Latinx Immigrant Youth’s Indirect and Direct Disclosures About Their Family-Undocumented Experiences,Received Emotional Support,and Depressive Symptoms

Abstract

Drawing upon sensitive interaction systems theory, this study investigated Latinx immigrant youth’s indirect and direct disclosures about their family-undocumented experiences (i.e., informing others of their own or family’s undocumented experiences, including details surrounding their undocumented situation) to a teacher(s) or a friend(s). Furthermore, this study examined how such disclosures relate to subsequent disclosure and to received emotional support across an academic year. Received emotional support, in turn, was theorized to predict youth’s depressive symptoms. Survey data were collected in the beginning, middle, and end of the 2015-2016 academic year from 410 Latinx family-undocumented students (ninth-12th grades). Latinx students reported more indirect than direct disclosures about their family-undocumented experiences, potentially as a way to seek support. They also reported greater indirect and direct disclosure about their family-undocumented experiences to a friend(s) compared with a teacher(s). Emotional support did not significantly mediate the associations between disclosure and depressive symptoms; however, indirect and direct disclosures to a friend(s) positively related to depressive symptoms within various waves.

Keywords

disclosure support privacy undocumented immigrants Latino longitudinal

The Pew Research Center estimates that among the 40 million immigrants residing in the United States, 11 million (over 25%) of them are undocumented¹ (López & Bialik, 2017; Passel & Cohn, 2017). Although the number of undocumented immigrants who are under 18 years of age is comparatively small (estimated at 775,000 in 2012), nearly 50% of undocumented adults have at least one child under 18 years of age in their household (Passel & Cohn, 2009; Passel, Cohn, Krogstad, & Gonzalez-Barrera, 2014). Thus, the number of children with undocumented family members is quite large. Family-undocumented youth (i.e., youth who are either undocumented or who have an immediate family member who is undocumented) often experience fear of deportation for self and family, discrimination, and financial strain, as well as limited access to health care, higher education, and employment (Gonzales, Suárez-Orozco, & Dedios-Sanguineti, 2013; Scranton, Afifi, Afifi, & Gangi, 2016).

In stressful situations, such as being undocumented or having a family member who is undocumented, supportive communication can buffer against stress and help people flourish in the face of adversity—a process known as thriving (Feeney & Collins, 2015). Emotional support is particularly effective at reducing the psychosocial consequences that accompany a variety of stressors (Burleson, 1984). To receive the benefits of supportive communication, individuals often have to seek it by disclosing their problems to others (MacGeorge, Feng, & Burleson, 2011). Nevertheless, several qualitative studies reveal that immigrant youth rarely discuss their family’s undocumented status and their experiences related to being undocumented with nonfamily members, which inhibits their access to support (Dabach, 2015; Jefferies, 2014).

We do not assume that disclosure is inherently positive or necessary; however, some theories of supportive communication assert that disclosing information about the stressor might be a necessary precursor to receiving quality assistance (Barbee & Cunningham, 1995; Gonzales et al., 2013; MacGeorge et al., 2011). We know little about how immigrant youth share their family-undocumented experiences with nonfamily members, such as a teacher(s) (i.e., one or more teachers) or a friend(s) (i.e., one or more friends), and whether such disclosures activate emotional support over time (Kam, Gasiorek, Pines, & Steuber Fazio, 2018). To understand how disclosure, received emotional support, and psychosocial outcomes are related for immigrant youth, we utilize three waves of longitudinal survey data collected from 410 Latinx family-undocumented high school students (ninth-12th grades) in the beginning, middle, and end of the 2015-2016 academic year (see online appendix for sample justification). Drawing from sensitive interaction systems theory (SIST; Barbee & Cunningham, 1995), we first examine the extent to which Latinx immigrant youth indirectly and directly disclose about their family-undocumented experiences (i.e., sharing information about being undocumented, which can include disclosing their own or a family member’s undocumented status, as well as disclosing details about their experiences being undocumented) to a teacher(s) or a friend(s). Second, because people can disclose differently at various times (Catona, Greene, Checton, & Carpenter, 2016), we investigate how indirect and direct disclosures and received emotional support relate to each other over an academic year. Finally, we consider how indirect and direct disclosures and received emotional support relate to depressive symptoms for family-undocumented youth.

SIST and Support Activation

MacGeorge et al. (2011) defined seeking support as “intentional communicative activity with the aim of eliciting supportive actions from others” (p. 330). People often disclose information about a stressor to seek support (Bodie, Vickery, Cannava, & Jones, 2015; Derlega, Winstead, Mathews, & Braitman, 2008). We consider disclosing about a family’s undocumented experiences as a precursor to and one means of eliciting supportive communication.

SIST (Barbee & Cunningham, 1995) postulates that one of the strongest predictors of received support is how people seek it. Although SIST suggests that seeking support can vary in additional ways, the direct versus indirect dimension has received the most research attention (Goldsmith, 1995; High & Scharp, 2015; Youngvorst & High, 2018). Whereas indirectly seeking support can be vague and might not specify the assistance people desire, directly seeking support describes a stressor and expresses a clear desire for comfort or advice (Barbee & Cunningham, 1995). This study does not consider indirect and direct support seeking, but it does examine the extent to which indirect and direct disclosures about family-undocumented experiences relate to emotional support. Despite not focusing on seeking support per se, SIST is still a valuable theoretical lens for this investigation because one of its central tenets is that how people talk about their stressors shapes the support they receive. Immigrant youth can indirectly disclose their family-undocumented experiences by engaging in incremental disclosures and hints about their family being undocumented and related experiences (e.g., challenges to being undocumented; Kam, Gasiorek, et al., 2018). They might also directly disclose by overtly telling a teacher or a friend about their family’s situation as undocumented immigrants (Kam, Steuber Fazio, & Mendez Murillo, 2018). Thus, a question that emerges is as follows: Do indirect and direct disclosures about family-undocumented experiences garner emotional support?

Recent research on supportive communication, and the reception of emotional support in particular, recognizes that supportive messages do not occur in isolation but are embedded in conversations among multiple parties (Bodie, Jones, Vickery, Hatcher, & Cannava, 2014). This recognition requires new considerations of how supportive communication operates. For example, whereas SIST suggests that messages used to seek support are either indirect or direct, studying support in conversation must acknowledge that people can seek support both directly and indirectly throughout the course of an interaction (i.e., the possible use of multiple strategies). Thus, when SIST is applied to disclosure, a series of direct and/or indirect disclosures (e.g., indirectly→directly, directly→indirectly, indirectly→indirectly, or directly→directly) can occur within a conversation or over several different conversations with the same or different potential support providers (Catona et al., 2016). This study extends SIST’s notion of directly or indirectly seeking support by measuring directly and indirectly disclosing as separate behaviors (rather than as opposite poles on a continuum) and assessing how people directly and indirectly disclose over time.

Potential Dangers to Disclosing About One’s Family-Undocumented Experiences

Although SIST (Barbee & Cunningham, 1995) suggests that indirect and direct support seeking can garner supportive resources from others, disclosing one’s family-undocumented experiences can also lead to negative consequences (Scranton et al., 2016). Thus, immigrant youth might not discuss their family-undocumented experiences with others for fear of deportation, negative evaluation, or because their family would be angry with them for sharing a secret (Kam, Steuber Fazio, & Mendez Murillo, 2018). Clearly, disclosing one’s family-undocumented experiences and talking about one’s situation comes with risks; however, in some cases, immigrant youth might disclose to access supportive resources (Gonzales, 2011; Jefferies, 2014).

People who believe they have a highly stigmatized identity tend to disclose private information indirectly (Afifi & Steuber, 2009). Given the vulnerability of disclosing one’s family-undocumented experiences, immigrant youth are more likely to disclose their family-undocumented experiences to a teacher(s) and a friend(s) indirectly rather than directly. Indeed, Kam, Gasiorek, et al. (2018) found that Latinx immigrant youth primarily used indirect disclosures, as opposed to direct disclosures, to inform school counselors of their family-undocumented status.²

When and how people seek supportive communication for a stressor also often differs based on the source they approach for support. For example, friends are often assumed to be more accessible and understanding sources of support compared with more formal support providers (e.g., teachers), or sources based on involuntary relationships (e.g., parents; Catona, Greene, & Magsamen-Conrad, 2015; Derlega, Winstead, Oldfield, & Barbee, 2003). This results in more direct disclosure to friends relative to other sources (Knobloch & Solomon, 2002). Immigrant youth, then, are likely more inclined to indirectly disclose to a teacher(s) compared with a friend(s) because of the greater relational intimacy characterizing friendships and due to the asymmetrical power distribution within the student-teacher relationship. Similarly, a friend(s) is likely to facilitate more direct disclosures than a teacher(s). Thus, the following hypotheses were set forth:

Hypothesis 1 (H1): Latinx immigrant youth will report greater indirect than direct disclosure about their family-undocumented experiences.

Hypothesis 2 (H2): Latinx immigrant youth will report greater indirect disclosure about their family-undocumented experiences with a teacher(s) than a friend(s).

Hypothesis 3 (H3): Latinx immigrant youth will report greater direct disclosure about their family-undocumented experiences with a friend(s) than a teacher(s).

The Processual Nature of Disclosure

In SIST, Barbee and Cunningham (1995) argued that the process of support involves five phases. First, someone with a problem seeks support. Second, a support provider offers supportive resources. Third, the support seeker processes those messages. Fourth, support seekers accept or reject the received messages, as well as clarify or restate their position. Finally, based on those reactions, support providers have another opportunity to reframe their assistance to the person in need. This multistage process reflects the reality that people likely disclose about or seek support at multiple times for the same stressor, and that they might enact these behaviors using different strategies both within and across conversations. Still, few studies actually embrace the processual nature of supportive communication. This study focuses specifically on disclosure as a means of eliciting emotional support. To date, little is known about how different means of disclosure and support reception are associated over time.

Afifi and Steuber (2009) proposed that the more individuals anticipate negative consequences to disclosing private information, the more likely they are to use indirect disclosures to “test the waters.” People can disclose indirectly and then gradually move to more direct disclosures. For example, immigrant youth might indirectly disclose their family-undocumented experiences to a teacher. If the teacher responds in a supportive way, the youth might be more inclined to directly discuss their family-undocumented experiences with that same teacher or a different teacher in the future because they might feel optimistic and encouraged. By contrast, if the teacher is unreceptive, the immigrant youth might be less likely to discuss their family-undocumented experiences with that same teacher or a different teacher in the future. If they do decide to disclose an aspect of their situation again, they might do so indirectly because they were discouraged by the first disclosure.

Another possibility is that immigrant youth might disclose directly about their family-undocumented experiences (e.g., “My mom and I will talk to a lawyer to get our papers.”) and then directly disclose additional information at a later date (e.g., “My mom wants to see my grandma in Mexico. But, if my mom visits my grandma, my mom might not be able to return because she doesn’t have papers.”). Given the processual nature of indirect and direct disclosures, we examine how disclosing varies according to both previous disclosures and the support received, recognizing that people can disclose directly and indirectly both within and across interactions. Based on the lack of prior evidence on which to base predictions, we propose the following research questions:

Research Question 1 (RQ1): How are immigrant youth’s indirect and direct disclosures about their family-undocumented experiences related over an academic year?

Research Question 2 (RQ2): How are immigrant youth’s indirect and direct disclosures about their family-undocumented experiences related to receiving emotional support over an academic year?

Linking Disclosure to Received Emotional Support and Depressive Symptoms

Although both indirect and direct disclosures might activate supportive communication (Mortenson, 2009), SIST suggests that they might produce varying degrees of success (Barbee & Cunningham, 1995). People who clearly indicate a need for support receive more sophisticated responses from support providers (Scott, Caughlin, Donovan-Kicken, & Mikucki-Enyart, 2013), and directly, relative to indirectly, seeking support corresponds with both increased incidence of supportive responses and decreased numbers of unsupportive comments (Williams & Mickelson, 2008). Applied to direct and indirect disclosures, it is plausible that directly disclosing makes the stressor clear, whereas indirectly disclosing renders the precise source of distress unclear. Consequently, potential support providers might not even recognize if a person requires support if they indirectly disclose about their stressor. Thus, although both direct and indirect disclosures might garner supportive resources, those resources are likely to be greater after direct disclosures. This thinking is reflected in the following hypothesis:

Hypothesis 4 (H4): The positive association between disclosure and received emotional support is stronger for direct disclosures compared with indirect disclosures.

A number of negative consequences can be alleviated if family-undocumented youth receive adequate emotional comfort. Although emotionally supportive messages cannot solve the problem by changing a family’s undocumented experiences, they can improve how youth feel about their family-undocumented experiences, perhaps by decreasing feelings of depression that often accompany this stressor (Gonzales et al., 2013). Along these lines, researcher shows that receiving emotional support reduces depression and increases people’s positive moods (Biehle & Mickelson, 2012). Accordingly, we propose the following hypothesis:

Hypothesis 5 (H5): Received emotional support will be negatively related to depressive symptoms.

To this point, we have proposed that disclosing a stressor is positively associated with received emotional support (H4), albeit more robustly for direct disclosures than indirect disclosures. We also hypothesized that received emotional support is negatively associated with depressive symptoms (H5). Taken together, these predictions describe a process in which disclosing about one’s family-undocumented experiences indirectly decreases depressive symptoms due to emotional support reception. Our final hypothesis tests this mediation model.

Hypothesis 6 (H6): Indirect and direct disclosures are indirectly and negatively related to depressive symptoms through received emotional support.

Method

This study comes from a larger longitudinal project on the stress, resilience, and thriving of immigrant youth from three high schools located in the western United States (Kam, Gasiorek, et al., 2018). Survey data were collected 3 times during the 2015-2016 academic year: November 2015 for Wave 1 (beginning of the academic year), in February 2016 for Wave 2 (middle of the academic year), and in May 2016 for Wave 3 (end of the academic year).

Participants

Over the three waves of data collection, a total of 658 immigrant high school students completed a survey. Using the school-district data, 85.9% (n = 550) of the 658 students were from a Latin American country. Other countries of origin included the Philippines, Sierra Leone, and Germany. For the present study, we only selected students who were of Latin American origin and (a) who self-identified as undocumented at any point across the three waves or (b) who reported that an immediate family member (parent or sibling) was undocumented at any point across the three waves. Based on these eligibility criteria, we had a sample of 410 Latinx family-undocumented high school students. The design of the larger project allowed participating students to enter or exit the study at any wave; therefore, among the 410 Latinx family-undocumented students, 16.6% completed a survey in only one wave, 37.3% completed a survey in two waves, and 46.1% completed a survey at all three waves.

With respect to the 410 students’ demographic breakdown, 92% were born in Mexico and 6% in El Salvador, with the remaining 2% born in another Spanish-speaking country. The average length of time students had resided in the United States was 7.32 years (SD = 4.04). Seventy-five percent identified as undocumented, 89% reported having a parent who was undocumented, and 83.5% reported that at least one sibling was undocumented. Fifty-five percent were male, and the mean age was 16.1 years (SD = 1.30). Ninety-nine percent participated in their school’s free-lunch program. The grade distribution was as follows: 36% in ninth grade, 24% in 10th grade, 23% in 11th grade, and 17% in 12th grade.

Procedures

After receiving approval from the university’s Institutional Review Board (IRB), the Department of Education website was used to locate public schools with a Latinx student population of at least 50%. To recruit participants, the first author called school principals, mailed information packets, and sent e-mails informing them of the study’s design and purpose. In response, a school district located in an agricultural city in a western state volunteered to participate to provide additional resources for its immigrant students. Three high schools comprise the district with approximately 7,000 total students enrolled. The Latinx student population ranged from 58% to 94% of the total student body across the three high schools.

Because the school district consists of thousands of students, only a portion of the immigrant students could be feasibly surveyed across the academic year. Thus, the first author and school district agreed to invite 700 immigrant students across the three schools to participate. The school district mailed an information letter, printed in Spanish and English, to the parents or guardians of 700 immigrant students. These students were selected based on how recently they immigrated to the United States until a sample size of 700 was met. The parent letter explained the study’s purpose and its voluntary nature, and it informed parents that their child would be invited to complete surveys 3 times during the academic year. The letter informed parents that their child would “answer questions on how they manage stressful experiences related to immigration (e.g., having a family member deported, migrating to the United States unaccompanied, and having to live in one country, while a parent(s) lives in another country).” Parents had 2 weeks to withdraw their child from the study by calling a university phone number created specifically for the study and leaving a voice message in any language of their choosing. For Waves 2 and 3, the school district used an automated phone messaging system to again inform parents in Spanish and English. Eight parents withdrew their children.

Prior to survey administration at each wave, the first author and research assistants described (in Spanish and English) the study’s purpose and its voluntary and confidential nature. At each wave, participating students completed an assent form and an online survey using a tablet or a school computer, which took approximately 30 to 45 minutes. We utilized Rogler’s (1989) back-translation method to establish translation fidelity between the Spanish and English surveys. Among the 410 Latinx immigrant students, 35% completed the survey in Spanish at Wave 1, 31% at Wave 2, and 32% at Wave 3. Participating students received a snack and university paraphernalia (e.g., stylus highlighter pens and notepads) at each wave.

Measures

For this study, we utilized condensed versions of previously-existing scales because of the time restrictions imposed by the schools’ class periods. Students responded to the disclosure questions 3 times, reflecting on a teacher(s), friend(s), and then a school counselor, respectively. Students responded to the emotional support questions 3 times, reflecting on a parent, a teacher(s), or, a friend(s), respectively. For this study, only items reflecting on a teacher(s) or a friend(s) were employed because school counselors were of focus for another manuscript (Kam, Gasiorek, et al., 2018), and the present study concentrates on disclosures to nonfamily members. In addition, Wave 1 asked students to report on past behaviors in general. Waves 2 and 3 asked about the time since the previous survey (i.e., 3 months). Descriptive data and reliability coefficients are presented in Table 1A (see online appendix).

Indirect disclosures of family-undocumented experiences

To measure indirect disclosures, students read the following introduction, “Thinking about your family member’s documentation status, how strongly do you disagree or agree with the following statements?” They then read two statements from Afifi and Steuber (2009): “You’ve provided pieces of information about your family’s situation to see how the following people [teacher(s) or friend(s)] would react,” and “You’ve revealed small details to the following people [teacher(s) or friend(s)] about your family’s situation to see how they would respond” (1 = strongly disagree to 5 = strongly agree).

Direct disclosures of family-undocumented experiences

To measure direct disclosures, students responded to two statements from Afifi and Steuber (2009): “You’ve directly told the following people [teacher(s) or friend(s)] about your family situation,” and “You’ve told the following people [teacher(s) or friend(s)] about your family situation, in person, face-to-face” (1 = strongly disagree to 5 = strongly agree).

Received emotional support

Four items, adapted from Xu and Burleson (2001), operationalized received emotional support. The four items asked how often (1 = never to 5 = very often) a teacher(s) or friend(s), “expressed understanding by sharing a similar situation that he or she experienced before,” “promised to keep problems you discuss secret,” “expressed sympathy for your situation,” and “provided you with hope or confidence.”

Depressive symptoms

At each wave, participants completed four items from the Center for Epidemiologic Studies Depression Scale (CES-D, Radloff, 1977). The four items asked how often participants, “felt depressed,” “felt that you could not stop feeling sad, even with help from family and friends,” “felt that you could not deal with the problems in your life,” and “had trouble staying focused on what you were doing” (never = 1 to very often = 5).

Control variables

During the panel analyses, we controlled for the following variables: school site (dummy coded), documentation status of adolescent (0 = no, 1 = yes, but a member of the immediate family is undocumented), school-reported sex (0 = male, 1 = female), school-reported grade level (ninth through 12th), and time spent in the United States at Wave 1.

Results

Missing Data

Of the 410 participants, 339 participated in Wave 1, 304 participated in Wave 2, and 298 participated in Wave 3. One-hundred and eighty-nine participants completed all three waves, 153 completed two of the three waves, and 68 completed only one wave. We conducted a Multivariate Analysis of Variance (MANOVA) with wave completion pattern (i.e., three-wave completion, two-wave completion, and single-wave completion) as the fixed factor and the main variables as the dependent. Results were not statistically significant, Wilks’ λ = .961, F = 1.059, p = .392, suggesting there were no significant differences on the main variables by different patterns of participation. All missing data for the confirmatory factor analyses and longitudinal structural equation modeling (SEM) were handled with full information maximum likelihood estimation (FIML), which is recommended for longitudinal data analysis (Little, 2013).

Longitudinal Confirmatory Factor Analysis

We tested the factorial validity and longitudinal invariance of our measures using a multistep process. We specified a series of latent factors across the three time points and constructed three increasingly restrictive models. The configural invariance model provides baseline fit, the weak invariance model constrains the factor loadings over time, and the strong invariance model further constrains the intercepts over time. The measurement is considered invariant over time if the change in the Confirmatory Fit Index (ΔCFI) between the increasingly stringent models is less than .01 (Little, 2013). Using AMOS 24.0 with maximum likelihood estimation, we ran two sets of models, one for the teacher(s) variables and one for the friend(s) variables. Both models also contained depressive symptoms. Good model fit and strong invariance standards were met for the teacher(s) and friend(s) models based on the fit statistics and ΔCFI from the configural to the weak to the strong models. The strong model fit for each model was as follows: teacher(s): χ² = 733.451, df = 526, p < .001, CFI = .967, ΔCFI from weak to strong models = .004, Tucker-Lewis Index (TLI) = .958, Root Mean Square Error of Approximation (RMSEA) = .031 (90% confidence interval [CI]: [.026, .036]); friend(s): χ² = 731.244, df = 526, p < .001, CFI = .969, ΔCFI from weak to strong models = .002, TLI = .961, RMSEA = .031 (90% CI: [.025, .036]).

Analytic Plan for Hypotheses and Research Questions

We tested H1, H2, and H3 with paired samples t tests for each wave. We tested the two research questions (RQ1 and RQ2) and H4 and H5 with two random intercept cross-lagged panel models (RI-CLPMs; Hamaker, Kuiper, & Grasman, 2015). RI-CLPMs are advantageous because they test cross-sectional and longitudinal associations at the between-person (i.e., stable relationships over time) and within-person (intraindividual change) levels.

Direct and Indirect Disclosure Differences

H1 proposed that family-undocumented youth will report greater indirect disclosure than direct disclosure. We tested this effect for teacher(s) and friend(s) separately and at each wave using paired samples t tests.³ For these tests, we lowered the p value for significance to .004 given the 12 tests conducted for H1 to H3. H1 was supported. To their teacher(s), in all three waves, students reported significantly higher indirect (M_W1 = 2.56, M_W2 = 2.27, and M_W3 = 2.40) than direct (M_W1 = 2.21, M_W2 = 2.00, and M_W3 = 2.11) disclosure (ts = 4.62 - 5.30, ps < .001). Furthermore, to friend(s), in all three waves, students reported significantly higher indirect (M_W1 = 2.92, M_W2 = 2.63, and M_W3 = 2.68) than direct (M_W1 = 2.72, M_W2 = 2.38, and M_W3 = 2.44) disclosure (ts = 2.974 - 3.80, ps < .001 - .003).

H2, which predicted that students would report greater indirect disclosure with a teacher(s) than a friend(s), was not supported. Instead, students reported significantly greater indirect disclosure with a friend(s) (M_W1 = 2.91, M_W2 = 2.61, and M_W3 = 2.68) than a teacher(s) (M_W1 = 2.56, M_W2 = 2.24, and M_W3 = 2.40) at all three waves (ts = −5.55 to −7.48, ps < .001).

H3 predicted that students would report greater direct disclosure with a friend(s) than with a teacher(s). H3 was supported. In all three waves, students reported greater direct disclosure with a friend(s) (M_W1 = 2.68, M_W2 = 2.37, and M_W3 = 2.44) than a teacher(s) (M_W1 = 2.19, M_W2 = 1.99, and M_W3 = 2.11; ts = −5.87 to −8.26, ps < .001).

Direct and Indirect Disclosures, and Emotional Support Over an Academic Year

Model setup and results

To test RQ1, RQ2, and H4 to H6, we set up two RI-CLPMs—one for teacher(s) and one for friend(s) with direct disclosure, indirect disclosure, emotional support, and depressive symptoms across the three waves. Each model was set up with four person-centered latent variables across the three time points related through intra-wave correlations and autoregressive and cross-lagged parameters over time. The autoregressive and cross-lagged parameters represent within-person change on the variables across time. The models also contain four random intercepts that are intercorrelated, which account for between-person correlations between the variables (Hamaker et al., 2015). The random intercepts load onto the observed means of the four variables at each of the three waves with loadings set to 1.0.

If tenable, Hamaker et al. (2015) recommend implementing several model constraints when setting up RI-CLPMs in efforts to preserve degrees of freedom, which can be limited in RI-CLPMs. For the teacher(s) and friend(s) models, tests of mean equivalence showed that models in which the Wave 1 means were unconstrained (while the Wave 2 and Wave 3 means were left constrained) fit the data significantly better based on Δχ² tests and ΔCFIs. Thus, for both models, only the W1 means were unconstrained, teacher: χ² = 9.081, df = 4, p = .059, CFI = .995, TLI = .903, RMSEA = .056 (90% CI: [.00, .105]), Δχ²(4) = 22.182, p < .001; ΔCFI = .018; friend: χ² = 1.196, df = 4, p = .879, CFI = 1.00, TLI = 1.05, RMSEA = .00 (90% CI: [.00, .036]), Δχ²(4) = 28.233, p < .001; ΔCFI = .019. Tests further showed that cross-time constraints were justified for the friend(s) model given that the constrained model fit the data well, χ² = 29.846, df = 26, p = .274, CFI = .997, TLI = .990, RMSEA = .019 (90% CI: .00, .045), and that the fit did not improve significantly when the time constraints were released, Δχ²(16) = 19.551, p = .241, ΔCFI = .003. Cross-time constraints were not justified, however, for the teacher(s) model given the unconstrained model—χ² = 16.872, df = 10, p = .077, CFI = .993, TLI = .948, RMSEA = .041 (90% CI: [.00, .074])—provided significantly better fit than the constrained model, Δχ²(16) = 39.974, p < .001, ΔCFI = .023. In sum, based on these tests, we added (a) partial mean constraints in both models and (b) cross-time constraints in the friend(s) model only.

With the appropriate model constraints determined, we constructed RI-CLPMs with the covariates of school site, documentation status of adolescent, sex, grade level, and time spent in the United States. The control variables were set as direct predictors of the W1 variables and correlated with the random intercepts. For the final models, we trimmed out direct control paths sequentially if they did not at least approach statistical significance (i.e., p < .10; see Little, 2013). The final models fit the data well for teacher(s), χ² = 72.889, df = 53, p = .036, CFI = .988, TLI = .960, RMSEA = .030 (90% CI: [.008, .046]), and friend(s), χ² = 59.860, df = 68, p = .749, CFI = 1.00, TLI = 1.012, RMSEA = .00 (90% CI: [.00, .022]). When interpreting the results from the RI-CLPMs, it is important to keep in mind that cross-time associations and intra-wave correlations are within-person (i.e., associations based on participants’ deviations from their expected levels). Random intercept correlations, however, are at the between-person level, or mean level. A positive between-person correlation indicates that participants who tended to report higher mean levels of one variable also tended to report higher mean levels of a second variable. Within-person parameters, intra-wave correlations, and between-person correlations are presented in Figure 1A (teacher model) and Figure 2A (friend model) in the online appendix.

RQ1

RQ1 asked how indirect and direct disclosures about family-undocumented experiences are related over time. In the teacher(s) model, there was a significant and positive longitudinal association between direct disclosure at W2 and indirect disclosure at W3 (B = .54, p < .001). Thus, youth who increased direct disclosure of their family-undocumented experiences to a teacher(s) at W2 also increased their indirect disclosure at W3. In the teacher(s) model, there were also significant intra-wave, or cross-sectional associations between direct and indirect disclosure at W1 (r = .43, p < .001), W2 (r = .56, p < .001), and W3 (r = .49, p < .001). In the friend(s) model, there were no significant cross-wave associations, but, consistent with the teacher(s) model, there were significant intra-wave correlations between direct and indirect disclosure at W1 (r = .47, p < .001), W2 (r = .42, p < .001), and W3 (r = .39, p < .001).

RQ2

RQ2 asked how indirect and direct disclosures are associated with received emotional support over an academic year. In the teacher(s) model, longitudinal results indicated that direct disclosure at W1 significantly positively predicted emotional support at W2 (B = .26, p < .007). Emotional support at W1, however, negatively predicted indirect disclosure at W2 (B = −.52, p < .011). Yet, intra-wave correlations showed that indirect disclosure and emotional support were significantly positively correlated within W1 (r = .37, p < .001) and W3 (r = .31, p < .001). There was also a significant and positive between-person correlation between the indirect disclosure and emotional support random intercepts (r = .54, p < .001). Turning to the friend(s) model, there were no significant cross-time parameters, but several correlations were significant. Specifically, direct disclosure and emotional support were positively correlated within W1 (r = .29, p < .026), W2 (r = .19, p < .041), and W3 (r = .20, p < .014). Indirect disclosure was significantly correlated with emotional support within W1 (r = .26, p < .038) and W3 (r = .29, p < .001). Finally, there was a significant correlation between the random intercepts of indirect disclosure and emotional support in the friend(s) model (r = .67, p < .001).

H4

H4 predicted that the expected positive association between disclosure and received emotional support is stronger for direct disclosure compared with indirect disclosure. H4 was not supported. There was only one significant cross-time parameter between disclosure and emotional support, and that was a positive path in the teacher(s) model from direct disclosure W1 to emotional support W2 (B = .26, p < .007). That path, however, was not significantly stronger than the corresponding indirect disclosure W1 to emotional support W2 path (B = .09, p < .108) in the teacher model (z = 1.16, ns). We also compared nonsignificant (in)direct disclosure with emotional support paths and there were no significant differences between those either (based on z scores).

H5

H5, which predicted that received emotional support will be negatively related to depressive symptoms, was not supported, as there were no significant relationships of any kind in the models between emotional support and depressive symptoms.

H6

H6 predicted that indirect disclosure and direct disclosure are indirectly and negatively related to depressive symptoms through received emotional support. We used RMediation (Tofighi & MacKinnon, 2011) to calculate indirect effects and 95% CIs of the predicted W1 (in)direct disclosure → W2 emotional support → W3 depressive symptoms pathways. Mediation effects were not significant in the teacher(s) model (direct disclosure: .01, 95% CI: [–.056, .086]; indirect disclosure: .04, 95% CI: [–.183, .268]) nor the friend(s) model (direct disclosure: –.00, 95% CI: [–.027, .017]; indirect disclosure: .00, 95% CI: [–.015, .018]). Thus, H6 was not supported.

Although the mediation effects were not significant, there were notable within-person associations between disclosure and depressive symptoms in the friend(s) model. Depressive symptoms positively predicted indirect disclosure (B = .24, p < .05), and indirect disclosure positively predicted depressive symptoms over time (B = .17, p < .05). Moreover, within W2 and W3, depressive symptoms were positively associated with direct disclosure (r_W2 = .26, p < .009; r_W3 = .30, p < .001) and indirect disclosure (r_W2 = .29, p < .006; r_W3 = .26, p < .002).

Discussion

Past research devotes little attention to the extent to which immigrant youth indirectly and directly disclose their family-undocumented experiences to people outside the family and whether such disclosures garner support and attenuate depressive symptoms over time. To provide insight into this process, we longitudinally examined Latinx immigrant youth’s indirect and direct disclosures of their family-undocumented experiences to a teacher(s) or a friend(s), and tested how such disclosures directly or indirectly related to emotional support and depressive symptoms. We consider the theoretical and practical contributions of our findings.

Mean Indirect and Direct Disclosures to a Teacher(s) and a Friend(s)

In support of H1, immigrant youth were more likely to indirectly than directly disclose information about their family-undocumented experiences to a teacher(s) or a friend(s) in three waves. This finding is consistent with the notion that people who have stigmatized identities are more likely to disclose indirectly than directly (Williams & Mickelson, 2008). Contrary to H2 and H3, immigrant youth were more likely to directly and indirectly disclose their family-undocumented experiences to a friend(s) than to a teacher(s). Perhaps the sensitive nature of being undocumented and the asymmetrical power distribution between youth and teachers makes immigrant youth less likely to disclose to a teacher(s) than a friend(s), regardless of the type of disclosure. Indeed, people coping with stigmatizing stressors and in need of support often perceive that they lack support in their networks and fear rejection from potential support providers (Williams & Mickelson, 2008). In sum, should immigrant youth choose to disclose information about their family-undocumented experiences, it seems they prefer to employ indirect disclosures. Furthermore, they are more likely to disclose to a friend(s) than a teacher(s).

Associations Between Indirect and Direct Disclosures

In addition to comparing cross-wave levels of indirect and direct disclosure, we found that the more family-undocumented youth directly disclosed to a teacher(s) in the middle of the academic year, the more likely they were to indirectly disclose at the end of the academic year (RQ1). With respect to within-wave associations, direct disclosures were positively related to indirect disclosures for teacher(s) and friend(s). Afifi and Steuber (2009), as well as Catona et al. (2015), posited that people with stigmatizing information or identities often disclose information incrementally (e.g., through hints or partial disclosure), assess the recipient’s reaction, and then decide whether to more fully and directly disclose information. Our results suggest that family-undocumented youth might use this multistep strategy. It is also possible that upon directly disclosing to a teacher(s), the youth might have reverted to indirect disclosures because their teacher did not respond in a satisfying way or because the youth were being cautious.

When interpreting our results, it is important to keep in mind that a conversation about one’s status is unlikely to only involve, for example, the phrase, “My parent is undocumented.” Instead, there are likely to be certain details surrounding the undocumented status and its related experiences that youth might discuss. Thus, youth could have started the conversation with a direct disclosure and then changed to indirect disclosure, depending on the teacher’s reaction, and the same variations in disclosure can exist across conversations about that stressor. As Barbee and Cunningham (1995) argued, support seekers might accept or reject received support and make another attempt to garner support, so that the provider can make another effort to help. They also argued that receiving limited or low quality support is likely to lead to indirect support seeking. If family-undocumented youth’s direct disclosure did not garner the desired response from the teacher, it might have been safer to later use indirect disclosures. It is also plausible that youth could have disclosed to different teachers across the year, and one teacher’s response could shape students’ subsequent disclosures to other teachers.

Associations Between Disclosures and Received Emotional Support

RQ2 asked how indirect and direct disclosures relate to received emotional support over time, and H4 argued that the association would be stronger for direct disclosure than indirect disclosure. H4 was not supported, and RQ2 mostly revealed within-waves cross-sectional associations. The similarity between the teacher(s) and friend(s) model is that for both sources, indirect disclosure was positively related to emotional support reception within most waves. Thus, it is possible that the more emotional support received, the more indirectly Latinx youth disclosed about their family-undocumented experiences. Because they received emotional support in general, they might not have needed to directly disclose their family-undocumented experiences. Alternatively, subtly and/or incrementally disclosing one’s family-undocumented experiences to a teacher(s) or friend(s) might have garnered immediate, but not long-term emotional support. Indirect disclosure did not predict received emotional support at subsequent waves, which means indirect disclosure might not have had a long-term significant effect. Perhaps indirectly disclosing is too subtle or vague to achieve overtime benefits in support.

Where the teacher(s) and friend(s) results differed most was with respect to the longitudinal associations. The friend(s) model revealed no significant longitudinal associations, whereas beginning-of-the-year direct disclosure to a teacher(s) positively predicted middle-of-the-year received emotional support. Yet, beginning-of-the-year received emotional support negatively predicted middle-of-the-year indirect disclosure. Thus, as immigrant youth directly disclosed to a teacher(s) in the beginning of the year, they were more likely to report received emotional support in the middle of the year, perhaps because the teacher was able to clearly discern that the youth needed support upon disclosing. Nevertheless, if immigrant youth reported emotional support reception at the beginning of the year, they were less likely to report indirect disclosure in the middle of the year, perhaps because they already received emotional support.

Implications for Depressive Symptoms

Together, H5 and H6 posited mediation effects such that indirect and direct disclosures would be positively related to emotional support, and in turn, emotional support would be negatively related to depressive symptoms. The mediation was not supported. Emotional support was not significantly related to depressive symptoms. This finding is noteworthy for several reasons. First, in our study, received emotional support did not follow the buffering model of support (Cohen & Willis, 1985) because it did not protect youth from the effects of the stress associated with their undocumented status. Past research, however, tells us that not all support is created equal, and support’s effectiveness depends on a host of conditions. For example, emotional support might not have been a significant mediator because teachers and friends engaged in different types of support (e.g., informational support). Past research suggests that matching people’s desires for support is consequential because unwanted support can be ineffective (High & Steuber, 2014). We studied emotional support because of its applicability and efficacy with a variety of stressors. Yet, receiving emotional support also can be face-threatening. This could be especially true for individuals confronting stigmatizing stressors, such as family-undocumented status. Also, although emotional support might make immigrant youth feel understood and cared for, emotional support cannot provide youth with valuable tangible resources, such as authorization to reside in the United States, financial aid for college, or stable employment, all of which can be related to depressive symptoms. Finally, the nonsignificant mediation might be attributable to the 3-month time intervals (Collins & Graham, 2002), which might have been too long or short to capture effects.

Despite the lack of significant mediation, significant direct associations between indirect disclosure and depressive symptoms warrant attention. The more family-undocumented youth engaged in indirect disclosures with a friend(s) at one time point, the more likely they were to report depressive symptoms at a subsequent time point. Furthermore, the more depressive symptoms immigrant youth reported at one time point, the more likely they were to engage in direct disclosures at a later time point. Similarly, within the middle and the end of the academic year, significant cross-sectional associations were found such that indirect and direct disclosures to a friend(s) were positively correlated with depressive symptoms.

Several possibilities exist that can explain the unexpected finding between both types of disclosure and depressive symptoms. First, it is possible that the more one is depressed, the more likely one is to disclose to a friend(s) to garner assistance. In fact, Barrera (1986) labeled this association the support mobilization model. Second, communication privacy management theory suggests that once people share private information, the recipient co-owns the information (Petronio, 1991). Family-undocumented youth might experience uncertainty, a lack of control, and a sense of vulnerability knowing that a nonfamily member has access to sensitive information that could threaten their family’s safety. Third, upon directly and indirectly disclosing, youth might engage in verbal rumination that heightened their depressive symptoms.

Limitations and Future Research Directions

Although this study sheds light on disclosure patterns among family-undocumented youth, it has several limitations. For example, indirect and direct disclosures were each measured with two items, yet the items referenced a wide variety of possibilities regarding disclosure (e.g., disclosing their family’s undocumented status and sharing details surrounding their family’s undocumented status; disclosing to the same teacher or to different teachers over time). Ideally, the survey would have also included measures that specifically referenced a range of family-undocumented experiences, such as challenges to being undocumented and unique needs of family-undocumented youth. In addition, SIST suggests that there are more ways in which people indirectly garner support through verbal or nonverbal behaviors (e.g., crying, sighing, or appearing agitated). We designed our surveys to work around time restrictions imposed by the school’s class periods. Future research, though, would benefit from using more comprehensive measures of indirect and direct disclosure and youth’s unique stressors. In addition, future research can test the extent to which we would obtain different results if we measured direct or indirect support seeking, rather than just disclosure. Moreover, given that people’s self-reported perceptions of support reception can differ from objective estimates of support reception based, for instance, on coded conversations (Cohen, Lakey, Tiell, & Nelly, 2005), we must keep in mind that alternative measurement approaches to support could yield different results.

This study also would have benefited from considering family-undocumented youth’s goals for disclosing to different sources and the potential cost-benefit analysis they engage in when making disclosure decisions (Afifi & Steuber, 2009). Family-undocumented youth might turn to teachers for certain types of support and friends for other types of support. Moreover, the emotional support items might not have captured the range of support behaviors that friends and teachers engage in. The means for teacher emotional support, for instance, were consistently lower than the means for friend emotional support, which could suggest friends provide more support to students. Yet, it is also possible that the lower scores on the teacher support items are the result of our measure not fully capturing unique types of support teachers offer to students.

Theoretical Implications

SIST is one of the few theories to recognize the processual nature of seeking support, such that support seekers have multiple opportunities to disclose details about their stressors. Barbee and Cunningham (1995) also asserted that a main predictor of the support people receive is how they seek support. Yet, these ideas have not received much empirical testing, particularly using longitudinal data. Although we did not explicitly assess support seeking, we drew from tenets of SIST to inform our theoretical model that connects indirect and direct disclosures to received emotional support. In so doing, we found that greater amounts of direct disclosure to a teacher(s) in the middle of the academic year corresponded with greater amounts of indirect disclosure at the end of the year. Teachers might have responded negatively after the first disclosure, making students reticent to directly disclose a second time. Another possibility is that disclosing is costly, and these costs could be amplified when people are compelled to disclose multiple times for the same problem.

Barbee and Cunningham (1995) asserted that a main predictor of the support people receive is how they seek support. When people are clear about their needs for support, they receive more effective supportive communication. Consistent with SIST, we also found that direct disclosures to a teacher(s) in the beginning of the academic year predicted greater emotional support in the middle of the year. Not only did our study support SIST’s proposition, our findings extended it by noting that the support people receive impacts how they subsequently disclose. People who reported greater emotional support in the beginning of the year engaged in less indirect disclosure at the end of the year. SIST predicts that direct support seeking elicits emotional support, and we found that receiving emotional support suppresses indirect disclosure. People who received emotional support at the outset might have experienced less stress and therefore sought less support later in the academic year. Receiving emotional support might lower the costs and face threat associated with disclosure, which resulted in less need for indirectly disclosing. Thus, we extend the premise of SIST to suggest that the support people receive shapes how they subsequently disclose about the stressor. Future work is needed on the reciprocal relations between disclosure and support to help scholars gain a better understanding of how supportive communication unfolds over time both within and across conversations.

Still, any conclusions related to SIST based on our results should be interpreted cautiously, given our focus on disclosure rather than support seeking. For example, the lack of difference between direct versus indirect disclosure and the mostly nonsignificant longitudinal links between disclosure and received emotional support might be due to our focus on direct and indirect disclosure rather than support seeking. Perhaps, directly disclosing does not convey the same desire or necessity for support than seeking support directly. The lack of longitudinal effects from friends was particularly surprising because SIST contends that how people talk about their stressors and ask for support shapes the support they receive. Our data revealed support for that premise among teachers but not friends. One possibility is that the time between surveys was too long to capture conversational dynamics among friends. The frequent contact and conversations among youth might mean that variations in their disclosure predict the support they receive shortly thereafter but not months later. Researchers can explore the temporal dynamics between disclosing and receiving support for various distinct sources of support.

Although the benefits of directly seeking support have been discussed previously, this is one of the first studies to document the beneficial effects of direct disclosure on receiving support over time. Being clear about the nature of a problem and the accompanying feelings could heighten a provider’s ability to communicate effective comfort by giving them more details about the stressor. Indirect disclosures, conversely, might hamper a provider’s ability by being vague about the source of distress and the corresponding feelings. Indirectly disclosing might also signal a lack of trust in a potential support provider, which might reduce motivation to help. Investigating the interpersonal influences on support providers’ ability and motivation could extend theorizing about the factors that promote effective support provision.

Emotional support is often touted as the most effective type of support and a default type of support to communicate (Xu & Burleson, 2001). In fact, some scholars contend that receiving emotional support is universally beneficial (Reinhardt, Boerner, & Horowitz, 2006). The current study, then, appears to be an outlier. Coping with the stress of the situation surrounding participants’ family-undocumented experiences might be too severe to be ameliorated by emotional support from a teacher(s) or a friend(s). Highly sophisticated emotional support messages, as assessed by their level of verbal person-centeredness often produce better outcomes than less sophisticated messages (High & Dillard, 2012). Thus, future studies should assess the verbal person-centeredness of the emotional support that youth receive.

Practical Implications

To support family-undocumented youth, teachers might need to recognize the indirect ways in which immigrant youth share their family’s situation. To this end, teachers might benefit from receiving culturally-grounded training that informs them of how to recognize signs of stress and trauma, what indirect disclosures might look like, and possible ways to respond to disclosure. Furthermore, it would be beneficial for teachers to know that how they respond to a disclosure can affect immigrant youth’s disclosure or support seeking in the future.

When considering the friend(s) results, this study revealed that family-undocumented youth are more likely to indirectly and directly disclose to a friend(s) rather than a teacher(s); however, indirect and direct disclosures to a friend(s) were positively related to depressive symptoms. Thus, any attempt to support family-undocumented youth might start by partnering with a friend(s) who shares a voluntary relationship built on trust, warmth, and liking because students disclosed more to a friend(s) than a teacher(s) (Derlega et al., 2003). Second, because indirect and direct disclosures to a friend(s) were positively related to depressive symptoms, it is crucial to teach friends or peers how to be effective allies.

Thus far, this study has placed a great deal of responsibility on teachers and friends, but teachers and friends also face systematic barriers that can inhibit their ability to support family-undocumented youth. For example, teachers often have large student-teacher ratios, have school curriculum that they must cover in a short amount of time, and have many students with diverse needs that teachers must accommodate. Moreover, teachers and friends are not trained therapists; they might not know how to address some of the challenges that immigrant youth face. Although beyond the scope of the current investigation, change also likely needs to occur at the school, community, and policy levels because family-undocumented youth face systematic oppression that cannot be addressed only at the individual or dyadic level.

Supplemental Material

Online_Appendix_4.17.19 – Supplemental material for Latinx Immigrant Youth’s Indirect and Direct Disclosures About Their Family-Undocumented Experiences, Received Emotional Support, and Depressive Symptoms

Supplemental material, Online_Appendix_4.17.19 for Latinx Immigrant Youth’s Indirect and Direct Disclosures About Their Family-Undocumented Experiences, Received Emotional Support, and Depressive Symptoms by Jennifer A. Kam, Andy J. Merolla and Andrew C. High in Communication Research

Footnotes

Acknowledgements

The authors would like to thank Maria Larios-Horton and her staff, as well as the authors’ research assistants, Cinthia Chicas, Daisy Figueroa, Hannah Gunter, Lyndsi Ibarra, Celine Jeremiah, Roselia Mendez Murillo, Debora Pérez Torres, and Rachyl Pines. The authors also thank the teachers and students who contributed to this study.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the Institute for Social Behavioral and Economic Research (ISBER) Social Science Research Grant Program and the Academic Senate Doctor Pearl Chase Grant for Local Community Development, Conservation, or Historic Preservation Research Projects at the University of California, Santa Barbara. Both grants were awarded to the first author.

Notes

ORCID iD

Jennifer A. Kam

Supplemental Material

Supplemental material for this article is available online.

Author Biographies

Jennifer A. Kam (PhD, The Pennsylvania State University) is an associate professor in the Department of Communication at the University of California, Santa Barbara. She studies how stressors—stemming from racial/ethnic identification, immigration, and acculturation—relate to children of immigrants’ academic, mental, and physical well-being. She utilizes a stress-resilience-thriving framework to identify psychological, individual action-oriented, and interpersonal communication processes that can attenuate the negative effects of stressors and/or directly promote academic, mental, and physical well-being.

Andy J. Merolla (PhD, The Ohio State University) is an associate professor in the Department of Communication at the University of California, Santa Barbara. His research examines the relationships between interpersonal communication and individual and relational well-being across varying contexts.

Andrew C. High (PhD, The Pennsylvania State University) is an associate professor in the Department of Communication Arts and Sciences at Penn State. His research focuses on supportive communication and computer-mediated communication. He is particularly interested in understanding the dynamics involved in communicating successful attempts at support seeking and provision, especially as those behaviors occur across different channels.

References

Afifi

Steuber

K. R.

(2009). The revelation risk model (RRM): Factors that predict the revelation of secrets and the strategies used to reveal them. Communication Monographs, 76, 144-176. doi:10.1080/03637750902828412

Barbee

A. P.

Cunningham

M. R.

(1995). An experimental approach to social support communications: Interactive coping in close relationships. Communication Yearbook, 18, 381-413. doi:10.1080/23808985.1995.11678921

Barrera

Jr. (1986). Distinctions between social support concepts, measures, and models. Community Psychology, 14, 413-445. doi:10.1007/BF00922627

Biehle

S. N.

Mickelson

K. D.

(2012). Provision and receipt of emotional spousal support: The impact of visibility on well-being. Couple and Family Psychology: Research and Practice, 1, 244-251. doi:10.1037/a0028480

Bodie

G. D.

Jones

S. M.

Vickery

A. J.

Hatcher

Cannava

(2014). Examining the construct validity of enacted support: A multitrait–multimethod analysis of three perspectives for judging immediacy and listening behaviors. Communication Monographs, 81, 495-523. doi:10.1080/03637751.2014.957223

Bodie

G. D.

Vickery

A. J.

Cannava

Jones

S. M.

(2015). The role of “active listening” in informal helping conversations: Impact on perceptions of listener helpfulness, sensitivity, and supportiveness and discloser emotional improvement. Western Journal of Communication, 79, 151-173. doi:10.1080/10570314.2014.943429

Burleson

B. R.

(1984). Age, social-cognitive development, and the use of comforting strategies. Communication Monographs, 51, 140-153. doi: 10.1080/03637758409390190

Catona

Greene

Checton

M. G.

Carpenter

(2016, November). Matters of the heart: An analysis of strategies for disclosing cardiovascular health updates in established romantic relationships. Paper presented at the Annual Meeting of the National Communication Association, Philadelphia, PA.

Catona

Greene

Magsamen-Conrad

(2015). Perceived benefits and drawbacks of disclosure practices: An analysis of people living with HIV/AIDS’ strategies for disclosing HIV status. Journal of Health Communication, 20, 1294-1301. doi:10.1080/10810730.2015.1018640

10.

Cohen

J. L.

Lakey

Tiell

Neeley

L. C.

(2005). Recipient-provider agreement on enacted support, perceived support, and provider personality. Psychological Assessment, 17, 375-378. doi: 10.1037/1040-3590.17.3.375

11.

Cohen

Willis

T. A.

(1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 2310-2357. doi:10.1037/0033-2909.98.2.310

12.

Collins

L. M.

Graham

J. W.

(2002). The effect of the timing and spacing of observations in longitudinal studies of tobacco and other drug use: Temporal design considerations. Drug and Alcohol Dependence, 68, 85-96. doi:10.1016/S0376-8716(02)00217-X

13.

Dabach

D. B.

(2015). “My student was apprehended by immigration”: A civics teacher’s breach of silence in a mixed-citizenship classroom. Harvard Educational Review, 85, 383-412. doi:10.17763/0017-8055.85.3.383

14.

Derlega

V. J.

Winstead

B. A.

Oldfield

E. C.

III Barbee

A. P.

(2003). Close relationships and social support in coping with HIV: A test of sensitive interaction systems theory. AIDS and Behavior, 7, 119-129. doi: 10.1023/A:1023990107075

15.

Derlega

V. J.

Winstead

B. A.

Matthews

Braitman

A. L.

(2008). Why Does Someone Reveal Highly Personal Information? Attributions for and against Self-Disclosure in Close Relationships. Communication Research Reports, 25, 115-130. doi: 10.1080/08824090802021756

16.

Feeney

B. C.

Collins

N. L.

(2015). Thriving through relationships. Current Opinion in Psychology, 1, 22-28. doi:10.1016/j.copsyc.2014.11.001

17.

Goldsmith

D. J.

(1995). The communicative microdynamics of support. Communication Yearbook, 18, 414-433. doi:10.1080/23808985.1995.11678922

18.

Gonzales

R. G.

(2011). Learning to be illegal: Undocumented youth and shifting legal contexts in the transition to adulthood. American Sociological Review, 76, 602-619. doi:10.1177/0003122411411901

19.

Gonzales

R. G.

Suárez-Orozco

Dedios-Sanguineti

M. C.

(2013). No place to belong: Contextualizing concepts of mental health among undocumented immigrant youth in the United States. American Behavioral Scientist, 57, 1173-1198. doi:10.1177/0002764213487349

20.

Hamaker

E. L.

Kuiper

R. M.

Grasman

R. P. P. P

. (2015). A critique of the cross-lagged panel model. Psychological Methods, 20, 102-116. doi:10.1037/a0038889

21.

High

A. C.

Dillard

J. P.

(2012). A review and meta-analysis of person-centered messages and social support outcomes. Communication Studies, 63, 99-118. doi:10.1080/10510974.2011.598208

22.

High

A. C.

Scharp

K. M.

(2015). Examining family communication patterns and social support seeking: Direct and indirect effects through ability and motivation. Human Communication Research, 41, 459-479. doi:10.1111/hcre.12061

23.

High

A. C.

Steuber

K. R.

(2014). An examination of support (in)adequacy: Types, sources, and consequences of social support among infertile women. Communication Research, 81, 157-178. doi: 10.1080/03637751.2013.878868

24.

Jefferies

(2014). Fear of deportation in high school: Implications for breaking the circle of silence surrounding migration status. Journal of Latinos and Education, 13, 278-295. doi:10.1080/15348431.2014.887469

25.

Kam

J. A.

Gasiorek

Pines

Steuber Fazio

K. R.

(2018). Latina/o adolescents’ family-undocumented-status disclosures directed at school counselors: A latent transition analysis. Journal of Counseling Psychology, 65, 267-279. doi:10.1037/cou0000259

26.

Kam

J. A.

Steuber Fazio

K. R.

Mendez Murillo

(2018). Disclosing one’s undocumented status to non-family members: Exploring the perspectives of undocumented youth of Mexican origin. Journal of Social and Personal Relationships. Advance online publication. doi:10.1177/0265407518815980

27.

Knobloch

L. K.

Solomon

D. H.

(2002). Information seeking beyond initial interaction: Negotiating relational uncertainty within close relationships. Human Communication Research, 28, 243-257. doi:10.1093/hcr/28.2.243

28.

Little

T. D.

(2013). Longitudinal structural equation modeling. New York, NY: Guilford Press.

29.

López

Bialik

(2017, May 3). Key findings about U.S. immigrants. Pew Research Center. Retrieved from http://www.pewresearch.org/fact-tank/2017/05/03/key-findings-about-u-s-immigrants/

30.

MacGeorge

E. L.

Feng

Burleson

B. R.

(2011). Supportive communication. In Knapp

M. L.

Daly

J. A.

(Eds.), Handbook of interpersonal communication (4th ed., pp. 317-354). Thousand Oaks, CA: SAGE.

31.

Mortenson

S. T.

(2009). Interpersonal trust and social skill in seeking social support among Chinese and Americans. Communication Research, 36, 32-53. doi:10.1177/0093650208326460

32.

Passel

J. S.

Cohn

(2009, April 14). A portrait of unauthorized immigrants in the United States. Pew Research Center. Retrieved from http://www.pewhispanic.org/2009/04/14/a-portrait-of-unauthorized-immigrants-in-the-united-states/

33.

Passel

J. S.

Cohn

(2017, April 25). As Mexican share declined, U.S. unauthorized immigrant population fell in 2015 below recession level. Pew Research Center. Retrieved from http://www.pewresearch.org/fact-tank/2017/04/25/as-mexican-share-declined-u-s-unauthorized-immigrant-population-fell-in-2015-below-recession-level/

34.

Passel

J. S.

Cohn

Krogstad

J. M.

Gonzalez-Barrera

(2014). As growth stalls, unauthorized immigrant population becomes more settled. Pew Research Center Hispanic Trends. Retrieved from http://www.pewhispanic.org/2014/09/03/as-growth-stalls-unauthorized-immigrant-population-becomes-more-settled/

35.

Petronio

(1991). Communication boundary management: A theoretical model of managing disclosure of private information between marital couples. Communication Theory, 4, 311-335. doi:10.1111/j.1468-2885.1991.tb00023.x

36.

Radloff

L. S.

(1977). The CES-D Scale; A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385-401. doi:10.1177/014662167700100306

37.

Reinhardt

J. P.

Boerner

Horowitz

(2006). Good to have but not to use: Differential impact of perceived and received support on well-being. Journal of Social and Personal Relationships, 23, 117-129. doi:10.1177/0265407506060182

38.

Rogler

L. H.

(1989). The meaning of culturally sensitive research in mental health. The American Journal of Psychiatry, 146, 296-303. doi:10.1176/ajp.146.3.296

39.

Scott

A. M.

Caughlin

J. P.

Donovan-Kicken

Mikucki-Enyart

S. L.

(2013). Do message features influence responses to depression disclosure? A message design logics perspective. Western Journal of Communication, 77, 139-163. doi:10.1080/10570314.2012.694007

40.

Scranton

Afifi

Gangi

(2016). Networks of passing: Experiences of undocumented Latin American immigrants’ identity negotiation. Journal of Intercultural Communication Research, 25, 449-469. doi:10.1080/17475759.2016.1217913

41.

Tofighi

MacKinnon

D. P.

(2011). RMediation: An R package for mediation analysis confidence intervals. Behavioral Research, 43, 692-700. doi:10.3758/s13428-011-0076-x

42.

Williams

S. L.

Mickelson

K. D.

(2008). A paradox of support seeking and rejection among the stigmatized. Personal Relationships, 15, 493-509. doi:10.1111/j.1475-6811.2008.00212.x

43.

Burleson

B. R.

(2001). Effects of gender, culture, and support type on perceptions of spousal social support: An assessment of the “support gap” hypothesis in early marriage. Human Communication Research, 27, 535-566. doi:10.1093/hcr/27.4.535

44.

Youngvorst

L. J.

High

A. C.

(2018). “Anyone free to chat?” Using technological features to elicit quality support online. Communication Monographs, 85, 203-223. doi:10.1080/03637751.2018.1426871

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.32 MB