ATTACHMENT INSECURITY AND ITS ROLE IN INTERCULTURAL ROMANTIC RELATIONSHIPS

ATTACHMENT INSECURITY AND ITS ROLE IN INTERCULTURAL ROMANTIC

RELATIONSHIPS

INTRODUCTION

Intercultural Romantic Relationships: The Role of Attachment Insecurity

Intercultural romantic relationships—typically formed between individuals from different

racial or ethnic backgrounds (e.g., Gaines et al., 2006) but also extending to individuals of

differing cultural backgrounds that are less salient (e.g., Streib, 2015)—are more prevalent now

than ever before (Pew Research, 2019). This type of relationship is seldom the topic of research

and the studies that have been done include conflicting findings, which may be due to unique

factors that individuals bring to each type of relationship. My dissertation examined one potential

factor, attachment insecurity, which is robustly associated with relationship outcomes but has not

been examined specifically within intercultural romantic relationships. This dissertation was

intended as the first step toward understanding the role of attachment insecurity with

intercultural romantic relationship initiation and maintenance. These studies may contribute to a

better understanding of intercultural romantic relationships relative to intracultural romantic

relationships.

Attachment

Attachment develops in infancy as a mechanism that facilitates survival for children who

are vulnerable without the support of a caregiver. Children seek food, shelter, safety, and comfort

from their caregiver and the attachment bond allows children to rely on this caregiver for these

needs. It also helps children understand when the world is safe or unsafe to explore on their own.

Based on the pattern of caregiving that a child receives, different levels of attachment security or

insecurity develop (Ainsworth, 1979; Bowlby, 1968) and remain fairly stable across the lifespan

(see McConnell & Moss, 2011 for a review) and serve as a model for future romantic

relationships (Hazan & Shaver, 1987).

Attachment Security

Higher levels of attachment security are most adaptive and allow individuals to thrive in

their adult romantic relationships. Theoretically, higher levels of attachment security present as

individuals who feel safe enough to express their own needs and expectations about the

relationship, and who trust that their partner will be responsive when they need them. Individuals

who are higher in attachment security are more likely to experience higher quality relationships

(e.g., Collins & Read, 1990), to trust their partner (e.g., Mikulincer, 1998), and to use adaptive

strategies to effectively manage conflict (e.g., Pistole, 1989).

Two important aspects of attachment security are the extent to which an attachment

figure acts as a (1) safe haven and a (2) secure base. When an attachment figure serves as a safe

haven, the attached individual sees them as a space of protection and care. One may seek out a

romantic partner after a tough day at work or a fight with a friend to receive comfort and support

for their emotional distress. When an attachment figure serves as a secure base, the attached

individual feels safe to explore their world away from the attachment figure. This is because the

reliable care that the attached individual has learned to expect leads them to trust that care will

consistently be available when called upon and the individual can safely explore their world

away from their secure base. These two factors are a hallmark of higher levels of attachment

security. However, when one develops an attachment that does not have a high level of security,

higher attachment insecurity may develop, which carries forward from childhood into adult

attachment relationships (e.g., Hazan & Shaver, 1987).

Attachment Insecurity

In social psychology, attachment insecurity is sometimes conceptualized as higher levels

of attachment anxiety or higher levels of attachment avoidance. Theoretically, individuals can be

high on both attachment anxiety and avoidance, but these constructs are most productively

understood separately, including in this dissertation, because they are theoretically and

conceptually distinct. In contrast, individuals who are low on both attachment anxiety and

avoidance are higher in attachment security.

As pioneered by Mary Ainsworth (1985; 1978) and since supported in the literature by

numerous studies, there are specific patterns of behaviors that children and later adults exhibit

when they develop higher levels of each domain of attachment insecurity. Higher levels of

attachment anxiety develop when a child experiences inconsistent caregiving such as when a

caregiver is sometimes responsive to needs but at other times unresponsive to needs. Under these

conditions, a child learns to be hypervigilant, worried, and fearful; they are unsure of whether

they will be comforted and adequately cared for so they may bid for caregiving in highly

distressing ways in an attempt to connect. Higher levels of attachment avoidance develop when a

child consistently experiences rejection and/or their needs are dismissed by their caregiver.

Under these conditions, a child learns that it is difficult rely on close others, to express emotions,

or to ask for support.

Different studies have offered support for these patterns in adulthood, which has been

synthesized in a textbook on adult attachment (Mikulincer & Shaver, 2007a). This paragraph will

summarize this synthesis. In adulthood, individuals who are higher in attachment anxiety worry

about the availability of their partner, and worry about whether their partner will be responsive,

aloof, or abandon them (e.g., break-up with them). Individuals higher in attachment anxiety are

particularly challenged by uncertainty and sensitive to relationship threat and are frequently

hypervigilant of things that could threaten their relationship (e.g., they experience high levels of

jealousy or anxiety). They also have a poor perception of self-worth, which can challenge the

stability of interpersonal relationships. Adults who are higher in attachment avoidance are more

likely to rely on themselves for comfort and support rather than close others. Individuals higher

in attachment avoidance tend to be self-sufficient, are less comfortable with commitment, and

engage in distancing behavior between themselves and close others to maintain independence.

Individuals higher in attachment avoidance value autonomy and freedom and regard the

interdependence that is characteristic of close relationships as confining and emotionally

suffocating. They also have a poor perception of others, which can challenge the stability of

interpersonal relationships.

Attachment insecurity is associated with relationship initiation (the beginning or birth of

a relationship), such as number of second dates (Poulsen et al., 2013) and is a robustly associated

with relationship maintenance (the continuation of a relationship via indicators of relationship

quality such as relationship commitment and satisfaction (e.g., Mikulincer & Shaver, 2007a)).

However, no work has established whether this is true within the context of intercultural

romantic relationships.

Intercultural Romantic Relationships

In the scientific study of relationships, similarity is well established as an important

predictor of relationship outcomes. The Similarity-Attraction Hypothesis (Davis, 1981), developed

when relationship science was in its infancy, posits that partner similarity is positively related to

attraction and negatively related to relationship dissolution (see Montoya et al., 2008 for meta-

analysis). Similarity is an important factor for close relationships because it reduces the

uncertainty that individuals might otherwise experience when interacting with their partners

(Parks & Adelman, 1983). It also facilitates social interactions; for example, by allowing couples

to enjoy time together spent on shared desires and interests (Burleson et al., 1994).

Culture is an aspect of human psychology that shapes our lives and relationships, and

allows individuals to see the world from their own perspective based on the schemas and patterns

they have experienced (Campos & Kim, 2017). Due to differences in culture, opportunities for

misunderstanding and conflict may arise more frequently for intercultural couples than is true for

intracultural couples. At first glance, intercultural couples who, by definition, are navigating

differences in attitudes and customs, would seem to be at a central relationship disadvantage.

However, this disadvantage may only be true when the coping resources of an individual are

exceeded, which is a circumstance that is likely true in the context of attachment insecurity.

Attachment insecurity can be classified as an individual level vulnerability that

challenges relationship outcomes. One aspect of the Vulnerability-Stress-Adaptation Model

(Karney & Bradbury, 1995) posits that internal vulnerabilities may contribute to a reduced ability

to maintain high quality relationships or avoid dissolution. It may be that individuals higher in

attachment insecurity faced with initiating an intercultural romantic relationship, or who find

themselves trying to maintain an already established intercultural romantic relationship, are

unable to navigate the uncertainty associated with cultural differences. This uncertainty may

activate their attachment system causing them to be reactive to partners. This would directly

impact their likelihood of initiating or ability to maintain an intercultural romantic relationship

that may feel unfamiliar. In this dissertation, I will be examining both insecure domains and my

predictions for all studies are in the same direction – that attachment insecurity is a vulnerability

for intercultural relationship initiation and maintenance. Although all predictions are in the same

direction, it is important to note that both insecure domains are unique and may have different

mechanisms that drive their links to poorer relationship outcomes. For this reason, I examine

these subscales separately.

AIMS OF THE CURRENT RESEARCH

The current research aimed to answer the overarching question: is attachment insecurity

associated with intercultural romantic relationship initiation and maintenance? When met with

the opportunity to initiate or maintain an intercultural romantic relationship, uncertainty may

pose challenges for intercultural romantic relationships that is likely exacerbated for insecurely

attached individuals. Using self-report data from college students, I aimed to examine the

following: the association of attachment insecurity with likelihood of individuals initiating an

intercultural relationship (Study 1); the association of attachment insecurity with individual’s

intercultural romantic relationship quality relative to individual’s intracultural romantic

relationship quality (Study 2); and the association of both partner’s attachment insecurity with

both partner’s intercultural relationship quality and conflict management relative to both

partner’s intracultural romantic relationship quality and conflict management (Study 3).

STUDY 1

Attachment and Intercultural Romantic Relationship Initiation

Intercultural romantic relationships were only recently legalized with the U.S. Supreme

Court case of Loving v. Virginia (1967). As these relationships became more accepted,

researchers began asking individuals their reasons for initiating1 these relationships. The few

studies that have been done find that individuals give specific reasons for initiating a relationship

with someone from a different culture. The goal of the current study was to examine a factor,

specifically an individual difference factor, that may relate to one’s likelihood of initiating past

and future intercultural romantic relationships.

Previous research on intercultural romantic relationship initiation has asked participants

about their motivations and thoughts about being in a relationship with someone from a different

cultural background. Some individuals report that they were not fully satisfied with their own

culture and found aspects of their partner’s culture as desirable. For example, in Fong and

Yung’s (1995) qualitative study, Asian Americans reported wanting to marry their White partner

to obtain distance from what they perceived as the patriarchal views prevalent in their Asian

culture. Another qualitative study found that women stated that they liked that their partner

brought something “different” to their life (Khatib-Chahidi et al., 1998). Others in intercultural

romantic relationships report that they feel their difference is not as important as their similarity

(Johnson & Warren, 1994).

These reports are in line with the Self-Expansion Model (Aron & Aron, 1986), which

posits that humans use close relationships to expand their horizons and reach new goals.

Generally, embarking on relationships can be a rich opportunity for people to expand

themselves and

1In this dissertation I will use the term “initiate” to refer to the birth or beginning of a romantic relationship. This is

in keeping with how the term is used in the literature on this topic and I would like to remain consistent. At the same

time, based on the measures used in this study, the more accurate term may be “entering” into romantic relationships

lest the term “initiate” connotate active pursuit of a relationship.

relational self-expansion is a frequently studied aspect of self-expansion that individuals engage

in (Aron & Aron, 1986, 1996, 2006a). Embarking on an intercultural romantic relationship can

also uniquely allow individuals to expand their knowledge and customs of a different culture in

ways that may be regarded as an opportunity to expand the self. Recent research has begun to

examine cultural self-expansion that is specific to intercultural romantic relationships as a source

of self-expansion (West et al., 2022). Alternatively, Social Exchange Theory posits that

relationship initiation will not occur if one perceives the relationship as more costly than

rewarding (see Cropanzano & Mitchell, 2005 for a review). In the scenario outlined by Social

Exchange Theory, the motivation for the natural drive for self-expansion may be infringed upon

if individuals experience higher levels of attachment insecurity (Aron & Aron, 2006a) because

these individuals may perceive a higher level of cost than reward from these relationships.

Individuals higher in attachment anxiety develop an understanding that the way they

expand themselves, or the way they explore their world, is partially dependent on the availability

of their partner because those higher in attachment anxiety believe their secure base is uncertain,

inconsistent, or unreliable (Aron & Aron, 2006b). These individuals may desire to self-expand

initially but may have trouble continuing to grow if they perceive their partner is not available to

meet their needs. Additionally, cultural self-expansion, expanding through the exploration of

cultural difference, is related to more conflict (West et al., 2022) potentially due to the growing

pains that are associated with the beginning of an uncertain relationship. Individuals who are

higher in attachment anxiety may not feel comfortable navigating these conflicts and may prefer

to avoid a relationship that has a high likelihood of such disagreements. A study examining

openness to diversity in terms of general intergroup relations found that individuals higher in

attachment anxiety cognitively appreciated cultural difference but did not feel emotionally

comfortable with those differences (Han, 2017). These findings together suggest that individuals

higher in attachment anxiety might want to seek out intercultural romantic relationships as an

opportunity to expand but ultimately end up deciding against this decision because of the

uncertainty associated with cultural difference. Conversely, individuals higher in attachment

avoidance may develop an understanding that they cannot rely on others for self-expansion

because they learned early on in development that close others are unresponsive or dismissive of

their needs and relying on the self or on less threatening close others for self-expansion (e.g.,

friendships) may be preferrable (Aron & Aron, 1986). In the same study of openness to general

diversity, the researchers found that individuals higher in attachment avoidance reported lower

openness to cultural difference, which was explained by a lack of trust for the out-group member

(Han, 2017). Interestingly, some evidence shows that individuals who are primed with security

report more positive evaluations of outgroup members (Mikulincer & Shaver, 2001), possibly

due to reduced feelings of threat (Mikulincer & Shaver, 2007b), which again supports the

premise that attachment insecurity may be a unique threat to intercultural romantic relationship

initiation.

The Current Study

The current study was drawn from my larger online cross-sectional self-report research

study that aimed to understand romantic relationships broadly. First, I examined whether

attachment insecurity (attachment anxiety and avoidance) was associated with likelihood of

initiating an intercultural romantic relationship in the past. Second, I examined whether

attachment insecurity (attachment anxiety and avoidance) was associated with likelihood of

initiating an intercultural romantic relationship in the future. All measures were obtained with

reference to one person’s actions and perceptions within their relationship. There are no partner

nor dyadic data in this sample.

H1: I hypothesized that individuals higher in attachment anxiety would have a lower

likelihood of initiating an intercultural romantic relationship in the (a) past and in the (b) future.

H2: I hypothesized that individuals higher in attachment avoidance would have a lower

likelihood of initiating an intercultural romantic relationship in the (a) past and in the (b) future.

Method

Participants

Undergraduates (N = 588) at a large research university on the West Coast of the United

States participated in a larger study on relationships. The current study included participants who

provided data to all questions of relevant demographics, attachment styles, and intercultural

romantic relationship initiation (n = 562)2. The final sample (Mage = 21.12, SDage = 2.91)

consisted of participants who identified as being in a romantic relationship (n = 331) and

individuals who identified as not being in a romantic relationship (n = 231). Additional

demographic information is presented in Table 1.

2Participant removal was based on listwise deletion: 27 participants did not report their age, 21 participants did not

report their ethnicity, 4 participants did not report on any of the attachment items, and 1 participant did not report

their past intercultural relationship likelihood.

Table 1

Study 1 Demographic Descriptives

Relationship Single

Age M SD M SD

21.142.6821.13.21

Ethnicity

Black

2.10

2.60

Chinese 51 15.4 50 21.60

Indian 9 2.70 3 1.30

Japanese 2 0.60 2 0.90

Korean 15 4.50 19 8.20

Mexican 73 22.10 39 16.90

Middle Eastern 11 3.30 8 3.50

Multi-Ethnic 57 17.20 33 14.30

Other 4 1.20 9 3.90

Other Latino 9 2.70 5 2.20

Pacific Islander 6 1.80 2 0.90

Southeast Asian 45 13.60 30 13.00

White 42 12.70 25 10.80

Gender n % n %

Men 115 34.74 110 47.60

Non-binary 3 0.01 1 0.40

Prefer not to state 1 0.00 0 0.00

Women 212 64.04 120 51.90

Procedure

All measures and procedures were approved by the University Institutional Review

Board. Participants enrolled in the online study using the university’s social science human

subjects research pool through the SONA system and were compensated for their study

participation with extra credit that they could apply to a participating course of their choice.

Upon enrolling in the study, participants were automatically directed to an online consent form.

Inclusion criterion for the larger study was that the participant was at least 18 years of age.

Participants were directed to the online survey if they fit this criterion and consented to

participate in the study. Relationship and demographic questions were asked at the end of the

questionnaire to ensure that self-identification did not influence responses to other study

measures.

Measures

Demographics

Age. Participants were asked to answer, “how old are you?” by selecting a numeric value

from a drop-down menu that ranged from “18” to “41 and above”.

Ethnicity. Participants were asked “What is your ethnicity (indicate more than one if

applicable)?” and were given the following options: Black (1), White (2), Chinese (3), Japanese

(4), Korean (5), Southeast Asian (6), Indian (7), Pacific Islander (8), Mexican (9), Other Latino

(10), Native American (11), Middle Eastern (12), and Other (13). If they indicated more than one

ethnicity, they were coded as belonging to a Multi-Ethnic category (14). For the purposes of

analyses, participants were grouped into pan-ethnic categories that corresponded as closely as

possible to United State Census categories (i.e., Asian, Latino, White, and Other). Individuals

who identified as Chinese, Indian, Japanese, Korean, Pacific Islander, and Southeast Asian were

included in the Asian category. Individuals who identified as Mexican and Other Latino were

included in the Latino category. Individuals who identified as Middle Eastern and White were

included in the White category. Individuals who identified as Black, Multi-Ethnic, Native

American, and Other were included in the Other Ethnicity category.

Gender. Participants were asked “which gender do you identify with?” and were given

the following options: Men (1), Women (2), Non-binary or gender nonconforming (3),

Transgender (4), Other (5), and Prefer not to State (6). For the purposes of analyses, participants

were grouped into larger categories (i.e., Men, Women, and Other Gender). Individuals who

identified as non- binary or gender nonconforming, transgender, and other were included in the

Other Gender category.

Attachment Insecurity

Attachment anxiety and avoidance were assessed using the Experiences in Close

Relationships – Short Form (Wei et al., 2007) with responses that ranged from Strongly Disagree

(1) to Strongly Agree (7). Attachment anxiety was computed from six items (e.g., I need a lot of

reassurance that I am loved by my partner). Attachment avoidance was also computed from six

items (e.g., I want to get close to my partner, but I keep pulling away). One anxiety item was

reverse scored (I do not often worry about being abandoned) and three avoidance items were

reverse scored (It helps to turn to my romantic partner in times of need; I turn to my partner for

many things, including comfort and reassurance; I usually discuss my problems and concerns

with my partner). Subscale items were averaged together to create composite scores. Higher

scores indicated higher attachment anxiety (α = .78) or avoidance (α = .81), respectively. This

scale has been reliable in different cultural groups (Wei et al., 2004).

Initiation of Past or Future Intercultural Romantic Relationships

Participants were presented with the following definition of an intercultural romantic

relationship and then asked to report whether they initiated any relationships in the past that were

intercultural romantic relationships and their endorsement of initiating an intercultural romantic

relationship in the future. The past question asked, “Based on the definition provided, have you

ever been in an intercultural romantic relationship?”. The future question asked, “Based on the

definition provided, would you be in an intercultural relationship in the future?”. Both questions

were answered on a dichotomous scale, Yes (1), No (0). The definition that was provided for an

intercultural romantic relationship is as follows:

By intercultural, we mean a relationship between individuals of different

races/cultures/ethnic groups. One example would be a relationship between an individual

of Asian background and an individual of Latino background. However, an intercultural

relationship may also be between two individuals of Asian background, for example

someone who is Chinese and someone who is Korean.

Self-identification of intercultural romantic relationship status can better capture couples

who may perceive cultural differences in their relationship that fall outside of census-based

race/ethnic categories (Shenhav et al., 2017).

Data Analytic Plan

All analyses were run using SPSS version 28. First, I examined the descriptive statistics

and bivariate correlations of demographic and key study variables to better contextualize the

data. I also compared key study variables between individuals in a relationship to those not in a

relationship using an independent means t-test. This allowed me to determine whether there were

mean differences in age, attachment anxiety, and attachment avoidance between the two groups.

To answer my research questions, I used binary logistic regression, which allowed me to

use the dichotomous variable of intercultural romantic relationship initiation (yes/no) as the

outcome variable. First, I examined whether attachment insecurity was associated with whether

participants have ever been in an intercultural romantic relationship in the past. Second, I

examined whether attachment insecurity was associated with their likelihood of being in an

intercultural romantic relationship in the future. In all models, attachment insecurity was entered

as the predictor variable and intercultural romantic relationship initiation as the outcome

variable. When examining one domain of attachment insecurity as the predictor variable, the

other facet of attachment insecurity was also added to the model to account for variation in the

outcome variable that may be related to both insecure domains. For example, when examining

attachment anxiety as a predictor, attachment avoidance was included in the model to produce

the main effect that is relevant only for attachment anxiety. This same strategy was used in

analyses examining attachment avoidance as the predictor. Higher levels of attachment anxiety

or avoidance represent higher levels of conceptually unique attachment insecurity while lower

levels represent lower levels of attachment insecurity, which can be conceptualized as higher

levels of attachment security. The recommended approach to examining adult attachment is

through a dimensional model rather than a categorical model, (Fraley et al., 2015; Fraley &

Waller, 1998) and statistically adjusting for each allowed for the examination of the unique

variation of the attachment anxiety dimension and attachment avoidance dimension in the

outcome variable.

In these models, I also tested how robust attachment insecurity was as a predictor of

intercultural romantic relationship initiation via sequential testing of sets of important conceptual

variables. The base models included no additional variables except for the alternate attachment

insecurity subscale. Next, I ran the models again with the addition of conceptually relevant

demographic variables (age, ethnicity, gender) to determine if any significant models still held

after statistically adjusting for important demographic variation (set one). Gender and ethnicity

were entered into each model as dummy coded variables with Asian and Women as the reference

groups as they were the largest groups in their respective variable. Lastly, to test whether there

were differences between the groups based on participant relationship status, I added current

relationship (yes/no) as an additional variable. This allowed me to determine if any significant

models still held after statistically adjusting for variation from one’s current relationship status

(set two).

Figure 1

Study 1 Binary Logistic Regression Conceptual Models

Note. This figure shows the conceptual models for the binary logistic regressions of attachment

anxiety or avoidance with likelihood of initiating an intercultural romantic relationship in the

past or future. Panel A depicts the base model; Panel B depicts the model that includes

demographic variables; Panel C depicts the model that includes demographics and current

relationship status variables.

Table 2

Study 1 Correlation Analyses of Key Study Variables

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

Results

Bivariate correlations between key study variables are presented in Table 2.

Measures 1 2 3 4 5 6

1. Age 1 .01

0.06 -.00 .03 -.06

2. Current Relationship Status 1 -.07 -.27** .15** -.05

3. Attachment Anxiety 1 .31** 0.03 .05

4. Attachment Avoidance 1 -.09* -.05

5. Initiation of Past Intercultural Romantic Relationship 1 .40**

6. Initiation of Future Intercultural Romantic Relationship 1

Descriptive frequencies found that 58.9% of the sample reported that they had initiated an

intercultural romantic relationship in the past (yes = 331) and 41.1% of the sample reported that

they had never initiated an intercultural romantic relationship in the past (no = 231).

Descriptive frequencies also found that 81.9% of the sample reported that they would be

likely to initiate an intercultural romantic relationship the future (yes = 460) and only 18.1% of

the sample reported that they would not be likely to initiate an intercultural romantic relationship

in the future (no = 102).

A cross-tabulation (see Table 3) found that a majority of participants who reported that

they would not be in an intercultural romantic relationship in the future, also reported that they

have not been in an intercultural romantic relationship in the past (85/102). Furthermore, a

majority

of participants who reported that they would be in an intercultural romantic relationship in the

future, also reported that they have been in an intercultural romantic relationship in the past

(314/331).

Independent means t-tests revealed that individuals in a current romantic relationship

were not significantly different in age [t(560) = -0.16, ptwo-sided = .87] or levels of attachment

anxiety [t(560) = 1.66, ptwo-sided = .10] relative to individuals who were not in a current romantic

relationship.

Descriptive statistics found that, in the overall sample, the average level of attachment

anxiety was 3.89 (SD = 1.24) while the average level of attachment avoidance was 2.90 (SD =

1.15). In the sub-sample of individuals in a current romantic relationship, the average level of

attachment anxiety was 3.82 (SD = 1.34) while the average level of attachment avoidance was

2.64 (SD = 1.15). In the sub-sample of individuals not in a current relationship, the average level

of attachment anxiety was 3.99 (SD = 1.09) while the average level of attachment avoidance was

3.27 (SD = 1.03).

231

331

562

Table 3

Study 1 Cross-Tabulation of Past and Future Initiation

Future Initiation

NoYesTotal

Past InitiationNo Yes

Total

102

146

314

460

Independent means t-tests revealed that individuals in a current relationship (M = 2.64, SD

= 1.15) reported significantly lower levels of attachment avoidance [t(560) = 6.68, ptwo-sided < .001],

relative to individuals who were not in a current relationship (M = 3.27, SD = 1.03).

Hypothesis Testing

Likelihood of Initiating an Intercultural Romantic Relationship in the Past (Models 1-3)

Attachment Anxiety. Contrary to hypotheses, after adding attachment avoidance to the

model, attachment anxiety was not significantly associated with initiating an intercultural

romantic relationship in the past (OR: 1.12, p = .12; see Table 4 Model 1). This pattern was the

same after adding demographics to the model (OR: 1.13, p = .10; see Table 4 Model 2) and after

adding current romantic relationship status to the model (OR: 1.13, p = .10; see Table 4 Model

3).

Attachment Avoidance. In line with hypotheses, after adding attachment anxiety to the

model, attachment avoidance was significantly associated with a lower likelihood of initiating an

intercultural romantic relationship in the past (OR: 0.82, p = .01; see Table 4 Model 1).

Specifically, with higher attachment avoidance, there was an 18% lower likelihood of initiating

an intercultural romantic relationship in the past. This held after adding demographics to the

model (OR: 0.83, p = .02; see Table 4 Model 2) but did not hold after adding current romantic

relationship status to the model (OR: 0.89, p = .16; see Table 4 Model 3).

Table 4

Study 1 Binary Logistic Regression Analyses for Past Initiation (Models 1-3)

Model 1. bS.E. Wald pExp(b) LLCI ULCI

Constant 0.51 0.32 2.59 0.11 1.67

Anxiety 0.11 0.07 2.38 0.12 1.12 0.97 1.29

Avoidance -0.20 0.08 6.44 0.01 0.82 0.7 0.96

Model 2. bS.E. Wald pExp(b) LLCI ULCI

Constant -0.12 0.76 0.03 0.87 0.89

Anxiety 0.13 0.08 2.78 0.10 1.13 0.98 1.31

Avoidance -0.19 0.08 5.23 0.02 0.83 0.71 0.97

Age 0.01 0.03 0.18 0.67 1.01 0.95 1.08

Latino 0.06 0.23 0.07 0.79 1.06 0.68 1.65

White 0.56 0.27 4.14 0.04 1.75 1.02 2.99

Other Ethnicity 0.65 0.24 7.15 0.01 1.91 1.19 3.07

Men 0.07 0.18 0.14 0.71 1.07 0.75 1.52

Other Gender -0.15 0.95 0.02 0.88 0.87 0.13 5.58

Model 3. bS.E. Wald pExp(b) LLCI ULCI

Constant -0.65 0.78 0.70 0.40 0.52

Anxiety 0.13 0.08 2.74 0.10 1.13 0.98 1.32

Avoidance -0.12 0.08 2.00 0.16 0.89 0.75 1.05

Age 0.01 0.03 0.17 0.68 1.01 0.95 1.08

Latino 0.01 0.23 0.00 0.98 1.01 0.64 1.57

White 0.54 0.28 3.83 0.05 1.72 0.99 2.96

Other Ethnicity 0.66 0.24 7.22 0.01 1.93 1.19 3.11

Men 0.13 0.18 0.52 0.47 1.14 0.80 1.63

Other Gender -0.27 0.95 0.08 0.78 0.76 0.12 4.90

Current Relationship Status 0.58 0.19 9.83 0.00 1.79 1.24 2.58

Likelihood of Initiating an Intercultural Romantic Relationship in the Future (Models 4-6)

Attachment Anxiety. Contrary to hypotheses, attachment anxiety was not significantly

associated with initiating an intercultural relationship in the future (OR: 1.17, p = .10; see Table

5 Model 4). This held after adding demographics to the model (OR: 1.19, p = .08; see Table 5

Model 5) and after adding current romantic relationship status to the model (OR: 1.19, p = .08;

see Table 5 Model 6).

Attachment Avoidance. Contrary to hypotheses, attachment avoidance was not

associated with initiating an intercultural relationship in the future (OR: 0.85, p = .12; see Table

5 Model 4). This held after adding demographics to the model (OR: 0.87, p = .17; see Table 5

Model 5) and after adding current romantic relationship status to the model (OR: 0.83, p = .08;

see Table 5 Model 6).

Table 5

Study 1 Binary Logistic Regression Analyses for Future Initiation (Models 4-6)

Note. The “.” depicts a lack of variation to compute the

effect.

Model 4 bS.E. Wald pExp(b) LLCI ULCI

Constant 1.37 0.40 11.71 <.001 3.94

Anxiety 0.16 0.09 2.78 0.10 1.17 0.97 1.41

Avoidance -0.16 0.10 2.48 0.12 0.85 0.70 1.04

Model 5 bS.E. Wald pExp(b) LLCI ULCI

Constant 2.01 0.87 5.40 0.02 7.45

Anxiety 0.17 0.10 3.09 0.08 1.19 0.98 1.44

Avoidance -0.14 0.10 1.91 0.17 0.87 0.71 1.06

Age -0.05 0.04 2.34 0.13 0.95 0.88 1.02

Latino 0.37 0.28 1.71 0.19 1.45 0.83 2.52

White 0.46 0.34 1.81 0.18 1.58 0.81 3.09

Other Ethnicity 1.18 0.37 10.3 0.001 3.27 1.59 6.74

Men 0.14 0.23 0.37 0.54 1.15 0.73 1.81

Other Gender . . . . . . .

Model 6 bS.E. Wald pExp(b) LLCI ULCI

Constant 2.38 0.90 6.93 0.01 10.78

Anxiety 0.17 0.10 3.14 0.08 1.19 0.98 1.44

Avoidance -0.18 0.11 2.98 0.08 0.83 0.68 1.03

Age -0.06 0.04 2.41 0.12 0.95 0.88 1.02

Latino 0.41 0.29 2.04 0.15 1.50 0.86 2.62

White 0.48 0.34 1.93 0.17 1.61 0.82 3.15

Other Ethnicity 1.20 0.37 10.42 0.01 3.31 1.60 6.83

Men 0.11 0.23 0.23 0.63 1.12 0.71 1.76

Other Gender . . . . . . .

Current Relationship Status -0.37 0.24 2.41 0.12 0.69 0.43 1.10

Discussion

The current study is the first to examine the role of attachment insecurity as it relates to

the initiation of intercultural romantic relationships relative to intracultural romantic

relationships. Two noteworthy patterns emerged. First, attachment anxiety was not associated

with intercultural romantic relationship initiation, but attachment avoidance was associated with

past intercultural romantic relationship initiation. Second, overall patterns differed in their

association with past and future

intercultural

romantic

relationship

initiation.

Altogether,

these

results

suggest

that

attachment insecurity may be an important individual difference that can inform understanding

of likelihood of initiating intercultural romantic relationships.

There was no significant effect of attachment anxiety with initiation of an intercultural

romantic relationship in the past. It could be that there is wide variation in attachment anxiety

that the association with intercultural romantic relationship initiation appears insignificant. For

example, it may be that a third variable is influencing the detection of the main effect (e.g.,

personality factors). On the contrary, as predicted, attachment avoidance was significantly

associated with a lower likelihood of initiating an intercultural romantic relationship in the past.

These individuals feel more comfortable with independence and less comfortable with intimacy

and so initiating a relationship of any kind may not be desired by our sample. This could be why

the effect was no longer significant when current relationship status was added to the model.

Initiation of any relationship for those higher in attachment avoidance may be key in this effect.

These findings together suggest that attachment avoidance may be the more problematic

attachment insecurity domain for relationship initiation and provides evidence for why it is

important to examine these domains separately.

There was no significant association of attachment anxiety or attachment avoidance with

likelihood of initiating an intercultural relationship in the future. However, my examination of

the base rates of individuals who reported not being likely to initiate an intercultural romantic

relationship in the future revealed that the numbers were very low. There may not be enough

variability in the sample to determine whether individual differences were associated with a

likelihood of initiating an intercultural romantic relationship in the future. More importantly, it is

possible that individuals self-report high likelihood of initiating intercultural romantic

relationships in the future even if their past behavior does not show a history of doing so. In this

scenario, (1) individuals may not have had the opportunity to initiate an intercultural romantic

relationship, (2) individuals may only be open to certain intercultural romantic relationships, (3)

individuals may be engaging in social desirability when self-reporting given the diverse

university population from which this sample was collected, or (4) responses may reflect

individuals’ inaccuracies in affective forecasting.

The current sample of participants may have not had the opportunity to initiate an

intercultural romantic relationship in the past, but they feel they would be likely to do so in the

future. The university where this sample was collected is highly diverse, but it is possible that

some students segregate themselves within the student body. Previous research finds that self-

segregation is common (even in ethnically diverse and desegregated schools) and is a continued

topic of discussion regarding promotion of positive intergroup relations (Conger, 2005; Thijs &

Verkuyten, 2014). Therefore, although the possible dating pool is diverse, the perception of

opportunity to date interculturally may be lower than reflected in population rates. Furthermore,

the mean age of participants was 21-years. As of 2011, the average age of marriage in the U.S.

was 26.5 for women and 28.7 for men (Pew Research, 2011). The participants in this sample

may have a number of years of dating ahead of them and their responses may reflect an openness

to the possibility of an intercultural relationship in the future even though they have not had a

chance to initiate such relationship in the past. On the contrary, some participants may have

regarded themselves as having no potential dating years ahead of them if they are currently in a

serious dating relationship, a domestic partnership, or married.

Another factor that may be relevant for understanding initiation of intercultural romantic

relationships is that participants may only be open to certain intercultural romantic relationships.

It could be that individuals would be likely to date individuals from some out-groups but not

others.

This could be relevant for religious background. Participants might be likely to date someone

from a different ethnic group but not another religious domination. Individuals may have not met

the right individuals in the past but are open to specific intercultural dating in the future.

Participants may be engaging in socially desirable ways when self-reporting their

likelihood to date interculturally. Given the diverse university population from which this sample

was collected, it is likely more accepting to be open to diversity in this population. Reporting a

complete unlikelihood of dating someone from a different cultural background could be

perceived as stemming from prejudice or racism. The individuals in this sample may have been

reporting, consciously or sub-consciously, a higher likelihood of initiating an intercultural

romantic relationship in the future than they truthfully feel. In future research on intercultural

relationship initiation, it is important to examine participants through other measures such as

behavioral intergroup interactions or experimentally presenting potential dating options and

determining if participants desire in-group or out-group members differently. This is another

reason why the reports of past behavior may be a more accurate metric for relationship initiation

rather than projected future likelihood.

Lastly, participants may be reporting inaccurately due to affective forecasting. Research

on affective forecasting has found that individuals do a poor job of predicting how they will feel

in the future as measured by comparing one’s prediction compared to how they feel when the

event occurs (Kushlev & Dunn, 2012). It could be that people are optimistic that they would be

likely to initiate an intercultural relationship but when the time comes, they are less likely. Work

also shows that individuals are relatively accurate in affective forecasting the valence of how

they will feel (e.g., negative, or positive) but less accurate at forecasting exactly how they will

feel (e.g., the extent to which they feel negatively; (Mathieu & Gosling, 2012)). Individuals

may have been

answering the question at face value, that they would be likely to initiate an intercultural

romantic relationship, but how likely they would be is still unknown.

Strengths & Limitations

To my knowledge, this is the first study to quantitatively examine intercultural romantic

relationship initiation. This study consisted of an ethnically diverse sample of individuals

currently in and not currently in a romantic relationship, which allowed me to determine if

initiation of any romantic relationship was a factor that might contribute to these findings. Lastly,

this study utilized a step-down process to isolate variation in the outcome variable and allowed

me to test whether attachment anxiety and attachment avoidance were associated with

intercultural romantic relationship initiation over and above age, ethnicity, gender, and current

relationship status. Limitations of this study include the following: (1) with the current questions,

I was unable to determine whether individuals who are not currently in a romantic relationship

are unlikely to initiate any type of romantic relationship or if they merely are currently single;

(2), the assessment of initiating an intercultural relationship was a single dichotomous item,

which reduced my ability to determine the degree of likelihood that someone would initiate an

intercultural romantic relationship; (3) I did not examine the role of satisfaction with past or

current intercultural romantic relationship experiences and how this may be related to previous

and future initiation of an intercultural romantic relationship; and lastly (4) this sample was of

college students in a diverse liberal community, which limits generalizability of these findings.

Future Directions

To address the current study limitations, I offer suggestions for future directions. It is

important for future studies to parse out whether individuals are likely (or unlikely) to initiate a

relationship at all or if their unlikelihood is directly related to the initiation of a romantic

relationship with someone from a different cultural background. In this study, this was indirectly

addressed by including individuals who were in a current relationship and those who were not in

a current relationship, but future studies should address this more directly. A qualitative

interview approach could be useful as well; this methodology could draw out details on the

considerations that shape whether an intercultural romantic relationship would be initiated or not.

Furthermore, qualitative data could elucidate why individuals have not initiated an intercultural

relationship in the past but might report a likelihood to do so in the future or vice versa. It may

be that past negative experiences reduce one’s future likelihood of engaging with particular

outgroup members. Although this has been done with a few studies cited in this dissertation, the

association of these reports with attachment insecurity is still unknown. Future studies should

also examine how different types of cultural variation are related to romantic relationship

initiation. It could be that some forms of cultural difference, such as different religions, are

perceived as less surmountable than other cultural differences. As such, understanding whether

the degree of difference matters for intercultural romantic relationship initiation is important.

One way to examine this degree of likelihood in future studies would be to assess participant

likelihood via a 0-100 scale, which would increase the variability in participant responses. Future

studies should also consider whether individuals have experience in intercultural romantic

relationships as this may color their view of specific out-group members. Lastly, it would be

interesting to collect data from samples that differ demographically such as their geographic

location, their age, their developmental stage, and their generation to determine if contextual

factors impact participant likelihood of initiating an intercultural romantic relationship.

Conclusion

This research lays the groundwork for future studies that can inform understanding of

what allows intercultural romantic relationship initiation with implications for future intercultural

romantic relationship maintenance. Factors that impact likelihood of being in an intercultural

romantic relationship may also apply to cohesion during intergroup interactions, which is a

significant implication of this study. The current study findings suggest that attachment

insecurity may be an important factor in understanding intercultural romantic relationship

initiation, yet much is still unknown regarding these relationships. Study 2 builds on Study 1 by

extending the role of attachment insecurity on relationship maintenance in intercultural romantic

relationships that are already established.

STUDY 2

Attachment and Intercultural Romantic Relationship Quality

Relationship commitment and satisfaction are two important aspects of relationship

quality. Relationship commitment is one’s dedication and investment to the longevity of their

relationship and their perception of it enduring long-term (Rusbult, 1980). Relationship

satisfaction is a global evaluation that is established when an individual feels content that their

needs are being met by their romantic partner (Fincham et al., 2018). Attachment insecurity is

robustly associated with relationship commitment and satisfaction (Mikulincer & Shaver,

2007a), but this association is unknown within the context of intercultural romantic relationships.

The goal of the current study was to examine whether attachment insecurity was associated with

relationship quality and whether this association varied between intercultural and intracultural

romantic relationships.

Studies of intercultural3 romantic relationship quality to date have yielded mixed patterns

of results. Some find that romantic relationship quality among intercultural couples is lower than

intracultural couples (e.g., Hohmann-Marriott & Amato, 2008). Two studies have found

romantic relationship satisfaction was similar when comparing interethnic to intraethnic couples

(Negy & Snyder, 2000; Shibazaki & Brennan, 1998). Contrary to expectations, at least one study

found that romantic relationship satisfaction was higher for interracial couples relative to

intraracial couples (Troy et al., 2006). This latter pattern has also been replicated in my work on

individuals in intercultural romantic relationships who reported higher average levels of romantic

relationship commitment and satisfaction relative to individuals in intracultural romantic

relationships (Froidevaux & Campos, Under Review). Interestingly, a recent meta-analysis

(Brooks, 2022)

3 For this section I will refer to intercultural couples the way the cited publication does (e.g., interethnic: partners

from different ethnic backgrounds).

argued that intercultural couples are not at heightened risk merely by being in an intercultural

relationship but that there may be other factors at play contributing to the mixed literature such

as geographic location where data is collected. The current study will examine whether an

important individual difference, attachment insecurity, may also contribute to the mixed findings.

In general, attachment insecurity is negatively associated with relationship quality. Those

higher in attachment anxiety tend to feel ambivalent about their relationship, which hinders their

ability to experience high relationship commitment (Etcheverry et al., 2013) and relates to lower

relationship satisfaction (e.g., Butzer & Campbell, 2008; Tucker & Anders, 1999). Those higher

in attachment avoidance are more comfortable with their independence and thus do not desire to

establish interdependence with a close other (Mikulincer & Shaver, 2007a). Individuals who

experience higher levels of attachment anxiety are more comfortable with more self-disclosure

(Hudson & Fraley, 2017), which can sometimes be perceived as too much too soon for their

partners. In contrast, those who experience higher levels of attachment avoidance feel

uncomfortable with self-disclosure (Pistole, 1993). Both self-disclosure patterns impact intimacy

and relationship quality.

The negative association of attachment insecurity and relationship quality may be

stronger for intercultural romantic relationships relative to intracultural romantic relationships.

Productively navigating differences in cultural background is likely more difficult for individuals

in intercultural romantic relationships who are also experiencing attachment insecurity. This is

because uncertainty and a fear of instability is activating for the attachment system, especially for

those who are higher in attachment insecurity.

The Current Study

The current study was drawn from my same larger online cross-sectional self-report

research study as reported in Study 1. First, I examined whether attachment insecurity

(attachment anxiety and avoidance) was associated with romantic relationship commitment and

satisfaction and whether this varied as a function of intercultural romantic relationship status. All

measures were obtained with reference to one person’s actions and perceptions within their

relationship. There are no partner nor dyadic data in this sample.

H1a-H1b: I hypothesized that attachment anxiety would be negatively associated with

relationship commitment (1a) and relationship satisfaction (1b) and that these associations would

be stronger for individuals in an intercultural relationship relative to those in an intracultural

relationship.

H2a-H2b: I hypothesized that attachment avoidance would be negatively associated with

relationship commitment (2a) and relationship satisfaction (2b) and that these associations would

be stronger for individuals in an intercultural relationship relative to those in an intracultural

relationship.

Method

Participants

Undergraduates (N = 588) at a large research university on the West Coast of the United

States participated in a larger study on relationships. The current study included participants who

provided data to all questions of relevant demographics, attachment styles, and relationship

quality (n = 563)4. The final sample (Mage = 21.14, SDage = 2.68) consisted of participants who

identified

4Participant removal was based on listwise deletion: 27 participants did not report their age, 21 participants did not

report their ethnicity, and 4 participants did not report on any of the attachment items.

Table 6

Study 2 Demographic Descriptives

InterculturalIntracultural

as being in a romantic relationship that began at least 3-months prior to their participation (n =

332). Additional demographic information is presented in Table 6.

Age M SD M SD

20.87 2.27 21.43 3.04

Ethnicity n % n %

Black 3 1.70 4 2.50

Chinese 20 11.60 31 19.40

Indian 5 2.90 4 2.50

Japanese 1 0.60 1 0.60

Korean 7 4.10 8 5.00

Mexican 28 16.30 45 28.10

Middle Eastern 8 4.70 3 1.90

Multi-Ethnic 39 22.70 18 11.30

Other 2 1.20 2 1.30

Other Latino 6 3.50 3 1.90

Pacific Islander 4 2.30 2 1.30

Southeast Asian 31 18.00 15 9.40

White 18 10.50 24 15.00

Gender n % n %

Men 58 33.70 57 35.60

Non-binary 2 1.20 1 0.60

Prefer not to state 0 0.00 1 0.60

Women 112 65.10 101 63.10

Relationship Length n % n %

3 months 27 15.70 16 10.00

6 months 24 14.00 25 15.60

Less than 1 year 37 21.50 25 15.60

1-2 years 30 17.40 37 23.10

2-5 years 42 24.40 40 25.00

over 5 years 12 7.00 17 10.60