The Affective Lives of Doubled-Up Latinx Youth: Influences of School Experiences, Familism, and Ethnic Identity Exploration

Abstract

Doubled-up Latinx youth experience many daily challenges associated with ethnic minority status and residential instability. Doubled-up youth share housing with non-custodial caregivers such as friends and/or extended family members primarily because of economic hardship and a breakdown in available parental support. Using data from baseline and 10 days of twice-a-day surveys, this study examined how in-school positive experiences, familism (i.e., a perspective that gives precedence to the family), and ethnic identity (i.e. affirmation, exploration, and resolution) influence after school positive (e.g. feeling joyful/happy) and negative (e.g., feeling stressed/anxious) affect among doubled up Latinx youth (70% female; M_age = 16.5). Results indicate that in-school positive experiences were associated with more after school positive affect and less after school negative affect. Additionally, youth with higher levels of familism reported experiencing less after school negative affect. However, gender moderated the relation of ethnic identity exploration and experiences of after school positive affect. Specifically, females with higher levels of ethnic identity exploration reported relatively lower levels of after school positive affect compared to males. Overall, study findings highlight the importance of both person-level and varying contextual influences on the affective lives of doubled-up Latinx youth.

Approximately 2.5 million youth in the U.S. are homeless according to the National Center for Homeless Education (NCHE, 2016), which highlights a pervasive inequality that merits increased attention and intervention. According to the NCHE, the majority of homeless youth (75%) share housing with a series of friends and/or extended family members because of economic hardship. Such youth are referred to as “doubled-up” and experience environmental instability while relying on others to meet their basic needs. Although doubled-up youth are homeless, they are often not included in researchers’ conceptions of homelessness or research on homelessness, because they do not fit typical stereotypes of homelessness such as living on the street) (see for example, Toolis & Hammack, 2015; Torino & Sisselman-Borgia, 2017). Being doubled-up, like being homeless, also places students at risk for being exposed to myriad environmental threats, as well as being victimized, exploited, and having negative long-term life outcomes (Sulkowski & Michael, 2014). For example, doubled-up homeless youth often have fragmented relations with caregivers, peers, and others (Curry et al., 2017). Consequently, residential instability can inflate risk and erode resilience among youth (Herbers et al., 2019; Sulkowski, 2016).

The risks that come with residential instability might be especially harmful for doubled-up Latinx students (Curry et al., 2017). Such youth represent 24% of the overall homeless youth population in the U.S (Henry et al., 2018), a number likely to increase because the U.S. Latinx population is projected to grow in forthcoming decades (U.S. Department of Commerce, 2018). Most importantly, at the intersection of doubled-up and minority statuses, doubled-up Latinx youth are at risk for experiencing daily emotional challenges (Hallett, 2010). With few exceptions, previous studies have not captured daily affective experiences (i.e., the range of positive and negative feelings) of doubled-up Latinx youth, nor have they investigated the factors that contribute to the levels and variability of affect in this population (Murphy & Tobin, 2011). Understanding doubled-up Latinx homeless youth’s experiences within the context of key intrapersonal (e.g., ethnic identity) and interpersonal (e.g., daily interactions at school) factors may enhance work within school settings aimed at fostering resilience while mitigating risk in populations of doubled-up Latinx youth.

Risk, Resilience, and Daily Affective Experience

The Risk and Resilience framework posits that youth can have healthy developmental trajectories despite experiencing adversity (Masten, 2001). Guided by the Risk and Resilience framework (Masten, 2001), this study investigated how doubled-up Latinx youth’s in-school positive experiences (i.e., feeling supported and socially connected at school), familism values (i.e., attachment and obligation to the family), ethnic identity (i.e., identification with customs and values common to their ethnicity), and gender relate to after school positive and negative affect among doubled-up Latinx youth. Affect refers to personal perceptions of wellbeing and the extent to which individuals experience positive (e.g. feeling joyful/happy) and negative (e.g., feeling stressed/anxious) feelings (Reavis et al., 2015). Past studies on homeless and highly mobile youth guided by the Risk and Resilience framework suggest that such youth display individual differences in resilience across multiple domains of functioning such as academic performance and emotional wellbeing (see Masten et al., [2015] for review). Resilience across multiple domains, however, may occur by maintaining daily positive affect while reducing daily negative affect which fosters doubled-up Latinx youth’s ability to adapt to their daily lives despite experiencing adversity. Therefore, the present study tests whether previously identified risk and resilience factors contribute to the unfolding of daily positive and negative affect among doubled-up Latinx youth to examine whether these mechanisms are relevant to fostering daily resilience among this population.

In addition to the general risks associated with residential instability such as abuse, victimization, and exploitation (Mawhinney-Rhoads & Stahler, 2006), doubled-up Latinx youth also are vulnerable to experiencing racism, discrimination, marginalization, and microaggressions because of their ethnic identity (Keels et al., 2017; Perreira et al., 2010). Their experiences within the school context may be particularly poignant because school may be the most stable environment that doubled-up students experience on a regular basis (Murphy & Tobin, 2011; Sulkowski, 2016). Therefore, in-school positive experiences, such as feeling connected to school and supported by peers, may have substantial influence on the after school positive and negative affect of doubled-up homeless students. In support of this notion, a recent study from the larger project from which the current subsample was drawn, found that receiving social support from teachers and peers differently influenced doubled-up homeless students’ end of day affect, and generally promoted positive affect while reducing negative affect (Griffin et al., 2019). Essentially, daily in-school experiences appear to foster resilience in doubled-up students’ wellbeing through the influence on affect later in the day.

Positive and Negative Affect

Personal perceptions of positive (e.g., happiness, hopefulness, calmness) and negative (e.g., depression, anxiousness, loneliness) affect are important to study for three important reasons. First, more positive and less negative affect is associated with academic functioning, prosocial relationships, and overall psychosocial adjustment (Reavis et al., 2015). Second, daily negative affect is associated with depression and anxiety among adolescents (Torres & Santiago, 2018); and third, positive and negative affect have been identified as critical cross-cultural indicators of affect (Wirtz et al., 2009).

Investigating predictors of affect is particularly relevant for research on resilience in doubled-up Latinx youth. Latinx youth report higher degrees of daily negative affect compared to Caucasian and African American youth (Kann et al., 2016). Further, higher reports of negative affect are associated with poverty and homelessness, which enhance daily stress and can cascade into negative mood states that carry over for subsequent days (Torres & Santiago, 2018). These findings suggest that a “double whammy effect” might exist in the intersecting experiences of being doubled-up and an ethnic minority, which result in doubled-up Latinx youth being at even greater risk for higher negative affect and lower positive affect.

School Experiences and Affect

In-school positive experiences such as receiving social support, acceptance, and respect for one’s ethnicity have shown to promote youth’s positive school outcomes (Perreira et al., 2010). Woolley et al. (2009), for instance, found that positive school experiences buffered against discrimination and psychosocial distress by increasing Latinx youth’s positive affect and simultaneously reducing negative affect. In the aforementioned study, both positive peer behavior and school satisfaction were associated with time spent on homework and academic achievement. Thus, the in-school positive experiences of doubled-up Latinx youth may be an important yet relatively underexplored way to diminish after school negative affect and promote after school positive affect.

Familism and Affect

Stein et al. (2015) state that familism is “a Latino value that promotes loyalty, cohesiveness, and obedience within the family” (p. 1254). Familism has been found to be directly associated and also moderates the influence of peer discrimination, acculturative stress, and economic distress on developmental and psychosocial outcomes of stably housed Latinx youth (Cupito et al., 2015; Hernández & Bámaca-Colbert, 2016; Rivas-Drake et al., 2014; Stein et al., 2015). These findings suggest that familism likely serves a protective role in the lived experiences of doubled-up Latinx youth. Further, a recent study by Torres and Santiago (2018) used a novel daily diary design approach, which can be used to assess daily risk and resilience factors in highly mobile populations, and found that person-level familism assessed at baseline was associated with lower daily negative mood among stably housed Latinx youth. Thus, given the salience of family among Latinxs, familism may index the bond youth have with their families, which could also serve as protective or promotive factor for doubled-up Latinx youth experiencing residential instability and family fragmentation. Limited research has focused on the role of familism among doubled-up Latinx youth and the current study aims to address this gap in the literature.

Ethnic Identity and Affect

Ethnic identity is a key multi-component construct (Umaña-Taylor et al., 2008), including exploration (i.e., exploring membership in an ethnic-racial group), affirmation (i.e., positive feelings toward one’s ethnic-racial group), and resolution (i.e., commitment made to one’s ethnic-racial identity), that can foster resilience or a positive adaptation to life circumstances (Rivas-Drake et al., 2014). Ethnic identity, which intersects with multiple facets of one’s identity, refers to how youth identify with their ethnicity due to the exploration, affirmation, and commitment to parts of their ethnicity that they have engaged in via customs, practices, languages, and values (Umaña-Taylor et al., 2004). Latinx youth’s ethnic identity’s exploration, affirmation, and resolution have been associated with positive self-esteem, higher academic achievement, reduced problematic behaviors, and lower levels of depressive symptoms (Berkel et al., 2010; Rivas-Drake et al., 2014; Umaña-Taylor & Updegraff, 2007). Little is known about the role of ethnic identity on daily affect, but a study with ethnically diverse youth showed that ethnic regard (e.g., individuals’ beliefs that their ethnic group is socially valued) was associated with higher levels of daily happiness and less daily anxiety (Kiang, Yip et al., 2006). Therefore, a link seems to exist between ethnic identity and daily affective experience. Because the role of ethnic identity for doubled-up Latinx youth’s daily positive and negative affect has not yet been empirically examined, the current study aims to fill this gap. More specifically, the influence of ethnic identity exploration, affirmation, and resolution were investigated as they may have a direct effect on after school affect in doubled-up Latinx youth.

Gender as a Moderator

It is critical to consider the role that gender plays in the processes of resilience and adaption for doubled-up Latinx youth as gender is known to shape their daily experiences (Flook, 2011; Weinstein et al., 2006). For instance, homeless females are at greater risk of being victimized and experiencing sexual abuse than males (Tyler et al., 2004). Research focused on Latinx youth has noted that the influence of familism and ethnic identity can be moderated by gender norms and expectations (Blanco et al., 2016; Cheng et al., 2016; Kulis et al., 2012). For instance, some studies find that familism and ethnic identity serve a protective role for females and not for males in promoting academic outcomes and decreasing substance use (Cupito et al., 2014; Kulis et al., 2012). These factors may be more protective for females in part as a result of the gender expectations of staying closer to the family and being the transmitters of culture, whereas males are not held to these same expectations; males are granted more freedom to explore who they are as individuals (Kulis et al., 2012; Raffaelli & Ontai, 2004). Given that gender is a critical factor in the intersecting aspect of doubled-up Latinx youth’s identity, the current study examined whether gender modified the effects of familism and ethnic identity on daily positive and negative affect.

Daily Diary Design, Risk, and Resilience

Previous research on doubled-up youth employs conventional design methods that investigate between-person differences (i.e., how one youth differs from another), yet fails to capture within-person variability (i.e., how a single youth differs from day-to-day; Jahng et al., 2008). Daily diary designs use multiple repeated assessments to capture within-person variability in experiences in near-real time as they occur in their natural environment. This design is uniquely able to investigate within-person variability in experiences, making it well-suited for studying dynamic processes by capturing experiences and behaviors on a timescale that more closely aligns with recently lived experiences.

Daily diary designs are one form of longitudinal data that allows for establishing a temporal order of events (Shiffman et al., 2008). Repeated participant assessments close to when an event occurs improves the ecological validity of responses by reducing recall bias (Bolger et al., 2003). Thus, daily diary assessments allow participants to report on context-bound phenomena (e.g., experiences in schools) that occur in individuals’ recently lived experiences. This design contrasts conventional survey designs (e.g., baseline or yearly assessments) that are better suited for assessing static characteristics or change over a long period of time (e.g., multiple years). In the current study, we implemented a daily diary design where participants completed two daily reports across ten days. The first daily report gathered data on youth’s in-school experiences and was completed right after school. The second daily report assessed youth’s after school social, behavioral, and affective experiences and was completed in the evenings.

Current Study

Consistent with the risk and resilience framework, the current study aims to fill an extant literature gap by investigating influences of in-school positive experiences, familism, and ethnic identity on doubled-up Latinx youth’s after school positive and negative affect. This study also employs a relatively novel design that was particularly calibrated toward understanding the lived experiences of doubled-up Latinx youth who are still attending school. In this regard, the daily diary design had participants provide data twice daily related to their proximal experiences.

The daily diary design allowed for an examination of in-school positive experiences and after school positive and negative affect of doubled-up Latinx youth; a highly mobile population of students that is difficult to assess using conventional methods because of the environmental instability they experience. This design overcomes existing barriers toward understanding risk and resilience and allows for the investigation of the following research questions: 1) How does within-person variation in in-school positive experiences relate to after school positive and negative affect?; 2) How do between-person differences in familism, ethnic identity, and gender relate to after school positive and negative affect?; and 3) Does gender moderate the effect of familism and/or ethnic identity on after school affect? The following are the proposed hypotheses:

• H1: Days with higher ratings of in-school positive experiences will be associated with higher after school positive affect and lower after school negative affect.

• H2: Doubled-up Latinx youth who have stronger familism values will have higher after school positive affect and lower after school negative affect.

• H3: Doubled-up Latinx youth who are higher on ethnic identity affirmation, exploration, and resolution will have higher after school positive affect and lower after school negative affect.

• H4: Familism and ethnic identity (affirmation, exploration, and resolution) associations with after school affect are expected to be moderated by gender, such that familism and ethnic identity increased after school positive affect and decreased after school negative affect for females. No directional hypotheses are made for males due to limited past work.

Method

Participants

Data on doubled-up Latinx youth (n = 64; 70% female; M_age = 16.5, SD_age = 0.76, range = 14 – 20 years) came from a larger project on risk and resilience in homeless youth (N = 98; 66% female; 66% Latinx, M_age = 17.02, SD_age = 1.43, range = 14 – 20 years; see Griffin et al., 2019 for more information about the larger project). Youth were identified as doubled-up (i.e., sharing housing or “couch surfing”) by a non-profit organization that provides resources to homeless students in a U.S. southwest state. The program director of the organization identified and ensured that participants met the following inclusion criteria: 1) were enrolled in one of twelve metropolitan high schools in the area; 2) were classified as “homeless” according to McKinney-Vento Act (MVA; 42 U.S.C. 11431 et seq.) criteria; and 3) lived with a non-biological parent. The sub-sample of self-identified Latinx youth was selected to examine the role of cultural factors on doubled-up Latinx youth’s daily affect. To respect the rights of participants according to existing privacy statutes, information on certain sensitive topics were not collected (e.g., country of origin, generational status, legal citizenship).

Table 1 presents the decomposition of participant responses related to their living situation. Most participants (83.4%) reported living with individuals who were not their parents, 6.3% reported living with a single parent, and none reported living with two parents. Most participants (78.2%) reported having lived in their current placement for at least six months (see Table 1). Participants reasons for being doubled-up varied, but included having parents that passed away, having parents who were nonviable (e.g., mentally ill, drug addicts, in prison, etc.), and being kicked out of the home. Although the living situations of doubled-up youth may make it challenging to identify them as homeless because they may live with relatives, neighbors, or friends, they still meet MVA criteria (Curry et al., 2017).

Table 1.

Descriptive statistics describing Doubled-Up Latinx Youth’s Experiences of Homelessness

Who are you currently living with?
	Frequency	Percent	Cumulative Percent
1. I live with both parents	0	0	0
2. I live with a single parent	4	6.3%	6.3%
3. I live with a relative	30	46.9%	53.1%
4. I live with a foster parent(s)	1	1.6%	54.7%
5. I live with an adult caregiver	9	14.1%	68.8%
6. I do not have a permanent home or caregiver	14	21.9%	90.6%
7. Missing	6	9.4%	100%
Mean = 3.98
SD = 1.40
How long have you lived in your current placement?
	Frequency	Percent	Cumulative Percent
1. 1 day - 1 week	4	6.3%	7.0%
2. 1 week - 1 month	3	4.7%	12.3%
3. 1– 6 months	22	34.4%	50.9%
4. Over 6 months	28	43.8%	100%
5. Missing	7	10.9%	----
Mean = 3.30
SD = .87
What is the longest amount of time that you spent in any home/placement/setting during the past year?
	Frequency	Percent	Cumulative Percent
1. Over 6 months	40	62.5%	66.7%
2. Over 3 months	11	17.2%	85%
3. Over 1 month	5	7.8%	93.3%
4. Less than 1 month	4	6.3%	100%
5. Missing	4	6.3%	----
Mean = 1.55
SD = .91

Participants were enrolled in one of twelve public high schools located in a large city in the U.S. Southwest. The schools the youth attended were spread among various school districts that ranged in demographic backgrounds. For instance, districts ranged between 11% to 40% in percentage of Hispanics attending the schools, 6% to 26.6% for the number of families whose income was below the federal poverty level, and finally 2.1% to 33.3% for families who utilized food stamps. Information specific to the participants in the study reveals that four students (6.3%) reported being in 9^th grade, seven students (10.9%) reported being in 10^th grade, nineteen students (29.7%) reported being in 11^th grade, and more than half of the sample (34 students, 53.1%) reported being in 12^th grade. The sample had an average GPA of 2.49 on a four-point scale. Eight youth reported having a GPA less than 2.0, eighteen youth reported having a GPA between 2.0 and 2.9, twenty-three youth reported having a GPA between 3.0 and 3.5, and six youth reported having a GPA between 3.5 and 3.9.

Procedure

Participants filled out a baseline survey assessing various factors (e.g., familism) prior to contributing daily diary data. After filling out the baseline survey, participants were provided a smartphone that had internet access, text messaging capability, and an application with the daily surveys. Participants received instruction on how to complete daily assessments by a member of the research team; training sessions lasted approximately 30 minutes.

The twice-a-day diary protocol was designed to provide data on behaviors, social experiences, and reactions to experiences in the participants’ natural environments in a fashion that would provide measures of in-school and after school phenomena closely following these experiences. Phones prompted participants to complete the surveys about in school and afternoon experiences with alarms at 3:30 P.M. and at 9:00 P.M. for 10 consecutive days across weekdays and weekends. The chosen after school assessment time avoided interfering with academic activities (e.g., class time, extracurricular activities). The afternoon assessment captured daily experiences while avoiding interfering with students’ nighttime routines (e.g., afternoon school responsibilities, homework completion). For each assessment, if participants did not complete the survey after the first alarm, they were alerted every 15 minutes for the next hour to complete the survey. After four reminders, the request to contribute data was terminated until the next assessment period.

To increase survey compliance, text messages were sent to participants during the 10-day survey period that thanked them for their participation and encouraged them to continue to complete assessments. A trained member of the research team was also available to address any questions participants might have about operating the phones or contributing data during the study. Participants were compensated $10 for filling out the baseline survey and $3.50 per each completed survey ($7.00 for completing both in a day), which resulted in as much as $80 for full study participation. All participants gave informed assent prior to study participation. A university-based IRB approved all study procedures.

Measures

Baseline Survey Measures

Familism.

Two subscales from the Mexican American Cultural Value Scale for Adolescents and Adults (MACVS; Knight et al., 2010) were included in the larger project. Included subscales were: Familism support (six items, α = .81; “Parents should teach their children that the family always comes first”) and familism referents (five items, α = .80; “Children should be taught to always be good because they represent the family”). Response choices range from 1 = not at all to 5 = completely important. Because of high correlations, r = .67, p < .001,the two subscales were combined. The combined subscales achieved an alpha coefficient of .80.

Ethnic Identity.

Youth completed the 17-item Ethnic Identity Scale (Umaña-Taylor et al., 2004), which assesses ethnic identity exploration (EI-E), resolution (EI-R), and affirmation (EI-A) on a four-point Likert scale ranging from “Does not describe me at all” to “Describes me very well.” For ethnic-identity exploration (EI-E; seven items), youth were asked about how often they engaged in behaviors that help them learn about their ethnicity (e.g., “I have participated in activities that have taught me about my ethnicity”; α = .77). The ethnic-identity resolution subscale (EI-R; four items) asked youth about their sense of clarity with their ethnicity (e.g., “I have a clear sense of what my ethnicity means to me”; α = .85). Lastly, ethnic-identity affirmation (EI-A; six items) assesses youth’s acceptance of their ethnic identity (e.g., “I wish I were of a different ethnicity”; α = .73).

Daily Diary Measures

In-School Positive Experiences (ISPE).

The in-school positive experiences measure was adapted from the National Longitudinal Study of Adolescent Health’s attitudes toward school measure (the Add Health study; Harris et al., 2009; see Russell & Toomey, 2013 for detailed measure information). The in-school positive experiences measure included four items to assess youth’s daily school positive experiences. These items included: “Today was a successful day at school”, “Today, I did my best at school,” “Today, I felt connected to school,” and “Today, I felt emotionally supported at school.” Youth responded on a 100-point slider-scale ranging from “Not at all” to “Very.” It is important to note that youth only responded to this measure every day after school (on days they attended school). On days that youth were not at school, this measure was excluded from their daily surveys. Thus, the current study only included data on this measure for days youth attended school and a mean of items was created for each day. Across 10 days of data collection, the internal consistency reached on the positive in-school experiences scale was α = .82. At the daily level, internal consistency ranged between α = .77–.86.

After School Positive Affect.

This scale included five items, adapted for substance and format, that were based on the Positive and Negative Affect Schedule for Children (PANAS-C; Watson & Clark, 1994) to assess youth after school positive affect. Items included: “After school today, did you feel: joyful, inspired, determined, happy, and calm?” Youth responded on a 100-point slider-scale ranging from “Not at all” to “Very.” Youth responded to this measure once daily during the end-of-day survey. The current study only included data on this measure for days youth attended school and a mean of items was created for each day. The internal consistency reached by the positive affect scale across 10 days was α = .85. At the daily level, internal consistency ranged between α = .77–.92.

After School Negative Affect.

This scale included seven items, adapted for substance and format, that were based on the Positive and Negative Affect Schedule for Children (PANAS-C; Watson & Clark, 1994) to assess youth after school negative affect. Items included: “After school today, did you feel: irritable, lonely, stressed, sad, anxious, ashamed, and angry?”. Youth responded on a 100-point slider-scale ranging from “Not at all” to “Very”. Youth responded to this measure once daily during the end-of-day survey. The current study only included data on this measure for days youth attended school and a mean of items was created for each day. The internal consistency reached by the negative affect scale across 10 days was α = .75. At the daily level, internal consistency ranged between α = .70–.88.

Data Analytic Plan

According to a procedure used for analyzing daily diary data, it was determined that participants must have baseline data (n = 64) and a minimum of four days of daily diary data (n = 52; Bolger & Laurenceau, 2013). The current study focused only on school days to understand the association between youth’s in-school positive experiences and after school affect, which are only accurately represented by weekday measurement occasions. This resulted in an analytic sample of 52 individuals and 286 measurement occasions.

Daily predictors were person-mean centered to ensure that within-person level associations are not confounded with between-person variance (Enders & Tofighi, 2007). The within-person level focuses on the variance occurring for one individual across time, whereas the between-person level focuses on variance between participants. Thus, each participant’s in-school positive experiences represented daily deviations from their average level of in-school positive experience across days (i.e., a within-person effect).

Diary designs produce multiple assessments per participant which violate the assumption that assessments are independent and identically distributed. Therefore, multilevel modeling (MLM) was used to accommodate the dependency created by the EMA data structure (Bolger & Laurenceau, 2013; Raudenbush & Bryk, 2002). MLM was conducted using maximum likelihood estimation in the MIXED procedure in SAS software version 9.4.

MLM was used to examine the three research questions. The main effect associations of in-school positive experiences, familism, ethnic-identity affirmation, ethnic-identity exploration, and ethnic-identity resolution with after school positive affect (Model 1) and negative affect (Model 3) were first tested. Whether the associations of familism, ethnic-identity affirmation, ethnic-identity exploration, and ethnic-identity resolution with after school positive (Model 2) and negative (Model 4) affect were moderated by youth’s gender was then examined. The relation between within-person (level 1) and between-person (level 2) associations is demonstrated by the following model equations:

$$\mathit{\text{After}}\mathit{\text{School}}{\mathit{\text{Affect}}}_{it}={\beta }_{0i}+{\beta }_{1i}In-\mathit{\text{School}}\mathit{\text{positive}}{\mathit{\text{Experiences}}}_{it}+{\epsilon }_{it}$$ Level 1:

$$\begin{gathered} {\beta }_{0i}={\gamma }_{00}+{\gamma }_{01}\mathit{\text{Familism}}+{\gamma }_{02}EI~R+{\gamma }_{03}EI~E+{\gamma }_{04}EI~A+{\gamma }_{05}\mathit{\text{Gender}}+{\gamma }_{06}\mathit{\text{Familism}}*\mathit{\text{Gender}}+{\gamma }_{07}EI~E*\mathit{\text{Gender}}+{\gamma }_{08}EI~R*\mathit{\text{Gender}}+{\gamma }_{09}EI~A*\mathit{\text{Gender}}+{\upsilon }_{0i} \\ {\beta }_{1i}={\gamma }_{10}+{\upsilon }_{1i} \end{gathered}$$ Level 2:

In the preceding model equations, i indicated individuals and t indicated time points. The models were tested with random-intercepts and slope (e.g., υ_0i and υ_1i) and residual error terms (e.g., ε_it). The random intercept captures the remaining random variability in the average level of positive or negative affect, whereas the random slope captures the remaining random variability in the association between in-school positive experiences and after school affect. The error structure of the model was left unstructured. The current model was compared to an autoregressive error structure and the unstructured model had a better fit.

Results

Intraclass correlation coefficients (ICCs) were calculated for the daily variables to determine the proportion of the total variance in each construct accounted for by the clustering of between-person variance. ICC estimates indicated that 46% of the variance in in-school positive experiences, 50% of the variance in after school positive affect, and 57% of the variance in after school negative affect was explained by between-person differences. The remaining proportion of variance, about 50%, was explained by within-person variation over time. As expected, positive and negative affect were inversely correlated, r = −.29, p < .01. The relatively modest level of this association indicates that these constructs were independent and should be examined independently. Descriptive statistics, ICCs, and correlations between all measures are presented in Table 2.

Table 2.

Descriptive Statistics and Correlations for Study Variables

	1	2	3	4	5	6	7	Range	M	SD	ICC
Between-Person
1. FAM								1–5	3.22	0.71	-----
2. EI-E	.04							1–4	2.62	0.71	-----
3. EI-R	.03	.66**						1–4	2.83	0.92	-----
4. EI-A	−.20	.01	.18					1–4	3.57	0.68	-----
5. Gender	−.09	−.13	−.20	.26*				0–1	0.70	0.46	-----
Within-Person
6. ISPE	−.04	.11	−.01	−.31	.26			0–100	66.19	24.28	.46
7. ASPA	−.01	.01	.03	−.29	−.08	.51**		0–100	59.27	27.78	.50
8. ASNA	.20	.28	.34*	.17	.08	−.17**	−.29**	0–100	25.90	22.86	.57

Note. FAM = Familism. EI-R = Ethnic identity resolution. EI-E = Ethnic identity exploration. EI-A = Ethnic identity affirmation. ISPE= In-School positive experiences. ASPA= After school positive affect. ASNA = After school negative affect.

^*p≤ 0.05,

^**p≤ 0.01.

In Model 1, in-school positive experiences, b = 0.34, p < .01, were positively associated with after school positive affect, indicating that on days when youth reported more in-school positive experiences than usual, after school positive affect was also higher (see Table 3). Model 2 revealed a significant interaction between gender and ethnic-identity exploration, b = −26.12, p <.01, on after school positive affect. Specifically, females who explored their ethnic identity more experienced relatively lower levels of after school positive affect on average across days, whereas males who explored their ethnic identity more experienced relatively higher levels of after school positive affect on average across days (see Table 3 and Figure 1).

Table 3.

Results for Doubled-Up Latinx Youth’s After School Positive and Negative Affect

	After School Positive Affect		After School Positive Affect		After School Negative Affect		After School Negative Affect
	Model 1		Model 2		Model 3		Model 4
	b	S.E.	b	S.E.	b	S.E.	b	S.E.
Fixed Effects
Intercept	64.96***	4.29	66.28***	4.81	27.32***	4.70	29.39***	5.80
ISPE	0.34***	0.07	0.31**	0.07	−0.16***	0.06	−0.14*	0.06
Familism	6.09	3.28	4.89	7.14	−13.79***	3.37	−16.02*	8.26
EI-R	6.30	3.53	−7.37	8.59	−5.13	3.56	−6.10	10.13
EI-E	−2.65	4.49	17.84*	9.03	8.80*	4.46	2.01	10.66
EI-A	−4.85	3.41	−5.66	4.69	4.08	3.59	2.72	5.60
Gender	−8.33	5.27	−10.02	5.41	3.87	5.57	1.83	6.45
G x Fam			−0.90	7.91			3.17	9.14
G x EI-R			17.21	9.33			0.32	10.88
G x EI-E			−26.12*	10.20			10.64	11.87
G x EI-A			5.52	6.47			2.48	7.66
Random Effects
Intercept	179.46***	52.05	147.65***	46.09	316.44***	76.27	290.63***	71.10
ISPE	0.07	0.05	−1.40	1.14	0.05	0.04	0.05	0.04
ISPE- intercept covariance	−0.60	1.15	0.06	0.05	−3.39*	1.67	−2.61	1.74
Fit Indices
AIC	2570.9		2570.0		2480.7		2486.7
BIC	2592.3		2599.3		2502.2		2516.0

Note. ISPSE = In-school positive experiences. EI-R = Ethnic identity resolution. EI-E = Ethnic identity exploration. EI-A = Ethnic identity affirmation. Fam= Familism. G = Gender.

^*p ≤ 0.05,

^**p ≤ 0.01,

^***p≤ 0.001.

Figure 1. Gender x Ethnic-Identity Exploration Effects on Doubled-up Latinx Homeless Youth After School Positive Affect.

Note. This figure demonstrates how ethnic identity exploration uniquely influences after school positive affect for males versus females.

In Model 3, in-school positive experiences, b = −0.16, p < .01, familism, b = −13.79, p < .01, and ethnic-identity exploration, b = 8.80, p < .05, were associated with after school negative affect (see Table 3). These findings revealed that on days when youth reported more in-school positive experiences than usual, after school negative affect was lower. Further, higher levels of familism values were associated with participants reporting lower after school negative affect, while higher levels of ethnic identity exploration were associated with higher average after school negative affect. Model 4 did not reveal significant interactions with gender.

Discussion

Doubled-up Latinx youth face considerable challenges as a result of residential instability. Despite these challenges, and understood within the Risk and Resilience framework, elements of their daily experiences and cultural background may foster resilience despite the adverse life experiences associated with residential instability. Through a daily diary design, the current study examined dynamic processes by which in-school positive experiences, familism, ethnic identity, and gender were associated with after school positive and negative affect in doubled-up Latinx youth. Overall, study results indicated that daily fluctuations in in-school positive experiences were associated with same-day reports of after school positive and negative affect. Also, youth with higher levels of familism reported less after school negative affect. In addition, gender was found to moderate the relation between ethnic identity exploration and positive affect. In this regard, females with higher levels of ethnic identity exploration reported relatively lower levels of after school positive affect compared to males. These key findings are contextualized and described in greater detail below.

Positive School Experiences

Consistent with expectations, daily in-school positive experiences promoted after school positive affect and diminish afterschool negative affect among doubled-up Latinx youth’s, such that on days when youth reported more in-school positive experiences during their afternoon assessments, after school positive affect was higher and after school negative affect was lower. Explanations for these findings are myriad. It may be that in-school positive experiences attenuate after school negative affect by buffering against the effects of distressing experiences, while also promoting positive affect and enhancing psychological wellbeing (e.g., Perreira et al., 2010; Potochnick et al., 2012). Additionally, in-school positive experiences may promote after school positive affect and reduce after school negative affect in doubled-up Latinx youth because such experiences are associated with school connectedness and wellbeing more generally (Potochnick et al., 2012; Woolley et al., 2009). The aforementioned results are in line with previous studies indicating that positive experiences in school, such as having a sense of belonging, positive academic performance, and emotional support, are associated with more positive affect and less negative affect (Martin & Huebner, 2007). Moreover, results highlighting the impact of daily in-school school experiences on affect are consistent with research on the importance of school connectedness or belonging as an important yet under-explored protective factor for highly mobile students (McNeely et al., 2002; Sulkowski et al., 2012).

Familism

Study results also indicated that stronger familism values were linked with lower levels of after school negative affect. This finding is in line with past empirical findings indicating that familism may buffer against deleterious outcomes such as negative affect and depressive symptoms (Cupito et al., 2015; Torres & Santiago, 2018). Results of the current study suggest that having higher familism values (or a perspective that involves giving precedence to one’s familial context) may promote resilience in doubled-up Latinx youth because of its ability to promote positive affect and lower negative affect at the daily level. This finding matters for school-based practice, especially related to fostering stronger school and community partnerships. A study by Azmitia et al. (2009) explored how familism increased perceived emotional support from parents, siblings, friends, and teachers and how such support predicted academic achievement in Latinx youth. This past research indicated that having a sense of familism values allowed youth to experience support from diverse individuals in their lives. Although, the current study’s population of doubled-up Latinx youth may not have ties with family members they are other individuals such as teachers that are important for fostering support given that school is one of the only stable contexts in their daily lives. Therefore, to facilitate resilience and reduce risk, schools could explore developing teacher-youth connections to capitalize on the influence this relationship can have in the lives of doubled-up Latinx youth.

Ethnic Identity

Surprisingly, and contrary to expectations, youth who reported higher ethnic identity exploration also reported more after school negative affect. Past research generally demonstrates the opposite pattern with stably housed youth, wherein ethnic identity exploration promotes positive outcomes such as healthy self-esteem, self-efficacy, and social competence (Rivas-Drake et al., 2014). However, a study by Torres and Santiago (2018) that utilized a daily diary design found that ethnic identity exploration exacerbated daily negative mood when youth were experiencing high economic stress. Thus, this finding parallels results of the current study in that ethnic identity exploration in low-income ethnic minority youth can be influenced by a range of contextual factors such as environmental instability, uncertainty about the future, and broader stereotypes, stigmas, and microaggressions about being Latinx and having limited financial and social capital (Kidd, 2007; Torino & Sisselman-Borgia, 2017; Torres et al., 2011).

Therefore, doubled-up Latinx youth who report higher levels of ethnic identity exploration may be more susceptible to experiencing enduring negative affective states, especially if they encounter negative perceptions or experiences related to their ethnicity. A study by Torres and Ong (2010) that also employed a daily diary design found that experiencing discrimination increased feelings of depression for more than a day following the event in Latino youth who were in the exploration phase of ethnic identity development. Collectively, this finding highlights the importance of exploring the intersection between Latinx youth’s experiences connected to their identities (e.g., homelessness, ethnic, undocumented status, etc.) and how the intersection of identities can influence whether cultural variables increase risk or promote resilience. Moreover, since doubled-up Latinx youth are limited in their ability to change environmental circumstances in their lives or systematic societal problems like racism and discrimination, the onus is on school personnel to ensure that these students feel safe, supported, respected, and included at school (Sulkowski, 2016; Sulkowski & Michael, 2014). An emerging consensus of research indicates that a positive school climate is associated with better student outcomes regardless of students’ ethnic background or mobility status (La Salle et al., 2017; Moore et al., 2018; Thapa et al., 2013).

It is important to note that, contrary to study hypotheses, no relation was found between ethnic identity resolution or affirmation and after school affect (both positive and negative) in the present study. Perhaps doubled-up Latinx youth with a clearer sense of what their ethnic identity means to them (identity resolution) or fixed perceptions about their ethnicity (identity affirmation) might not experience as much cognitive dissonance associated with exploring their ethnic identity. In general, cognitive dissonance is associated with negative affective states related to holding two or more conflicting beliefs, ideas, or values (Festinger, 1962). Therefore, although in need of research, doubled-up Latinx youth who are more settled in their ethnic identity beliefs—whether positive or negative—may experience less daily dissonance associated with their ethnic identity, which might then neutralize daily affect associated with such beliefs.

Ethnic Identity and Gender

As a key finding, study results revealed that gender moderated the association between ethnic identity exploration and after school positive affect. Specifically, higher levels of ethnic identity exploration were associated with decreased after school positive affect for doubled-up Latinx females, whereas higher levels of identity exploration were associated with increased after school positive affect for males. Although speculative, these differential effects may be related to how residential instability uniquely impacts female doubled-up Latinx youth’s exploration of their ethnic identity, an identity that traditionally is tied to females establishing and maintaining close family ties (Cupito et al., 2015; Raffaelli & Ontai, 2004). Therefore, breakdowns in traditional family structures experienced by being doubled-up may be more emotionally poignant for Latinx females compared to their male peers. Results from a study by Peña et al. (2011) suggest that being in a tight-knit family (i.e., high cohesion and low conflict) versus an intermediate-knit or loose-knit family protects against suicide behavior in Latina youth. Furthermore, in the same study, familism was found to increase the likelihood of being in a tight‐knit family—a protective factor. Therefore, similar to as in extreme cases, daily affective experiences of female doubled-up Latinx youth also may be influenced by how ethnic identity occurs within the context of one’s family constellation.

Doubled-up youth who experience breakdowns in family ties may benefit from receiving additional supports such as being in group counseling or working individually with a therapist. In this regard, Villalba (2011) recommends a strength‐based approach for counseling Latinx youth that focused on Latinx cultural strengths, including interpersonal bonds, bilingual abilities, and bicultural skills. Sapp (2006) also advocates for such a model, which involves acknowledging the reality of ethnic, gender, and social class discrimination in at-risk youth populations and systematically building resilience by considering personal and cultural strengths. A culturally responsive strength-based model may help Latinx youth who regularly experience negative affect or diminished positive affect while they are exploring their ethnic identity.

In contrast to their female counterparts, Latinx males are traditionally provided with more freedom and are encouraged to pursue opportunities outside of the family context that promote greater degrees of individuality and personal autonomy (Raffaelli & Ontai, 2004). Therefore, doubled-up male Latinx youth with higher levels of ethnic identity exploration appear to have comparably more daily after school positive affect. Additionally, because traditional Latino cultural norms encourage male freedom, male doubled-up Latinx youth may feel more comfortable exploring aspects of ethnic identity. Providing partial support for the former, a study by Umaña-Taylor et al. (2012) found that ethnic identity exploration was a protective factor in Mexican-origin male youth that buffered against the negative effects of discrimination. Of course, relations between ethnic identity and gender are complex, speculative, and in need of further research, especially in highly mobile populations of youth who might not be as heavily influenced by traditional family norms compared to their stably housed peers.

Limitations

It is important to interpret study findings within the context of some limitations. First, the sample comprised doubled-up youth seeking services from a community organization in a southwest city in the U.S. Thus, findings may not generalize to dissimilar populations of youth (e.g., doubled-up youth who have stopped attending school and are not linked to a community program). Second, given the focus on doubled-up Latinx youth, it is important to consider the role of acculturation in the associations tested in the current study. Youth’s level of acculturation to U.S. culture may play a role into how their cultural orientations can serve as risk and resilience factors. Third, daily assessments could potentially influence participants’ perceptions and reporting of daily experiences or behaviors, a phenomenon known as the reactivity bias (Bolger & Laurenceau, 2013). Future research should investigate doubled-up youth’s reactivity to daily assessment procedures to understand the strengths and weaknesses of this design with highly mobile populations. Fourth, gender was treated as a dichotomous variable, which does not acknowledge the fluid nature of gender. An important area for future research involves studying the intersection of diverse aspects of gender, ethnicity, and mobility experiences in youth.

Conclusion

A significant number of Latinx youth experience the deleterious effects of homelessness and high mobility (Henry et al., 2018). However, not all youth are influenced by these effects in the same way and to the same degree. Consistent with a Risk and Resilience framework, this study examined how in-school positive experiences, familism (i.e., a perspective that gives precedence to the family), and ethnic identity were linked to after school positive and negative affect among doubled up Latinx youth. Overall, results suggest that daily in-school positive experiences are associated with increases in after school positive affect and decreases in after school negative affect. Therefore, schools should actively work toward ensuring that school climates are safe, supportive, welcoming, and culturally responsive.

In addition, doubled-up Latinx youth with higher levels of familism reported less after school negative affect, which highlights the importance of fostering healthy connections between teachers and doubled-up youth. Lastly, gender moderated the relationship between ethnic identity exploration and after school positive affect in that females with higher levels of ethnic identity exploration reported relatively lower levels of positive affect compared to males. This finding indicates that it is important to ensure that such students are provided with appropriate supports to minimize risk and foster resilience.