20.0, MSB = 19.1, MBS = 18.6, and MBB = 19.1; F(3, 346) = 3.896, p < .05; η2 = .03) subscales.
ANOVA and Tukey tests also showed significant birth order differences, but only for the LSRS. Indeed, third-born (or more) participants reported significantly higher scores than first- or second-born participants on LSRS total score (MI = 161.5, MII = 161.9, and MIII = 171.8; F(1, 349) = 3.519, p < .05; η2 = .02).
These results support Hypotheses 4 and 5, although they do not fully confirm the hypotheses, in light of the low effect sizes (Cohen, 1992), which indicate a slightly significant relationship between birth order, siblings’ gender constellation, attitudes toward sibling ties, and some dimensions of romantic relationship quality.
Furthermore, ANOVA showed significant relationship status differences (engaged/cohabiting) but only in terms of Avoidance. Indeed, engaged participants reported higher scores than cohabiting participants on Avoidance (ME = 3.3 vs MC = 3.2; F(1, 349) = 8.438, p < .01; η2 = .03).
ANOVA and Tukey tests also showed significant differences regarding LSRS total scores for participants’ adult attachment styles, and for romantic relationship satisfaction. Indeed, participants with more negative attitudes toward sibling relationships reported significantly higher scores on Avoidance (MN = 3.5, MM = 3.2, and MP = 3.2; F(1, 349) = 6.559, p < .01; η2 = .04) than participants with medium or positive attitudes toward sibling relationships. Furthermore, participants with more positive attitudes toward sibling relationships reported significantly higher scores on total RQS (MN = 82.1, MM = 86.6, and MP = 89.3; F(1, 349) = 10.988, p < .001; η2 = .06).
Finally, ANOVA showed significant differences regarding RQS for participants’ avoidant attachment style. Indeed, participants showing lower scores on Avoidance reported higher romantic relationship satisfaction (MI = 87.8 vs. MII = 82.3; F(1, 348) = 31.266, p < .01; η2 = .08).
These results partly support Hypotheses 1 and 3, indicating a significant relationship between attitudes toward past and current sibling relationships, avoidant attachment style, and romantic relationship satisfaction.
Contrary to our expectations, no differences were found regarding age, number of siblings, parents’ marital status (married/separated/divorced), and romantic relationships’ duration.
Regression Analysis
Based on previous results, hierarchical multiple regression analysis was conducted in order to determine the extent to which each predictor variable (age and gender, siblings’ gender constellation, birth order, romantic relationship status, LSRS, Avoidance) contributed to the model predicting young adults’ romantic relationship quality above and beyond the others (see Table 4). In the baseline step, we inserted age and gender, neither of which contributed significantly to the regression model. In the second to fourth steps, we consecutively inserted siblings’ gender constellation, birth order, and romantic relationship status, again not contributing significantly to the regression model. In the fifth step, the addition of LSRS contributed significantly to the regression model by increasing the accounted variance by 5.5% (β = .240; R2 = .051, p < .001), thus indicating that more positive attitudes toward sibling relationships predicted higher romantic relationship quality. In the sixth step, the addition of Avoidance contributed significantly to the regression model by increasing the accounted variance by 11.3% (β = −.348; R2 = .164, p < .001), thereby indicating that a massive use of behaviors associated with avoidant attachment style predicted lower romantic relationship quality. In the final step, the addition of Anxiety did not contribute significantly to the regression model. Age, gender, siblings’ gender constellation, birth order, romantic relationship status, LSRS, and adult attachment styles, combined, accounted for 18.3% of the variance in relationship quality, and for 16.9% of the variance when adjusted for sample size and the number of predictors.
Stepwise Hierarchical Multiple Regression Analysis Summary for Age and Gender, Siblings’ Gender Constellation, Birth Order, Romantic Relationship Status, Lifespan Sibling Relationship Scale (LSRS), Avoidance, and Anxiety Predicting Romantic Relationship Quality (N = 350).
p < .001.
These findings partly support Hypothesis 6, thus indicating that more positive attitudes toward sibling relationships and a lower use of behaviors associated with avoidant attachment style positively influence perceived romantic relationship quality and predict its scores.
Testing of the Hypothesized Conceptual Model
We used structural equation modeling to test the structural relationships.
The hypothesized model for predicting the quality of young adults’ romantic relationships was tested (see Figure 1) and the results confirmed our model, with strong fit indices. As hypothesized, attitudes toward sibling relationships were positively associated with romantic relationship quality. On the contrary, as expected, avoidance and anxiety had a negative relationship with romantic relationship quality.

Testing of the conceptual model for the prediction of young adults’ romantic relationship quality.
Results indicated a good fit between the theoretical and the empirical models: χ2(df) = 143.34(59), p = ns; χ2/df = 2.43; CFI = 1.00; NNFI = 1.02; RMSEA = .06; 90% confidence interval (CI) for RMSEA: lower limit = .05; upper limit = .08; SRMR = .067. These results strongly support Hypotheses 2, 3, and 6, indicating that overall attitudes toward sibling relationships, as well as adult attachment styles, predict the quality of young adults’ romantic relationships. In particular, a more positive attitude toward sibling relationships is correlated with higher levels of romantic relationship satisfaction, while massive use of behaviors associated with avoidant and anxious adult attachment styles is correlated with lower levels of romantic relationship satisfaction.
Discussion
Research has strongly highlighted the importance of romantic ties for socioemotional development throughout the life span, as well as the significance of family and peer experiences for individuals’ adult romantic relationships (e.g., Scharf & Mayseless, 2001; Seiffge-Krenke & Connolly, 2010). Little research, however, has examined the role of sibling ties in adult attachment styles and in romantic relationships. This study has allowed us to deepen our understanding of the relationship between perceived sibling relationship quality, adult attachment styles, and young adults’ romantic relationships.
Building on previous research on middle childhood and adolescence (e.g., Buhrmester & Furman, 1990; Dunn, Slomkowski, & Beardsall, 1994), our findings indicate that sister pairs reported significantly more positive overall attitudes toward siblings. In particular, our research findings expand this framework to young adulthood, demonstrating that, also in this phase, sister–sister pairs show the highest levels of warmth, intimacy, and closeness (e.g., Riggio, 2000; Spitze & Trent, 2006), most likely due to, among other factors, their higher perception of similarity.
Contrary to previous research findings (e.g., Del Giudice, 2011; Doughty, McHale, et al., 2015; Feeney, 2016; Finzi-Dottan & Cohen, 2011), no significant age or birth order differences in attitudes toward sibling relationships were found, nor were there significant gender differences in romantic attachment, nor siblings’ gender constellation differences in perceived romantic relationship quality. We can interpret these results by referring to the limitations of the present study (see the following paragraph), which are also indicated by the reported small or moderate effect sizes. In particular, we can hypothesize that the sample’s gender disequilibrium (68.6% females, 31.4% males) and the sample’s mean age may have had an uncontrolled effect on study results.
More significantly, our findings indicate a link between attitudes toward sibling relationships, and avoidant attachment style, which suggests that siblings may also play a role as attachment figures and thereby influence future attachment relationships (e.g., W. A. Collins & Sroufe, 1999; Kriss et al., 2014). We also found a significant relationship between attitudes toward sibling relationships and romantic relationship quality, suggesting the further hypothesis of a congruent conceptualization of sibling and romantic relationships (e.g., Doughty, McHale, et al., 2015; Doughty, Lam, et al., 2015). In this sense, our results indicate that sibling relationships may continue to play an important role in young adults’ socioemotional development. Moreover, in particular, it seems that, by representing a context for exploring more complex social interaction, sibling relationships can have significant influence on later relationships outside the family, such as peers and romantic ties (e.g., Lockwood, Kitzmann, & Cohen, 2001; Updegraff, McHale, & Crouter, 2002).
Finally, in line with previous research findings (e.g., N. L. Collins & Read, 1990; Feeney & Noller, 1990; Fraley & Davis, 1997; Gleeson & Fitzgerald, 2014; Holmes & Johnson, 2009; Simpson, 1990; Tucker & Anders, 1999), the results showed a significant relationship between avoidant attachment style and romantic relationship quality, indicating that adult attachment styles may play an important role in the quality of couple relationships. Indeed, as indicated by a meta-analysis by Li and Chan (2012), both anxiety and avoidance tend to be detrimental to relationship quality in the emotional, behavioral, and cognitive dimensions; in particular, avoidance was significantly more negatively related to relationship quality and its positive indicators than anxiety.
Strengths, Limitations, and Future Research
The principal strength of the present study is that it is one of the very few international studies, and the only Italian study, to investigate the relation between attitudes toward sibling relationships, adult attachment styles, and romantic relationship quality in young adults.
The first general limitation is related to sampling. Our sample only included university students, and future investigations might try to replicate findings with other emerging Italian adults, such as those who work. Moreover, community-based sampling implies specific possible biases: for example, volunteers’ bias (related to the special characteristics of individuals who voluntarily participate in a study). We can hypothesize that the small or moderate effect sizes reported in the present study have their origin in this general limitation. In this vein, future research could implement other sampling strategies (i.e., balancing sample gender composition, extending investigation to other Italian contexts).
Another possible bias in the study is that of mono-method, related to the fact that, having assessed all variables of the study by using self-report instruments, there could be an inflation in observed association. Moreover, the use of a self-report instrument that is also retrospective (LSRS) may have produced personal biases and faulty memories.
A future longitudinal research design could be envisaged, which would also allow for causal inferences to be carried out, which was not possible in this study due to its cross-sectional nature. Future investigations could also integrate quantitative with qualitative data, to deepen different aspects of the adult siblings’ relationships. Furthermore, since our sample only includes university students aged 19 to 30 years, future research could address samples of older adults (e.g., middle-aged or elder adults), in order to understand the highlighted changes in importance of sibling relationships, as well as in adult attachment styles and in romantic relationship quality throughout the lifespan. Finally, future research could address sibling and couple dyads, utilizing dyadic analyses, to deepen understanding of the interactional nature of sibling and romantic relationships.
Conclusion
Taken together, our results suggest that, as stated by Doughty and colleagues, affectional, behavioral, and cognitive dynamics learned within the family can be reflected in other significant relationships outside the family system (e.g., Fraley, & Davis, 1997), thus extending previous research that has indicated a connection between positive sibling relationships and social competence with friends and peers (e.g., Dunn, 2007; Lockwood et al., 2001; Sherman, Lansford, & Volling, 2006), as well as between experience with peers and romantic relationships (e.g., Brumbaugh, 2017; Connolly, Furman, & Konarski, 2000).
Our results also confirm the role of affectional, behavioral, and cognitive attitudes toward sibling relationships, as well as of avoidant attachment style in predicting romantic relationship quality (e.g., N. L. Collins, Cooper, Albino, & Allard, 2002; Doughty, Lam, et al., 2015; Li & Chan, 2012; Robertson et al., 2014; Stackert & Bursik, 2003).
In sum, our data indicate that sibling ties and adult attachment styles may play an important role in young adults’ romantic relationship satisfaction, demonstrating the necessity of further analyzing the impact of sibling relationships and attachment dynamics on young adults’ development, as well as of further investigating the transference mechanisms of attachment styles from family members, especially siblings, to significant others in future life.
Our findings also have practical implications. Indeed, by demonstrating the importance of positive attitudes toward siblings in the perception of romantic relationship quality, our results illustrate the importance of sibling relationships for romantic competence and satisfaction. In this sense, families and practitioners need to be aware of these implications for young adults’ development; promoting sibling intimacy, as well as reducing sibling conflict, could be envisaged as an important focus in prevention/intervention programs developed with the aim of promoting young adults’ positive development.
Furthermore, our results indicate and confirm the importance of focusing on attachment processes in therapy as a way to contribute to the building of security for individuals and couples in counseling and in therapy. The improvement of attachment-based interventions in individual and marital counseling and therapy, as well as the conceptualization of attachment needs and processes for individuals and couples in distress, may represent an opportunity for clinicians to provide more effective help, especially by referring to emotionally focused therapy (Johnson, 2004), which is the attachment-based model, which has the strongest empirical support.
Finally, according to research findings indicating that the forming and maintaining of satisfactory romantic relationships is one of the critical tasks of early adulthood (e.g., Fincham & Cui, 2011), our results indicate the necessity of implementing relationship education programs and interventions among emerging adults, thus facilitating their ability to initiate and sustain successful romantic relationships.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
