Relational mobility is a socio-ecological construct quantifying how much freedom and opportunity a society affords individuals to choose and dispose of interpersonal relationships. Past research has confirmed that relational mobility varies across nations, but no large-scale study has examined the degree to which relational mobility may vary within a single nation. We report two studies (Study 1, N = 647; Study 2, N = 7343) exploring within-country similarity or variability in relational mobility across all 27 states and five geo-socio-political regions in the continent-size country of Brazil. Results confirmed the measurement equivalence of the Relational Mobility Scale across respondents from all Brazilian states. Notably, relational mobility scores were uniform across Brazilian regions and states, indicating a common national culture regarding the amount of opportunities Brazilians have in selecting new relationship partners within their social context. Replicating existing findings, relational mobility was positively associated with pro-active tendencies that help people retain relationships—levels of intimacy and self-disclosure toward a close friend—indicating that friends tend to feel closer intimacy to their close friends, and reveal serious personal information to a larger degree in social contexts where opportunities to find and retain relationships with like-minded others are greater.
Citation: Milfont TL, Thomson R, Yuki M (2020) Does relational mobility vary across national regions? A within-country examination. PLoS ONE 15(7): e0235172. https://doi.org/10.1371/journal.pone.0235172
Editor: Sergi Lozano, Universitat de Barcelona, SPAIN
Received: August 28, 2019; Accepted: June 9, 2020; Published: July 2, 2020
Copyright: © 2020 Milfont et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This research was supported by a Foreign Researcher Invitation Grant (外国人研究者等招致助成) from the Heiwa Nakajima Foundation (平和中島財団) and by a sabbatical leave from Victoria University of Wellington awarded to the first author.
Competing interests: The authors have declared that no competing interests exist.
Socio-ecological contexts vary in the degree they afford mobility with regards to interpersonal relationships and group memberships. While high-mobility contexts give individuals the opportunity and freedom to choose new and discard old relationships, individuals in low-mobility contexts have fewer opportunities and less freedom to select interaction partners based on personal preferences. Drawing on this observation, scholars have conceptualised relational mobility as a socio-ecological construct reflecting how much freedom in interpersonal or intergroup choice a given social environment affords individuals [1, 2]. This conceptualization follows a socio-ecological approach to behaviour and psychological tendencies, where those tendencies are viewed as adaptive strategies individuals follow (either consciously or not) in order to adapt to the level of relational mobility in their society.
This conceptualisation warrants objective measurement of relational mobility with archival material or observations, but such measurements have not been attempted to date. Relational mobility has been typically measured with self-report surveys accessing individuals’ perceptions of how much freedom and opportunity their social milieu affords individuals to choose or dispose of interpersonal relationships. Hence, self-report measures of relational mobility assess individuals’ perceptions of the “the amount of opportunities people have to select new relationship partners in a given society or social context” (1 p. 3). Yuki and colleagues  developed the Relational Mobility Scale (RMobS) to access individuals’ perceptions regarding how easy or difficult it is for people in their social and ecological setting to enter into, move out of, and form new relationships and group memberships. Notably, the focus is not on measuring how mobile individuals think they are, personally, in choosing relational partners but rather on measuring individuals’ perceptions of how easy it is in general for people in their society to make and change interpersonal relationships.
Using the RMobS, and other proxies for relational mobility, small-scale studies have examined relational mobility both within  and across nations (see 2, for a review). The first large-scale study of relational mobility examined variation in relational mobility across 39 societies . Using the RMobS, this study found relational mobility to be higher in North American, Western European and Latin American countries, and lower in East Asian, South Asian, North African, and Middle Eastern countries. Thomson and colleagues  also examined predictors and outcomes of relational mobility. They found that relational mobility was lower in societies that both practice settled, interdependent subsistence styles and have stronger ecological and historical threats, and that individuals in societies with higher relational mobility report more pro-active interpersonal behaviours and psychological tendencies that help them acquire and retain relationships.
Although past research has focused mainly on variability in relational mobility across nations, Schug, Yuki and Maddux  argued that within-culture variability in relational mobility is likely and consequential, and initial evidence from Yuki et al. (3, Study 2) supports this argument. These authors used regional-average service years (i.e., job tenure, or the average number of years respondents in 11 regions within Japan had worked for their current employer) as an objective indicator of the stability of relationships in the area, and a proxy indicator of relational mobility. They observed variability in job tenure between the regions, and that the positive association between individuals’ self-esteem and happiness was stronger in more relationally mobile regions where people more often move to new workplaces and change their social networks.
Thomson et al.’s  large-scale examination of relational mobility corroborate Schug, Yuki and Maddux  and Yuki et al.’s (3, Study 2) suggestion that relational mobility could vary within a country. Across the 39 countries and 16,939 people included in their study, only 9% (ICC = 0.09) of variance in relational mobility was accounted for by country of residence. This leaves 91% of variance unaccounted for. However, the degree to which relational mobility varies within a nation has not yet been investigated systematically.
Why are within-country examinations of psychological phenomenon important? Research examining cultural similarities and differences in psychological phenomena has flourished in the last decades with large cross-cultural projects that allow the characterisation of nations based on dimensions of cultural variability. These dimensions include values [6, 7], beliefs , and social norms , to name but a few. Although these dimensions of cultural variability are useful in describing and contrasting the psychological make-up and behavioural tendencies of individuals from distinct societies, it is possible that national cultures are not uniform within their boundaries. Indeed, an increasing number of studies have used similar large-scale approaches to examine the cultural variability that might exist within single nations. To illustrate, variability in the individualism–collectivism dimension has been observed within the United States [10, 11], and individualism–collectivism variability within China has been liked to specific agricultural practices . When considered as a whole, the United States and China are characterized as individualistic and collectivistic countries, respectively, but these findings provide evidence of within-country variability. Extending relational mobility research, we provide a large-scale examination of relational mobility within Brazil.
Brazil cultural makeup
Drawing from this growing line of research examining the extent to which dimensions of cultural variability differ within a single national culture [10–12] as well as preliminary evidence that relational mobility may vary within a country [3, 5], we provide the first systematic examination of relational mobility within a single country. We examine the within-country similarity/variability in relational mobility across all federative units (hereafter referred to as states) as well as the five geo-socio-political regions in Brazil, which is a continent-size country both in terms of territory [official territorial area of 8,515,759 km2; 13] and population [207,660,929 inhabitants; 14]. Brazil was colonized by the Portuguese in the 16th century, and six major groups make up the Brazilian population: native peoples (mainly from Tupi and Guarani language groups), Portuguese, Africans, other Europeans, Middle Eastern, and Asian immigrant groups. Brazil is the only Portuguese-speaking nation in the Americas, and its primary religion is Roman Catholic.
Anthropological and psychological accounts indicate great variability of psychological tendencies across states and geo-socio-political areas within Brazil [see, e.g., 15–18]. To illustrate, Torres and colleagues  observed that individuals within Brazilian regions with earlier histories of immigration and settlement more strongly endorse values of conformity, tradition and security than individuals in newer settlements in the Center-West region of the country. Hofstede et al.  also noted that Masculinity was greater in the North than in the Northeast geo-socio-political regions of the country, which they attributed to the Indigenous and Afro-Brazilian roots of these regions, respectively.
However, there is also evidence indicating homogeneity of Brazilian culture. In contrast to other national cultures, Brazil is often characterized as a collectivistic culture and recent research indicates Brazilian participants display more holistic thinking (i.e., greater attention to context more common in collectivistic societies) than both Chinese and US participants ; but Brazil is better described as above average on both Power Distance and Uncertainty Avoidance . Notably, although noting small regional distinctions, Hofstede et al.  concluded that their findings provided evidence of a common Brazilian national culture.
Should relational mobility vary across Brazilian regions? On one hand, it is possible that relational mobility is not uniform across the country, considering the large variability in geo-physical and historical contexts between regions within Brazil. In fact, it would be unsurprising to observe variability in psychologies across this diverse and large country in similar ways as has been observed in other large countries such as China and the US [10–12]. Recent studies exploring variability of psychologies within Brazil have documented some distinctions between regions with earlier (versus newer) histories of immigration and settlement [17, 18]. It is thus possible that relational mobility would vary as a function of settlement history and ethnic make-up of the Brazilian geo-socio-political regions. On the other hand, Brazil scored highly on relational mobility as compared with other nations , and this high mobility may be reflected in anthropological accounts of Brazilian sociability and hospitality . Existing research also suggests a common Brazilian national culture [6, 15], indicating that Brazilians might perceive the same level of relational mobility in their social context across all regions of the country. We conducted two exploratory studies to examine these possibilities.
The present study
We report two online studies conducted in Brazil. The research was approved by the Ethics Committee of the Center for Experimental Research in Social Sciences at Hokkaido University. The data were analyzed anonymously. In Study 1 (N = 647) we perform secondary analysis of survey data from Brazil previously collected as part of the aforementioned project examining relational mobility in 39 societies . In this first study, we examine the measurement model of the RMobS as well as the variability of relational mobility across the Brazilian geo-socio-political regions.
Study 2 (N = 7343) extends the first study in three significant ways. We first examine the measurement invariance of the RMobS across all 27 states in Brazil. Once we established whether participants from all states respond to the scale in statistically comparable ways, we then examine whether scores of relational mobility vary across the Brazilian states and regions, or whether perceptions of relational mobility are the same across this vast country, replicating or not Study 1 findings. Finally, we use multilevel modelling to examine the associations between relational mobility and outcome variables linked to relational mobility. As Thomson et al.  reported, individuals within socio-ecological contexts high in relational mobility are more likely to report pro-active interpersonal behaviours as well as psychological tendencies that help them build and retain relationships. They also found that perceived similarity between one’s best friend and oneself is higher in societies that are higher than lower in relational mobility, which can be a relational consequence of high freedom in relational choice. In line of these reasonings, Thomson and colleagues observed a positive association between relational mobility and intimacy, self-disclosure and interpersonal similarity, indicating that relationally mobile societies tend to share personal information more quickly (e.g., self-disclosure) and report higher intimacy and feelings of interpersonal similarity. The final goal of Study 2 is to test whether these cross-national results replicate within a nation-state.
Participants and procedure.
The World Relationships Study  recruited participants using Facebook advertisements in 39 societies (see more information in the S1 File). Societies were selected in order to maximize variation in geography and cultural blocks, as well as Facebook penetration rate to maximize sample diversity within each country. Responses from 499 participants from Brazil were analysed in the Thomson et al.’s  study, from a total of 738 participants who clicked past the initial survey landing page. While Thomson et al.’s  goal during data cleaning was to minimize inclusion of participants who had dropped out early in the survey, in the current study we were concerned primarily with maximizing the number of participants who had, at a minimum, responded to the RMobS. In the survey flow, this scale was presented to participants on the first page of the survey after the informed consent landing page, while demographic questions, including self-reported state of residence in Brazil, were on the last page of the 6-page survey.
Thomson et al.  removed responses with missing information regarding country of residence. Beyond country of residence, our focus is on the state of residence in Brazil, but the state of residence question appeared in the last survey page and this information was missing for 214 participants (out of the 738 participants who clicked past the initial survey landing page). We increased sample size across the Brazilian regions by using an online tool to identify the participants’ missing states from their IP addresses (http://www.bulkseotools.com/bulk-ip-to-location.php).
We removed cases with duplicate IP addresses (retaining the first response on file), those from outside Brazil (as identified by the IP address), cases with more than 33% of data missing for the RMobS, and cases that completed the full survey under three minutes. Because the RMobS is balanced with respect to positively and negatively keyed items, we also removed cases where 80% of responses included only the extreme response values (1, 2, 5 or 6). After these procedures, we had relational mobility data from 647 participants (92% female; Mage = 22.31, SDage = 9.57).
Table 1 presents the distribution of participants for each Brazilian state, ranging from two participants in Roraima (and none in Amapá) to 101 participants in both São Paulo and Minas Gerais. Considering the low sample in most states, we grouped participants to the five geo-socio-political regions of the country: Centre-West (n = 61), North (n = 36), Northeast (n = 160), South (n = 131), and Southeast (n = 259). We also considered the regrouping of the Federal District within the Southeast region instead of the Centre-West region as suggested by Hofstede et al. , but this regrouping had no effect on the results reported in both studies so we report results only when groupings were according to the traditional allocation of the Federal District within the Centre-West region.
Yuki and colleagues  developed the 12-item RMobS to access individuals’ perceptions of the degree to which people in a society or social context have the freedom and opportunity to choose and dispose of relationships based on personal preference. The scale includes five items for a “meeting” factor capturing the degree to which a society or social context affords opportunities for individuals to meet new people and forge new relationships (e.g., “They (the people around you) have many chances to get to know other people”), and seven items for a “choosing” factor capturing the degree to which people have the freedom to choose and leave relationships based on personal preference (e.g., “They (the people around you) are able to choose, according to their own preferences, the people whom they interact with in their daily life”). Items are rated on a 6-point agreement scale anchored by 1 (strongly disagree) and 6 (strongly agree).
Thomson et al.  used a translation/back-translation procedure to translate the English version of the survey into each of the other 19 languages. Separate professional translators first translated and back-translated surveys, and then English-bilingual collaborators for whom the target language was their native tongue double checked both translations and approved of the final version. The Brazilian Portuguese version of the RMobS is available here: https://osf.io/e5hm9/.
Results and discussion
Fig 1 depicts the measurement model of the RMobS that has been tested and validated cross-nationally by Thomson et al. . The model includes four latent factors: two first-order content factors, one second-order content factor, and one common-method bias factor modeling acquiescent response style. Instead of focusing on the sub-dimensions of relational mobility of “meeting” and “choosing”, this model focuses on an overall latent construct.
We were unable to run a multi-group confirmatory factor analysis to test the measurement invariance of this model across participants clustered in states or regions given the complexity of the model and small sample sizes or regions. We provide a thorough examination of this issue in Study 2. Here we only examined the fit of the model across all participants with robust maximum likelihood estimation in Mplus (version 7.4). The model had good fit to the data: S-B χ2 (50) = 93.845, p < .001, *CFI = .951, *SRMR = .040, *RMSEA = .037 (90% CI = .025, .048). The average standardized loading was .478 and the lowest was for Item 9 in the “choosing” factor, which was still statistically significant (b = .276, p < .001).
The factor scores generated from this model and the raw scores (averaging over the items after reversing coding the relevant items) were highly correlated (r = .980, p < .001; N = 646). We used the factor scores to examine whether relational mobility varied across the five geo-socio-political regions. A one-way ANOVA showed that the factor scores did not vary statistically across the regions (F = .139, p = .968). This lack of variability is evidenced by the respective raw and factor scores aggregated for each region: North (4.340, .00036), Northeast (4.419, .02019), Centre-West (4.353, -.01867), South (4.355, -.00690), and Southeast (4.365, -.00457).
The results from Study 1 suggest a common social reality within Brazilian regions regarding relational mobility. The small sample sizes across the geo-socio-political regions reduced our confidence in the findings, however. To overcome this shortcoming, we examined whether the observed uniformity in relational mobility replicates in a more systematic investigation by examining possible variations not only in the five broader Brazilian regions but also across all 27 states. Thus, Study 2 examines the measurement invariance of the RMobS across all 27 states in Brazil, whether scores of relational mobility vary or not across the Brazilian states and regions, and associations between relational mobility and outcome variables.
Participants and procedure.
We used the same Facebook advertisement strategy to recruit participants as in Thomson et al. , but this time the explicit focus was on recruiting participants from all states in Brazil. Since Internet penetration (and therefore Facebook usage) varies across states, we used three survey waves to maximize sample sizes in all geographic locations. Wave 1 was collected in Oct/Nov 2015 (N = 3688; 83.5% female; Mage = 20.94, SDage = 6.82), Wave 2 in Jan/Feb 2016 (N = 2522; 90.5% female; Mage = 31.23, SDage = 16.47), and Wave 3 in Sep/Oct 2017 (N = 1133; 82.9% female; Mage = 37.48, SDage = 15.88). These numbers represent the final sample for each wave after exclusions. Similar to the procedures used in Study 1, we removed cases with duplicate IP addresses or those from outside Brazil, cases with more than 33% of data missing for the RMobS, cases where 80% of responses included only one of the extreme response values (1, 2, 5 or 6), and participants who completed the survey under 3 minutes.
The final sample comprised 7343 participants (85.6% female; Mage = 26.30, SDage = 13.53) from all Brazilian states. We used the same online tool as in Study 1 to increase sample size across the Brazilian for participants with missing information on their state of residence (n = 2051). Sample sizes per state varied from 84 (Acre) to 560 (São Paulo; see Table 1), and comprised 1000+ participants in each region: Centre-West (n = 1060), North (n = 1209), Northeast (n = 2453), South (n = 1066), and Southeast (n = 1555).
Study 2 examines relational mobility plus the three measures detailed below (see S1 File for additional measures not examined here). Items of all four measures targeted either a participant’s “closest friend” or “romantic partner”. Advertisements focusing on either target were randomly presented to Facebook users targeted by Brazilian state. A roughly equal number of participants was obtained for each version of the survey.
Relational mobility. The same 12-item RMobS described in Study 1 was used in this study. Items are rated on a 6-point agreement scale anchored by 1 (strongly disagree) and 6 (strongly agree). Relational mobility scores were statistically higher for participants who responded to the friendship version of the survey (N = 4,027, M = 4.47, SD = .65) than for those who responded to the romance version (N = 3,316, M = 4.40, SD = .65), t(7341) = 4.81, p < .001, d = .11. Considering that this difference was small, that Thomson et al.  did not observe such a difference in their 39-nation study, and that exploration of this difference is outside the scope of our study, we did not investigate this further. We thus collapsed the responses to the two versions of the survey to examine how general perceptions of relational mobility impacts intimacy, self-disclosure and interpersonal similarity for both a closest friend and romantic partner.
Intimacy. We used a 10-item version of Sternberg’s triangular love scale to assess intimacy . Some example items in the scale are “I receive considerable emotional support from [friend/romantic partner]”, and “I feel that [friend/romantic partner] really understands me”. Items were rated on a 5-point agreement scale anchored by 1 (strongly disagree) and 5 (strongly agree), and Cronbach’s alphas were .92 and .91 for the friendship and romance versions, respectively.
Self-disclosure. We used five items to measure self-disclosure . Participants were asked to indicate the degree to which they have revealed themselves to their friend/romantic partner in term of secrets, embarrassing experiences, failures, worries, and bad things that had happened to them. Items were answered on a 5-point scale anchored by 1 (I have not revealed any information at all) to 5 (I have revealed even the most serious information), and Cronbach’s alphas were .88 for both survey versions.
Interpersonal similarity. Participants were asked to rate how similar they are with their friend/romantic partner in five aspects: personality, hobbies, values, behaviours, and lifestyles . Items were rated on a 7-point scale anchored by 1 (not similar at all) and 7 (very similar), and Cronbach’s alphas were .83 and .81 for the friendship and romance versions, respectively.
Results and discussion
Measurement invariance of measures.
We started by examining whether participants from all Brazilian states used the four measures in the same way. We compared the fit of the model in Fig 1 against alternative models, and then ran multi-group analysis to confirm configural, metric, and scalar invariance of the RMobS. Table 2 presents the results of all nine models tested. In brief, the fit statistics support the measurement model depicted in Fig 1 across all participants and confirms the measurement equivalence of the RMobS across participants from all Brazilian states. The model also had good fit across the Brazilian regions: S-B χ2 (333) = 851.52, p < .001, *CFI = .947, *SRMR = .036, *RMSEA = .033 (90% CI = .030, .035).
Examination of the psychometric properties of the other three scales confirmed the one-factor structure and good loadings of the scale items (see Table 3). The results also confirmed the configural, metric, and scalar invariance of the measures across participants from all 27 Brazilian states irrespective of social target (see Table 4).
Variability of relational mobility across Brazilian states and regions.
We used the factor scores and latent means (from Model 6 in Table 5) to extend the results of Study 1 and examine variability of relational mobile across the Brazilian states and the five geo-socio-political regions. As in Study 1, the factor scores and the raw scores were highly correlated (r = .977, p < .001, N = 7343).
Starting with the state comparisons, the results indicated that less than 1% of the variability in scores of the RMobS are attributed to variability across the Brazilian states (ICC = .009). A one-way analysis of variance (ANOVA) on participants’ relational mobility latent factor scores was statistically significant, as expected given the large sample size, but the effect size was very small, F(26, 7316) = 3.32, p < .001, η2 = .012. Similarly, a one-way analysis of variance (ANOVA) on participants’ relational mobility latent scores across the five geo-socio-geographic regions indicated a statistically significant but very small effect, F(4, 7338) = 3.78, p = .005, η2 = .022. The lack of variability is evidenced by the respective factor and raw scores of relational mobility aggregated for each state and region reported in Table 6 (see also S1 File for a visual inspection). Hence, these results indicate homogeneity in relational mobility across regions and states in Brazil.
Relational mobility and outcome variables.
Results from both studies indicate low variability in relational mobility scores in Brazil. At the same time, the conceptualization and measurement of relational mobility as a socio‐ecological construct implies that individuals’ perceptions of relational mobility in their surroundings (a contextual, level‐2 variable) can have downstream consequences on individual difference measures (level‐1 variables). We thus employed multilevel analyses to examine the associations between relational mobility (modelled using latent means at the state level), and intimacy, self-disclosure and interpersonal similarity (modelled at the individual level). We included gender, age, and household socio-economic level as individual-level control variables. Excluding these variables from the models did not alter the main findings and interpretations, and in most cases, the coefficients were stronger when these variables were excluded (all results with and without covariates, as well as the model syntax and output, are available here: https://osf.io/9xuvy/).
Table 7 presents the results of multilevel analyses for each target at the state level. Latent mean levels of relational mobility at the state level were significantly and positively associated with self-disclosure, similarity and intimacy when the target was a close friend. No reliable associations were observed for state-level relational mobility and these variables when the target was a romantic partner. This is unsurprising since relational mobility is conceptually more relevant for friendship than romantic relationships , and Thomson et al.  also observed stronger associations between country-level relational mobility and all three measures when the target was a close friend than when it was a romantic partner. Our results thus replicate these findings across states within a single country.
Relational mobility is a socio-ecological construct quantifying how much freedom and opportunity a society affords individuals to choose and dispose of interpersonal relationships, and a large-scale study recently showed meaningful variability in relational mobility across 39 societies . We reported two exploratory studies examining variability in relational mobility within Brazil, which is a continent-size country consisting of 27 geographically and culturally distinct states and 5 geo-socio-political regions.
First, results from both studies confirmed the construct validity of the RMobS––a 12-item measure accessing individuals’ perceptions of how easy or difficult it is for people in their social and ecological setting to enter into, move out of, and form new relationships . Moreover, results from Study 2 confirmed the measurement equivalence of the measure for participants across all 27 Brazilian states. Together with the cross-cultural findings reported by Thomson et al. , the results from this study provide further evidence of the psychometric properties of the scale as well as its within-culture validity.
Second, Thomson et al.  observed that levels of intimacy, similarity and self-disclosure toward a close friend were higher in relationally mobile societies. Study 2 replicated these findings by showing a positive association between these three pro-active tendencies that help people retain relationships and relational mobility in Brazil. Taken together, these findings support hypotheses regarding the psychological consequences of relational mobility [see 23]: friends tend to feel closer intimacy, more similar to their close friends, and reveal serious personal information to a larger degree in social contexts where opportunities to find desirable and terminate undesirable relationships are greater.
Notably, the main goal of our study was to explore variability in relational mobility within Brazil. Overall our results confirmed the very high degree of individual-level variability observed by Thomson et al.  and did not find any meaningful variance within Brazil. Indeed, our results indicate that relational mobility does not vary across regions (Studies 1 and 2) or states (Study 2) in Brazil. This reveals that our participants’ perceptions of the amount of opportunities Brazilians have to select new relationship partners within their social context are fairly uniform across Brazil, and are not dependent on one’s residence in any particular state or region.
This uniformity is consistent with anthropological and sociological work arguing that despite decades of migration in Brazil, a single civic and political entity emerged, which has eventually lead to cultural uniformity and a unitary national conscience [16, 24]. It is thus possible to argue for a common Brazilian national culture where states in Brazil are “much more similar to each other than to Latin American countries, let alone countries worldwide” (18 p. 347). Our findings provide support for this view by showing that the common Brazilian national culture also includes abundant freedom and opportunity afforded to Brazilians to choose and dispose of interpersonal relationships, which might explain their noted sociability and hospitality .
As indicated in Fig 1, relational mobility theoretically comprises the two interrelated subcomponents of “meeting” and “choosing”. The interplay and relationships between the two subcomponents are well outside the scope of the current paper but, following a reviewer’s request, we conducted the same analyses reported in Table 6 for the subcomponents (see S1 File). When split into its constituent subcomponents, only the choosing factor shows consistent and statistically significant associations with intimacy, self-disclosure and interpersonal similarity. In line with findings regarding the higher-order relational mobility factor, this is only the case for dependent variables where the target is a close friend. Moreover, the finding of a statistically significant association between the choosing factor (but not the meeting factor) and intimacy is consistent with results found by Yamada, Kito and Yuki . These findings should be explored in future studies to tease out the dynamics of choice and opportunity in relational mobility’s role in interpersonal behavior.
It is worth noting the gender distribution in our study was heavily skewed towards women (92% and 85.6% female participants in Study 1 and 2). We believe this reflects self-selection bias in the sampling strategy used with Facebook advertisement as well as the survey focus on close interpersonal relationships. The unbalanced gender distribution was similar to the 39-nation study by Thomson et al.  where 86.2% of the sample were women, but their analyses showed that gender explained less than .05% the variance in relational mobility scores. We control for gender in the critical analyses reported in Table 6, but future studies should seek a better gender distribution in their sampling. Finally, we used an online tool to identify the participants’ missing states from their IP addresses in order to increase the sample size across the Brazilian regions (representing 28.9% and 27.9% of the final samples in Studies 1 and 2, respectively). Although we believe this was a reasonable strategy, it is possible that some participants masked their correct IP location when completing the survey and/or that the tool did not accurately identified the state location of some participants. Future studies should replicate our findings across all Brazilian states with a better measure of participants’ actual location.
To conclude, the present study provides the first systematic within-nation examination of relational mobility. The findings indicate very high degree of individual-level variability and very low degree of within-country variance in relational mobility in Brazil. Hence, our findings provide support for a common Brazilian national culture regarding relational mobility. Considering the links of relational mobility with historical variations in subsistence styles and ecological threats , thinking styles , and faster spread of infectious viruses , we hope this study will motivate further relational mobility research within and across nations.
- 1. Yuki M, Schug J, Horikawa H, Takemura K, Sato K, Yokota K, et al. Development of a scale to measure perceptions of relational mobility in society. Work. Pap., Hokkaido Univ., Sapporo, Japan. 2007. Available from https://lynx.let.hokudai.ac.jp/cerss/workingpaper/download/75.html
- 2. Yuki M, Schug J. Relational mobility: A socioecological approach to personal relationships. In: Gillath O, Adams G, Kunkel A, editors. Relationship science: Integrating evolutionary, neuroscience, and sociocultural approaches. Washington, DC: American Psychological Association; 2012. p. 137˗151.
- 3. Yuki M, Sato K, Takemura K, Oishi S. Social ecology moderates the association between self-esteem and happiness. J Exp Soc Psychol. 2013; 49(4): 741˗6. https://doi.org/10.1016/j.jesp.2013.02.006
- 4. Thomson R, Yuki M, Talhelm T, Schug J, Kito M, Avanian A, et al. Relational mobility predicts social behaviors in 39 countries and is tied to historical farming and threat. Proc Natl Acad Sci U S A. 2018; 115(29): 7521˗6. pmid:29959208
- 5. Schug J, Yuki M, Maddux W. Relational mobility explains between- and within-culture differences in self-disclosure to close friends. Psychol Sci. 2010; 21(10): 1471˗8. pmid:20817913
- 6. Hofstede G. Culture’s consequences: Comparing values, behaviors, institutions and organizations across nations. Thousand Oaks, CA: Sage; 2001.
- 7. Schwartz SH. Cultural value differences: Some implications for work. Applied Psychology: An International Review. 1999; 48: 23˗48. https://doi.org/10.1111/j.1464-0597.1999.tb00047.x
- 8. Leung K, Bond MH. Social axioms: A model for social beliefs in multi-cultural perspective. Adv Exp Soc Psychol. 2004; 36: 119˗97. https://doi.org/10.1016/S0065-2601(04)36003-X
- 9. Gelfand MJ, Raver JL, Nishii L, Leslie LM, Lun J, Lin BC, et al. Differences between tight and loose cultures: A 33-nation study. Science. 2011; 332(6033): 1100˗4. pmid:21617077
- 10. Harrington J, Gelfand M. Tightness–looseness across the 50 united states. Proc Natl Acad Sci U S A. 2014; 111(22): 7990–5. pmid:24843116
- 11. Vandello JA, Cohen D. Patterns of individualism and collectivism across the United States. J Pers Soc Psychol. 1999; 77(2): 279–92. http://dx.doi.org/10.1037/0022-3518.104.22.1689
- 12. Talhelm T, Zhang X, Oishi S, Shimin C, Duan D, Lan X, et al. Large-scale psychological differences within China explained by rice versus wheat agriculture. Science. 2014; 344(6184): 603˗8. pmid:24812395
- 13. IBGE, Instituto Brasileiro de Geografia e Estatística [Internet]. ww2.ibge.gov.br. 2019 [cited 24 August 2019]. Available from: https://ww2.ibge.gov.br/english/geociencias/cartografia/default_territ_area.shtm
- 14. Neto L. IBGE releases population estimates of municipalities for 2017 [Internet]. IBGE—Agência de Notícias. 2019 [cited 24 August 2019]. Available from: https://agenciadenoticias.ibge.gov.br/en/agencia-press-room/2185-news-agency/releases-en/16134-ibge-releases-population-estimates-of-municipalities-for-2017
- 15. Ribeiro D. [The Brazilian people: The formation and meaning of Brazil]. São Paulo: Companhia das Letras; 1995.
- 16. Fischer R, Milfont TL, Gouveia VV. Does social context affect value structures? Testing the intra-cultural stability of value structures with a functional theory of values. J Cross-Cult Psychol. 2011; 42(2): 253˗70. https://doi.org/10.1177/0022022110396888
- 17. Hofstede G, de Hilal AVG, Malvezzi S, Tanure B, Vinken H. Comparing regional cultures within a country: Lessons from Brazil. J Cross-Cult Psychol. 2010; 41(3): 336˗52. https://doi.org/10.1177/0022022109359696
- 18. Torres CV, Porto JB, Vargas LM, Fischer R. A meta-analysis of basic human values in Brazil: Observed differences within the country. Revista Psicologia: Organizações e Trabalho. 2015; 15(1): 89˗102. https://doi.org/10.17652/rpot/2015.1.356
- 19. de Oliveira S, Nisbett RE. Beyond East and West: Cognitive style in Latin America. J Cross-Cult Psychol. 2017; 48(10):1554˗77. https://doi.org/10.1177/0022022117730816
- 20. Parker RG. Bodies, pleasures and passions: Sexual culture in contemporary Brazil. Boston: Beacon; 1991.
- 21. Kanemasa Y, Daibo I. Effects of early adult attachment styles on intimate opposite-sex relationships. Japanese J Soc Psychol. 2003; 19(1): 59˗76. https://doi.org/10.14966/jssp.KJ00003724884
- 22. Schug J, Yuki M, Horikawa H, Takemura K. Similarity attraction and actually selecting similar others: How cross-societal differences in relational mobility affect interpersonal similarity in Japan and the USA. Asian J Soc Psychol. 2009; 12(2): 95˗103. https://doi.org/10.1111/j.1467-839X.2009.01277.x
- 23. Yuki M, Schug J. Psychological consequences of relational mobility. Curr Opin Psychol. 2020; 32: 129˗132. https://doi-org.ezproxy.waikato.ac.nz/10.1016/j.copsyc.2019.07.029
- 24. DaMatta RA. For an anthropology of the Brazilian tradition, or “A virtude está no meio.” In: Hess DJ, DaMatta RA, editors. The Brazilian puzzle: Culture on the borderlands of the Western world. New York: Columbia University Press; 1995. p. 270˗292.
- 25. Yamada J, Kito M, Yuki M. Relational mobility and intimacy in friendships and romantic relationships: A cross-societal study between Canada and Japan. Jpn J Exp Psychol. 2015; 55(1): 18˗27. https://doi.org/10.2130/jjesp.1409
- 26. San Martin A, Schug J, Maddux WW. Relational mobility and cultural differences in analytic and holistic thinking. J Pers Soc Psychol. 2019; 116(4):495˗518. pmid:30614727
- 27. Salvador C, Berg M, Yu Q, Alvaro SM, Kitayama S. Relational mobility predicts the increased speed of the spread of COVID-19: A 38 country study. PsyArXiv; 2020.