Cultural processes and social networks (Path A, Fig 1).

Ruiz and colleagues [1] propose that Latinxs’ cultural processes are associated with different social network characteristics that distinguish them from the White-European American culture. Latinxs are collectivistic and value social interactions with others [2, 3]. Latinxs also value familism, which suggests that Latinxs have strong identification and attachment with their nuclear family and extended families [7–9]. For example, Valenzuela and Dornbusch [11] asked Mexican-origin and White-European adolescents living in the United Stated to indicate the number of adult relatives living within an hour’s drive of the respondent’s home on a six-point scale from “none” to “20 or more.” The results showed that Mexicans-origin adolescents reported between six and ten adult relatives living close to home, as opposed to between three and five among White-European adolescents. Similarly, Almeida and colleagues [12] compared foreign-born Latinx, U.S.-born Latinx and non-Latinx white samples in terms of perceived social support from friends and family. The authors found that foreign-born Latinxs were more likely to perceive higher social support compared to the other samples and this difference remained significant even after controlling for socioeconomic status.

Latinxs show this association between their cultural processes and their socialization with others in their everyday natural behaviors as well. For example, Ramírez-Esparza and colleagues [2] asked undergraduate students living in Mexico and undergraduate student living in the United States to wear a digital recorder to capture snippets of ambient sounds as they went about their lives. The authors then asked coders to identify different social interaction categories and the results demonstrated that cultural processes were associated with different forms of socialization across the groups. Specifically, Mexicans, in accordance with collective values, spent more time in face-to-face interactions (both in dyads and in groups), whereas White-European Americans, in accordance with individualistic values, spent more time having remote interactions (e.g., on the phone) and more private interactions (e.g., in their home as compared to public places).

The above mentioned studies inform about the association between cultural processes and social networks by using ethnicity (i.e., the cultural group individuals identified with) as a proxy of cultural processes. However, other studies provide evidence for this association by including actual assessments of cultural processes. For example, Campos and colleagues [9] found that self-reported familism among Latinxs living in the U.S. is strongly associated with higher perceived social support, than among individuals from European backgrounds. In a more recent study, Campos, Rojas Perez and Guardino [13] compared a sample of Latinxs with samples from European and East-Asian backgrounds and found that only among the Latinx sample, self-reported familism was associated with higher romantic relationship quality through an association with more secure attachment.

Social networks and health (Path B, Fig 1).

Ruiz and colleagues [1] provide evidence for the association between social networks and health outcomes, including other psychosocial indicators of health and well-being. It is well-known that social support buffers the effects of psychological stress on a diverse range of health outcomes [14]. For example, research has shown that lower levels of support following a heart attack are associated with a greater risk of subsequent cardiac morbidity and mortality [15]. Furthermore, social support is associated with lower levels of stress and lower physical pain [16], and lower systolic blood pressure [17]. Close relationships (e.g., with friends and acquaintances) have also been shown to have a positive effect on health [18]. Ruiz and colleagues [1] argue that although most of the studies that have tested the relation between social networks and health outcomes do not speak directly to the relation between culture and health, they do leave open the possibility that social networks can be a common mediational pathway.

Cultural processes and health (Path C, Fig 1).

Finally, Ruiz and colleagues [1] propose that there is an association between cultural processes and health. That is, values like familism and collectivism lead to better health outcomes. Studies that have used ethnicity as a proxy for cultural processes have shown that Latinxs as a group experience better health and live longer than non-Latinxs [1, 19]. Latinxs’ cultural processes have been also associated with other behavioral and psychological indicators of health, such as well-being. Bowman [20] conducted a study with a diverse sample of freshman college students from colleges and universities across the U.S. and found that being Latinx was associated with greater well-being upon entering college. Adams and Boscarino [21] found that Hispanics were less likely than their White counterparts to develop symptoms of major depression in the aftermath of the September 11^th attacks, after controlling for meaningful variables such as individual stressors. At the macro level Latin American countries exhibit higher levels of perceived well-being, compared to Non-Latin countries of similar income level [22].

Other relevant studies have assessed well-being in combination with actual cultural processes. For example, Campos and colleagues [9] found that pregnant Latina women in comparison with White-European pregnant mothers reported higher familism; and familism contributed to better social support, as well as reduced stress and pregnancy anxiety. Similarly, Schwartz and colleagues [23] carried out a study that included participants who identified themselves as White-European American and participants who identified themselves from other ethnic groups (e.g., Latinxs, Blacks, Asians). The results showed that although White-Europeans endorsed collectivistic values to a lesser extent than other ethnic groups, they also benefited from endorsing collectivism. That is, across ethnic groups, collectivist values was associated with well-being outcomes.

Summary.

There is evidence that Latinxs’ cultural processes lead to higher social cohesion and stronger social networks, which both arguably derive from distinct cultural processes relative to White-European Americans. There is further evidence that both being a Latinx and having more cohesive social networks are each associated with more positive physical health and well-being outcomes. The sociocultural resilience model suggests that stronger social networks have a positive mediating influence on these health outcomes, ultimately stemming from the influence of cultural processes. A couple of published studies lend support to the interplay of variables that fit within the sociocultural resilience model.

Campos and colleagues [24] assessed as cultural process, familism in White-European, Latinx, and Asian samples, and as social networks, the authors used assessments on perceived closeness with family members, and social support. Finally, as a health outcome they included an overall assessment of psychological health (i.e., a combination of perceived stress, general mental health, and depressive symptoms). The results showed that as expected Latinxs scored higher in familism; however, all groups benefited from endorsing more familism. Specifically, higher familism was indirectly associated with higher psychological health via a pathway through both; higher perceived closeness with family members, and higher perceived social support. In a recently published study, Campos and colleagues [25] found that familism buffered cortisol responses via higher perceived social support. In this study, however, this association was only beneficial for Latinxs as compared to other ethnic groups (i.e., White-European, and Asians).

The Latinx health paradox is widely supported in the literature; however, as cultural scientists point out, the understanding of the interpersonal/cultural mechanisms that contribute to this phenomenon are still in its beginnings [1, 26, 27]. In this study, we sought to provide additional evidence in support of the sociocultural resilience model. Specifically, we compared Latina mothers with White-European mothers in terms of the quality of the conversations they have in their natural environments. We also examined how the quality of these interactions relate to one specific psychosocial indicator of well-being: behavioral laughter, as it occurs in natural environments. Although we did not examine a physical health outcome, we propose here that laughter is a behavioral indicator of, or precursor to, well-being analogous to other psychological indicators highlighted by Ruiz and colleagues’ sociocultural resilience model (e.g., stress, smoking). In other words, smoking can be regarded as a behavior that leads to poor health, but it is not a physical assessment of physical or psychological health in and of itself. Likewise, laughter is a behavior that can lead to a positive sense of well-being.

Laughter as a function of cultural processes.

Although laughter has not been studied before in the context of cultural differences, there is evidence showing that Latinxs value positive affectivity. For example, research has shown that Latinxs living in the U.S. report experiencing more positive emotions (such as joy, happiness, excitement) and less negative emotions (such as sadness, worry, irritation) as compared to other cultural groups (including European-Americans, Asian-Americans, Japanese and Indians) [34]. Similarly, Mexican exchange students in Canada endorse more excited states (such as elatedness, happiness) as compared to Chinese students in Canada who value more calmed states (such as serenity, restfulness) [35]. Latinx not only demonstrate positive affectivity in questionnaires, but they also conduct themselves in a manner that facilitates positive social relationships. Specifically, Latinx behave in a polite, respectful and positive way when they interact with others; they emphasize friendliness and prefer to avoid conflict by accentuating positive behaviors and deescalating negative behaviors that might lead to it [3–6]. This specific way of behaving has been defined as the cultural value of Simpatía, which has been used to describe a pattern of social interaction that characterizes people of Latin American descent [36]. In sum, Latinx value positive emotional expressivity, as measured by questionnaires, and they also behave in a manner that emphasizes positive social interactions; therefore, we can assume that in the service of fomenting positive social interactions, Latina mothers might laugh more than White-European mothers during those social interactions. This argument is further supported by research that demonstrates that laughter, in natural conversations, serves the purpose of facilitating the flow of the interaction and signals affiliation [37]. Similarly, laughter has been shown to be a stronger predictor of the quality of interactions, rather than an overall sense of positive affect [38].

In this study, we aimed to employ a novel approach and a unique focus on women, which could elucidate some of the complex relationships between psychosocial variables and overall well-being in women. Although the studies reviewed above on laughter and well-being have used different methodologies and are not void of nuances, we argue that the more opportunities mothers have to experience behavioral laughter as a function of the quality of their conversations in their everyday lives, the more likely they will experience better well-being.

Ethnicity and quality of the conversations (Path A).

We tested the relationship between ethnicity and small talk-father (Path A1); ethnicity with small talk-other(s) (Path A2); ethnicity and substantive conversations-father (Path A3); and ethnicity with substantive conversations-other(s) (Path A4). We expected an association between ethnicity and substantive conversations, and no association between ethnicity and small talk. Specifically, we hypothesized that being a Latina mother would be associated with substantive conversations. This expectation is consistent with the cultural value of familism, which suggests that Latinx have strong identification and attachment with their nuclear family and extended families [7–9] and tend to live close to their relatives [11]. Thus, if Latina mothers have more opportunities to interact with close others, then they would likely engage in more substantive interactions than White-European Americans.

Quality of the conversations and behavioral laughter (Path B).

We tested the relationship between small talk-father and behavioral laughter (Path B1); small talk-other(s) and behavioral laughter (Path B2); substantive conversations-father and behavioral laughter (Path B3); and substantive conversations-other(s) and behavioral laughter (Path B4). We expected that substantive conversations would be correlated with behavioral laughter, independent of the group. In contrast, we expected that small talk would be unrelated to behavioral laughter, independent of the group. This expectation was drawn from Mehl and colleagues’ study [43] where they reported a positive association between substantive conversations and well-being, but small talk was not associated with well-being. Although behavioral laughter was not directly assessed by Mehl and colleagues, we assume that laughter may occur more frequently when mothers have substantive conversations while interacting with adults indicating affiliation and closeness [37,38], and if mothers have more opportunities to have substantive conversations and laugh during those conversations then they will experience an overall sense of well-being [30, 31].

Ethnicity and behavioral laughter (Path C).

We tested the relationship between ethnicity and behavioral laughter (Path C). Consistent with the literature on the Latinx health paradox [1], we expected that being a Latina mother would be positively associated with behavioral laughter. Although behavioral laughter has not yet been examined as a Latinx well-being advantage, being a Latinx has been associated with positive psychological health outcomes [19–24]; and with Latinxs’ cultural processes associated with positive expressivity [3–6, 37, 38]. Furthermore, laughter has been linked to positive physical health outcomes [28], and well-being benefits [30].

Latinx families’ language and cultural characteristics.

Twenty-three of the 26 participants responded to a language background questionnaire.

Regarding the mothers, one mother was born in Peru, one in Puerto Rico, one in El Salvador, two in Venezuela, three in Colombia, seven in Mexico, and eight in the United States. Including the mothers who indicated that they were born in the United States, the average number of years that the mothers had been living in the United States was 16.20 years (SD = 9.68). Ten mothers preferred to use Spanish in daily life, four preferred English, and nine preferred both languages.

Regarding the fathers, one father was born in Peru, one in El Salvador, one in Guatemala, one in Venezuela, one in Ecuador, two in Colombia, two in Puerto Rico, seven in Mexico, and seven in the United States. Including the fathers who indicated that they were born in the United States, the average number of years that the fathers had been living in the United States was 14.79 years (SD = 10.33). Thirteen fathers preferred to use Spanish in daily life, five preferred English, and five preferred both languages.

Data collection.

Families received two digital language processors (DLPs) and an armband to hold the DLP. Of the 26 Latinx families, 25 mothers wore the recorder and one father wore the recorder; and of the 24 White-European families, 21 mothers wore the recorder and three fathers wore the recorder. They were instructed to record eight continuous hours each day for four days (including two weekdays and two weekend days), yielding approximately 32 hours of recorded audio from each family. The participant who wore the recorder was also asked to complete a daily activity diary, noting the most relevant activities for each day. In the diary, the participant was asked to enter the day of the week, the date, and what they were doing in each hour of the day (e.g., at home having lunch, at a friends’ house at a playdate, having dinner, etc.) for each of the four days that they wore the recorder.

Data preparation.

The audio data were transferred from the DLP to a computer and analyzed by LENA software. The software was used to analyze language input and to efficiently locate intervals with the language activity of interest (i.e., adult speech) in each participant’s dataset to use for further conversation-quality analysis. The LENA algorithms produced a total adult word count across all four days for each participant in the study. The accuracy of these values for the English language has been estimated in previous studies [46, 47]. For the Spanish language, Weisleder and Fernald’s [48] research team transcribed 60-minute samples from ten participants in their study. Their analysis of these transcriptions showed a strong positive correlation (i.e., r = .80) between automated estimates and transcribed word counts.

We then used the same approach as other studies [39–42]. Specifically, using the LENA Advanced Data Extractor Tool (ADEX) each participant’s dataset of recorded audio was segmented into 30-second intervals, to automatically estimate an adult word count for each interval. For example, an eight-hour recording yields approximately 600–960 intervals with adult word counts after the data are segmented into 30-second intervals. Intervals with no adult words were removed, and 50 intervals were selected from the remaining intervals across the entire day (i.e., 200 across the four days), chosen from those with the highest adult word counts. Although the adult word count is an estimate done by the software, intervals were selected based on adult word count to ensure that there is language activity that will allow coding of social behaviors. Using this approach, we avoided selecting intervals for coding where there was no social activity, but only silence or noise (e.g., the adult is alone at home; the adult wasn’t wearing the recorder). Furthermore, we segmented the data based on 30-second intervals, since other studies have assessed social behaviors in adults from 30-second snippets [2, 49, 50] and was employed in previous studies with infants and children [39–42]. The average word count across the selected intervals for the Latina mothers was 77.77 (SD = 25.48) and for the White-European mothers was 84.72 (SD = 37.66). No significant differences were found across the groups (t = .78).

Compliance.

Ideally, the final data set would include for coding a total of 200 intervals for each participant; however, sometimes fewer intervals were selected if data was not available for participants (e.g., participant did not wear the recorder). On average, 190.42 (SD = 17.28) intervals were coded for the Latinx families and 184.25 (SD = 33.05) for the White-European families. No significant difference was found between the number of intervals coded across groups (t = .84, p = .40). Across all 26 Latinx families a total of 4,951 intervals were coded and across all 24 White-European families, a total of 4,422 intervals were coded.

Intervals were excluded from analyses if there were problems noted during coding of the recording (e.g., participant wasn’t wearing the recorder, excessive noise; M = .03, SD = .03 for the Latinx families; M = .04, SD = .05 for the White-European families). An average of 184.38 (SD = 17.71) intervals for the Latinx families and an average of 177.46 (SD = 33.17) intervals for the White-European were included in the analyses. No significant difference was found between the number of intervals included for analyses between groups (t = .93, p = .36).

As a behavioral marker of obtrusiveness, research assistants coded how often the participants mentioned the recorder to other people during recording. On average, White-European families mentioned the recorder 2.73% of the time (SD = 1.71%) and on average Latinx families mentioned the recorder 1.91% of the time (SD = 1.58). This replicates past studies showing that behavioral data collection using recorders operates unobtrusively [51, 52].

Mean differences by group.

In order to examine mean differences as a function of ethnic group (i.e., Latina mothers vs. White-European mothers), we performed t-tests for each of the raw categories and transformed categories (see Table 2 the t-values and effects sizes). Results showed that for the raw categories the percent of coded intervals for substantive conversations was significantly higher (t = 2.05, p < .05) for the Latina mothers (Mean = 49.89, SD = 14.13) than White-European mothers (Mean = 39.82, SD = 20.31). Likewise, the percent of coded intervals for laughter overall was significantly higher (t = 3.32, p < .01) for the Latina mothers (Mean = 29.74, SD = 8.35) than White-European mothers (Mean = 20.24, SD = 11.52). For the transformed categories the percent of substantive conversations-other(s) was significantly higher (t = 2.25, p < .05) for the Latina mothers (Mean = 22.59, SD = 9.68) than the White-European mothers (Mean = 15.88, SD = 11.41). Likewise, the percent of laughter that occurs when talking to the father and/or other adult(s) was significantly higher (t = 2.19, p < .05) for the Latina mothers (Mean = 16.21, SD = 6.08) than White-European mothers (Mean = 11.74, SD = 8.23). None of the other categories were significantly different as a function of the ethnic group.

Mean differences by social networks and ethnicity.

In order to learn if the mean levels for each conversational style variables and behavioral laughter change as a function of social networks and as a function of the ethnic group, a repeated measures ANOVA was completed independently for the percentage of small talk, substantive conversations and behavioral laughter. Social networks were examined as within-participants (i.e., father vs. other(s)) and ethnicity (Latina vs. White-European mothers) was examined as between-participants. Since we do not assume sphericity, we report F-values after using the Greenhouse-Geiser correction.

Concerning small talk, the social network by ethnicity interaction was not significant F(1, 48) = 2.80, p = .10, η_p² = 0.06. Furthermore, the main effects of social networks and ethnicity were not significant, F(1, 48) = 2.57, p = .12, η_p² = 0.05 and F(1, 48) = .10, p = .75, η_p² < .001, respectively (also see Fig 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Differences for social networks and ethnicity.

Main effects and interactions for social networks (i.e., father vs. other(s) and ethnicity (i.e., Latina mothers vs. White-European mothers). Error bars represent standard errors from the mean.

https://doi.org/10.1371/journal.pone.0214117.g003

Concerning substantive conversations, the social network by ethnicity interaction was not significant F(1, 48) = 1.59, p = .21, η_p² = 0.03. However, the main effects for social networks and ethnicity were significant, F(1, 48) = 6.45, p = .01, η_p² = 0.12, and F(1, 48) = 5.22, p = .03, η_p² = 0.10, respectively. Fig 3 shows that both Latina and White-European mothers have more substantive conversations with other adult(s) than with the father. Furthermore, Latina mothers have more substantive conversations than do White-European mothers.

Concerning behavioral laughter, the social network by ethnicity interaction was not significant F(1, 48) = 0.04, p = .84, η_p² < .001. However, main effects for social networks and ethnicity were both significant, F(1, 48) = 9.67, p = .003, η_p² = 0.17; and F(1, 48) = 4.56, p = .04, η_p² = 0.09, respectively. Fig 3 shows that both Latina and White-European mothers laugh more with other adult(s) than with the father. Furthermore, Latina mothers laugh more than White-European mothers do.

Intercorrelations among conversation style variables.

In order to learn about the intercorrelations among the conversational style variables, we conducted correlation analyses between the four conversation style variables independently for Latina and White-European mothers. Table 3 (columns 2–4) shows that for the Latina mothers there was a significant correlation between small talk-others and substantive conversations-other(s). Latina mothers who have more small talk with others also have more substantive conversations with others (r = 0.57, p < .01). For White-European mothers, small talk-father correlated significantly with small talk-other(s), substantive conversations-father and small talk-other(s). Mothers who have more small talk with the father also have more small talk with other(s) (r = 0.80, p < .001); less substantive conversations with the father (r = -0.48, p < .05), and with others (r = -0.43, p < .05). Finally, small talk-others correlated significantly with substantive conversations-others. Mothers who have more small talk with others use less substantive conversations with the father (r = -0.54, p < .01).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Correlations between conversation style variables and laughter for both Latina and White-European mothers.

https://doi.org/10.1371/journal.pone.0214117.t003

In order to learn about the association between conversational style variables and behavioral laughter, we conducted correlation analyses between the four conversation style variables and the three laughter outcomes independently for both Latina and White-European mothers. Table 3 shows that when considering laughter across different social networks (columns 6 and 7), significant correlations were found for Latina mothers between small talk-father and laughter-father (r = 0.63, p < .001); small talk-other(s) and laughter-other(s) (r = 0.62, p < .001); substantive conversations-father and laughter-father (r = 0.78, p < .001); substantive conversation-other(s) and laughter-other(s) (r = 0.79, p < .001). For White-European mothers, a significant correlation was only found between substantive conversations-other(s) and laughter-other(s) (r = 0.65, p < .001).

Furthermore, Table 3 shows that when considering laughter collapsed across social networks (column 8), for Latina mothers, significant correlations were found between small talk-father and laughter (r = 0.45, p < .05), as well as between substantive conversations-father and laughter (r = 0.45, p < .05). For White-European mothers, only the correlation between substantive conversations-other(s) and laughter was significant (r = 0.51, p < .01).