Validation of the Revised BSI-18 With Latino Migrant Day Laborers

Abstract

Objective:

This study examines the factor structure of the Brief Symptom Inventory-18 (BSI-18; Derogatis, 2001) in a hard-to-reach population of mainly undocumented migrant Latino day laborers. While, the BSI-18 has been found to be a valid and reliable measure of psychological distress, cross-cultural assessments in sub-groups of Latinos are scarce withthis being the first study to examine the factor structure of the Spanish version of the BSI-18 with this population.

Methods:

A series of principal axis exploratory factor analytic procedures examining one to three-factor models were performed with a street recruited sample of 150 Latino migrant day laborers.

Results:

A one-factor model emerged and four items were dropped due to low factor loadings. The internal consistency coefficients for the 14-item BSI was α = .87.

Conclusions:

Findings suggest that the revised BSI-18 appears to measure a single dimension of general somatic-psychological distress in this sample of Latinos.

Keywords

factor analysis psychometric study Hispanics population men mental health

Despite the estimated 11.2 million unauthorized immigrants that live in the United States (Passel & Cohn, 2011), there is an absence of a well-defined understanding of this population’s mental health status. Recent studies suggest that Latino migrant day laborers, many of whom are undocumented, are particularly at heightened risk for psychological distress due to their distinct social, situational and contextual circumstances (Negi, 2013; Organista & Kubo, 2005; Quesada, 1999). Day laborer populations are largely comprised of single, Latino, undocumented migrant men primarily working in construction and demolition jobs (Valenzuela, 2003; Valdez, Cepeda, Negi, & Kaplan, 2010). Latino day laborers work in the informal economy in often unregulated conditions with widespread workers’ rights abuses and victimization (Walter, Bourgois, Loinaz, & Schillinger, 2002). Day labor work is also highly inconsistent as it is often negotiated in verbal contracts at informal day labor sites (or public street corners) and, as a result, day laborers often earn less than the poverty level (Valenzuela, Melendez, & Gonzalez, 2006). Confounding their experiences, this population tends to settle in some of the most disadvantaged and impoverished neighborhoods and experience high levels of victimization (Cepeda et al., 2012).

Although the Latino immigrant health paradox suggests better health outcomes for first-generation immigrants (Warner et al., 2006), day laborers employed in the informal economy may not fit this model. Specifically, this distinct subpopulation of Latino migrant day laborers appears to be at conferred vulnerability for psychological distress as a result of the often stressful contexts within which they work and live in (de Castro, Voss, Ruppin, Dominguez, & Seixas, 2010; Negi, 2011). With Latinos having some of the lowest rates of access to and utilization of mental health services (Cabassa, Zayas, & Hansen, 2006), and those with limited English proficiency having even more significant disparities in mental health care (Sentell, Shumway, & Snowden, 2007), research on the social epidemiology of mental health problems are needed to minimize emerging health disparities among Latino migrant day laborers. Subsequently, valid and reliable assessment instruments are needed to begin to more accurately understand the mental health needs of Latino migrant day laborers living in the United States.

Brief Symptom Inventory-18 (BSI-18) With Spanish-Speaking Populations

While there is a growing number of valid and reliable psychological distress measures, few have been normed with Latino populations and even fewer on subpopulations such as Latino male migrant day laborers. Psychological distress measures developed and normed with European Americans or other minority populations cannot be assumed to be valid and reliable for Latinos, given cultural and linguistic differences. To address this need, this is the first study we are aware of that examines the factor structure of the Spanish version of the BSI-18 (Derogatis, 2001) with a community sample of Latino male migrant day laborers.

The BSI-18 is a widely used brief measure of psychological distress and has been found to consist of three factors: (1) somatization or distress due to perceived bodily ailments; (2) depression which includes depressed mood, self-deprecation; and (3) anxiety including nervousness and tension (Derogatis, 2001). This is consistent with Lazarus and Folkman’s (1984) theory of stress and coping which defines psychological distress as an adaptive outcome as a result of exposure to stress. The factor structure of the BSI-18 was examined by Derogatis’ (2001) with a community sample of 1,134 adults using principal component analysis and found four factors (somatization, depression, anxiety, and panic). However, Derogatis indicated that panic related disorder marginally made the retention criteria for a separate factor and that it would likely load with anxiety items in future analysis. Similarly, Zabora and colleagues’ (2001) study of 1,543 cancer patients used principal component factor analysis and found three factors (somatization, depression, and anxiety), as well as a fourth factor consisting of 1 item that generally loads on depression, “suicidal ideation.” However, this 1-item factor did not meet the Eigenvalue criteria but was included as it passed the scree test. As the race, ethnicity, and socioeconomic status of these normative samples were not provided, it is difficult to ascertain whether the underlining factor structure of the BSI-18 operates similarly or differently for Latinos more broadly or Latino migrant day laborers more specifically.

There have been a handful of studies that have begun to explore the psychometric characteristics and the posited factors structure of the BSI-18 with Latinos in general. Prelow, Weaves, Swenson, and Bowman’s (2005) study examined the construct validity and reliability of the BSI-18 with 1,115 low-income Latina women using confirmatory factor analysis (CFA). They found that the factor correlations between anxiety and depression, anxiety and somatization, and depression and somatization were very large with questionable discriminant validity. A one-factor model was recommended, as the BSI-18 was found to measure general psychological distress among their sample of Latinas. Asner-Self, Schreiber, and Marotta (2006) explored the factor structure of the BSI-18 with 100 Central American immigrants and found that women had higher scores for somaticism than the male participants as well as the norm group, and the Central American men were a little above one standard deviation from the U.S. American male norm. They also concluded that the BSI-18 may be measuring one underlying factor—general psychological distress. Similarly, Wiesner et al. (2010) conducted a CFA with mean structures with Latinas as well as White and African American women. They found that the hypothesized three-factor structure of the BSI-18 held for both White and African American women. However, significant redundancy existed among several of the factors with inadmissible parameter estimates for the Latina sample. Finally, Torres et al.’s (2013) study examined the factorial invariance of the BSI-18 with a sample of 923 adults of Mexican descent recruited from university and community settings across nativity status, language format, and gender. Although they found that the three-factor structure exhibited better model fit for the total sample and other groups excluding the foreign-born sample, the large factor relationships rendered it untenable and a one-factor structure was assessed to be the best fit for all samples.

Although the studies reviewed suggest that the BSI-18 appears to measure a single dimension of general psychological distress, Latinos are not a homogenous group. Literature indicates that male Latino migrant day laborers in particular may be at higher risk of experiencing psychological distress due to their difficult work and life conditions (Negi, 2013; Organista & Kubo, 2005). Despite being a potentially vulnerable group experiencing a significant amount of psychological distress, measures including the BSI-18 might not fully capture the distinct distress that this population experiences. In particular, the undocumented immigrant status of many Latino migrant day laborers may heighten stress on this population as it places them within a stigmatized category that is often distrusting of mainstream institutions (including police and other authorities) thereby increasing vulnerability and risk for threatening or demanding events such as discrimination and victimization. Latino migrant day laborers’ fear of being exposed as undocumented and subsequently being deported may further serve as a chronic source of stress with deleterious consequences on this population’s health (Arbona et al., 2010). It is then possible that the amalgamation of sociocultural factors (such as poverty, undocumented immigration status, acculturation, and culture) may influence differential response patterns to the specific items on the BSI-18. This is especially compelling, as the scale characteristics of the BSI-18 may be different when used with a population with a lower distress level from those in a population with a higher distress level (Meijer, deVries, & van Bruggen, 2011).

Given Latino migrant day laborers’ distinct experiences, it makes theoretical sense that the factor structure and item loadings may differ for this distinct group (Guillemin, Bombardier, & Beaton, 1993). Consequently, further examination of the factor structure of the BSI-18 has been recommended in order to determine its utility as a research and screening tool (Asner-Self, Schreiber, & Marotta, 2006). The purpose of this current study is to investigate the validity and reliability of the Spanish BSI-18 for use with a high-risk subgroup of Latinos, Latino migrant day laborers. This study’s focus on men is additive as most studies on the factor structure of the BSI-18 with Latino immigrants have been conducted with Latinas (Prelow, Weaves, Swenson, & Bowman’s 2005; Wiesner et al., 2010), Latino college students (Torres et al., 2013), or a small sample of Latino immigrant men (n = 47; Asner-Self et al., 2006). Findings are then likely to further our limited understanding of the psychometric properties of the Spanish BSI-18 as well as help refine this measure through this study’s focus on exclusively Spanish-speaking, low-income Latino migrant men with low levels of acculturation and educational attainment.

Method

Participants

A total of 150 male Latino migrant day laborers participated in the survey, of which 68% were born in Mexico, 31% in Central America, and 1% in the United States. As shown in Table 1, participants ranged in age from 18 to 68 years, generally reported lower levels of acculturation (Marin, Sabogal, Marin, Otero-Sabogal, & Perez-Stable, 1987) and less than a high school education. The majority were married (59%), with their wives living in their country of origin.

Table 1.

Background Characteristics of Latino Day Laborers.

Characteristics (N = 150)	% M	SD
Place of birth
Mexico	68%
Central America	31%
United States	<1%
Age (M)	37 years	10.68
Acculturation^a	4.99	1.65
Years education
Middle school	35%
Some high school	22%
Some grade school	16%
High school	10%
Advanced degree	3%
No formal education	7%
Marital status
Married	59%
Single	41%
Spouse living abroad	89%
Spouse living in the United States	11%
Income (weekly)
<$100	9%
$101–200	16%
$201–300	35%
$301–400	33%
$501–1,000	8%
Income sent home (M)	43.46%	23.05
Total years in the United States (M)	7 years	8 years
Workers’ rights abuses
Wage theft	63%
Physical or verbal threats	26%

^a 1 = Highly acculturated, 3 = Bicultural, 5 = Low levels of acculturation.

Design

This study is part of a larger mixed-methods study with results presented elsewhere (Negi, 2011; Negi, 2013). The larger study was comprised of three phases including Phase I which used ethnographic methods to elucidate participant identified protective factors, phase II which implemented a quantitative survey, and Phase III which involved qualitative member checking. This study utilizes secondary data collected during the administration of the quantitative survey (Phase II), which included the Spanish BSI-18 among other background demographic questions and two other brief quantitative scales, including those measuring discrimination and social isolation.

As day laborers are a highly transient population, creative methodological approaches were used to recruit participants (Valenzuela, 2000). To build entrée with this hard-to-reach population, a local organization that works with day laborers was identified to serve as a gatekeeper (Sifaneck & Neaigus, 2001). This local organization provided essential information regarding the location and composition of day labor sites in the large Southwest city where this study was conducted. To avoid any conflict of interest, any direct involvement with the organization was discontinued while conducting research.

Based on gatekeeper input as well as field observations, the three largest and most active day labor sites in the Southwest city were selected for purposive sampling. Four research assistants and the principal investigator (PI) administered all phases of the research at the identified day labor corners. Field visits were conducted at different times between 7:00 a.m. and 3:00 p.m. to facilitate inclusion of the maximum amount of workers moving in and out of the street corner seeking employment. All interviews were administered in Spanish. As some Latino migrant day laborers may have limited reading and writing abilities and may be unfamiliar with completing survey instruments, the surveys were conducted face-to-face and in the field at day labor corner sites. Face-to-face interviews were chosen, as it allowed the interviewer an opportunity to establish rapport and gain the cooperation and trust of the person being interviewed. This was especially essential when targeting a mainly undocumented population for inclusion into the study. The survey lasted between 15 and 20 minutes. All researchers, including the PI, are fluent in Spanish.

Inclusion Criteria

To be eligible to participate in the study, participants had to be Latino, male, aged 18 or older, and currently working as a day laborer. Participants were informed that data collection could be terminated immediately if an employment opportunity presented itself. All survey participants, including those who terminated the interview because of work-related reasons, were provided with a $10 gift certificate to a grocery store.

Measures

Demographics

Specific demographic items included age, country of birth, number of times immigrated to the United States, years in the United States, education, weekly income, remittances sent, number in household, and marital status. A majority of the demographic questions were taken from a Spanish language survey developed specifically to study Latino day laborers (Organista & Kubo, 2005).

Psychological Distress

The BSI-18 (Derogatis, 2001) was used to assess psychological distress. The BSI-18 is an 18-item measure that assesses level of distress over the previous 7 days on a 5-point Likert-type scale (0 = not at all to 4 = extremely). Each dimension (somatization, depression, and anxiety) contains 6 items. The items that correspond to each dimension are noted in Table 2. The sum of the scores of all three dimensions is the global severity index (GSI) of distress and ranges from 0 to 72 with higher sums indicating higher levels of psychological distress. The internal consistency reliability coefficient of the BSI-18 Spanish version for this current study was α = .87.

Table 2.

Factor Loadings for the Spanish Brief Symptom Inventory-18 Items.

			One-Factor Model	Two-Factor Model		Three-Factor Model
Item	M	SD	F₁	F₁	F₂	F₁	F₂	F₃
1. Faintness (S)	0.28	0.69	.54	.55	.01	.50	.03	–.08
2. No interest (D)	0.41	0.79	.27	.10	.22	.02	.15	–.18
3. Nervous (A)	0.79	1.02	.65	.88	–.21	.78	–0.18	–.11
4. Heart or chest ache (S)	0.31	0.70	.28	.20	.10	.04	–0.02	.10
5. Lonely (D)	1.34	1.20	.61	.42*	.25	.03	–0.04	–.82
6. Stress (A)	1.18	1.28	.56	.25	.40	.10	.27	–.35
7. Nausea or stomach ache (S)	0.44	0.87	.34	.11	.30	.15	.33	.02
8. Depressed (D)	1.07	1.12	.71	.53	.24	.25	.03	–.61
9. Scared all of a sudden (A)	0.54	1.02	.62	.54	.13	.46	.12	–.16
10. Work to breathe (S)	0.24	0.64	.45	.28	.22	.16	.13	.26
11. Useless (D)	0.25	0.63	.31	.04	.34	.13	.41	.11
12. Fear or panic (A)	0.44	0.87	.62	.79	–.15	.71	–.12	–.09
13. Pins and needles (S)	0.28	0.64	.49	.43	.10	.39	.10	.08
14. Without hope for future (D)	0.64	1.00	.44	.00	.55	.02	.54	–.06
15. Restless (A)	0.61	0.94	.64	.57	.12	.57	.17	.00
16. Weakness in parts of body (S)	0.57	0.91	.62	.51	.16	.61	.26	.13
17. Suicidal (D)	0.20	0.65	.45	–.14	.75	–.17	.64	–.22
18. Fearful (A)	0.43	0.77	.71	.34	.48*	.20	.36	–.36

Note. Boldface values are factor loadings .40 or greater. The letters in parentheses correspond to each dimension of the BSI; D = Depression, S = Somatic, and A = Anxiety. The two cross-loadings with .20 difference in the two-factor model is noted by the asterisk (*).

Data Analytic Plan

Prior to conducting the analysis, two common factor extraction methods including principal components analysis (PCA) and common factor analysis, or principal axis factoring (PAF), were considered. The goal of PCA is to reduce the number of items in the measure while retaining as much of the original item variance as possible (Worthington & Whittaker, 2006), while the purpose of PAF is to identify the underlying dimensionality of the item set or latent factors that account for shared variance among items. Accordingly, following best practice guidelines, PAF was selected since we were interested in identifying the latent factor/factors and evaluating the construct validity of the BSI-18 Spanish version with Latino male day laborers (Fabrigar, Wegener, MacCallum, & Strahan, 1999; Miller, Kim, Chen, & Alvarez, 2012; Worthington & Whittaker, 2006). Although there is much debate over the preferred method (Gorsuch, 2003) and the fact that PCA and PAF often produce similar findings (Velicer & Jackson, 1992), PAF has been found to be the superior method under several conditions, and the results more optimally generalize to CFA results and more stable factor loadings (Floyd & Widaman, 1995). The direct oblimin rotation was selected, given that the factors were expected to correlate with one another (Gorsuch, 1983).

The primary factor structure extraction decisions used were based on the minimum average partial test (MAP) and parallel analysis (O’Connor, 2000). Although the Cattell (1966) and Kaiser (1958) Eigenvalues greater than 1.0 are widely used approaches to determine how many factors to retain, parallel analysis (Horn, 1965) and the MAP test are well-validated procedures that are widely recommend by statisticians (O’Connor, 2000). That is, Eigenvalues-greater-than one rule tends to overestimate and sometimes underestimate the number of factors (Zwick & Velicer, 1986). Accordingly, parallel analysis is the recommended empirical procedure for deciding how many factors to retain (O’ Connor, 2000; Worthington & Whittaker, 2006; Zwick & Velicer, 1986). Using the parallel analysis procedure, the researchers randomly ordered the participants’ item scores and conducted a factor analysis on both the original data and the randomly ordered scores (O’Connor, 2000). The user of the parallel analysis program specifies how many random data sets are generated; the larger number of data sets specified provides more reliable and precise estimates; thus, we specified in the program to generate 1,000 random data sets in order to compare the random data Eigenvalues with the actual Eigenvalues. Accordingly, the scree plot or the Eigenvalues of each factor is plotted in descending order. The researcher then located the break and plateau in Eigenvalues as an additional visual supplemental method to determine the factor structure. The criteria used to retain or delete an item was based on factor loadings of less than .40 or cross-loadings of less than .20 difference from the item’s highest factor loading (Tabachnick & Fidell, 2001).

Results

Descriptive Statistics

Table 2 lists each item on the BSI-18 and the average mean with their standard deviations. The mean GSI score on the BSI-18 was 10.49 (SD = 9.6; range 0–46), while the scores on each of the three dimensions were 3.87 (SD = 3.36, range 0–16) for the depression dimension, 2.1 (SD = 3.36, range 0–16) for the somatic dimension, and 4.5 (SD = 5.01, range 0–24) for the anxiety dimension.

A series of principal axis factor analyses using a direct oblimin rotation examining one- to three-factor models were performed (Table 2). The Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy value was .85 and the Bartlett’s Test of Sphericity was significant, χ = 838.26 (df = 15.), p < .001, which indicated that factorability of the data was acceptable. The participant-to-item ratio was 8.3:1 suggesting that the current sample size was adequate to perform a PAF and likely to yield a stable factor solution (Gorsuch, 1983; Miller et al., 2012). Items with factor loadings lower than .40 and items that cross-loaded, or were highly correlated with one another, were removed. The results revealed that three factors had Eigenvalues greater than 1.0; however, this solution was rejected by the parallel analysis. Based on the parallel analysis with 1,000 randomly generated data sets (Figure 1), the findings suggested that only the first factor explained a greater percentage of variation in the items than could be expected on the basis of chance (Table 2). Based on the aforementioned criteria, a one-factor model emerged that accounted for 32% of the variance in the BSI-18 items. All the factor loadings greater than .40 are in boldface in Table 2. The single factor appeared to reflect general somatic-psychological distress and the three highest loadings included fearful (0.71), depressed (0.71), and nervousness (0.65). Accordingly, 4 items were dropped due to low factor loadings resulting in the 14-item BSI, which we term the revised BSI-14 Spanish version. Of the 4 items that failed to load adequately, the 2 dropped items, “faintness or dizziness” and “pains in heart or chest”, fell into the original Somatization factor, while “feeling worthless” and “feeling no interest in things” are items that were in the original Depressed factor. We also compared and provided the factor loadings derived from Prelow et al.’s (2005) study, which are provided in Table 3.

Figure 1.

Parallel analysis with 1,000 randomly generated data sets.

Table 3.

Factor Loadings for the Revised Brief Symptom Inventory-14 Spanish ITEMS compared to Prelow et al.’s (2005) Study.

	Current Study	Prelow et al.
Item	F₁	F₁
1. Faintness (S)	.54	.46
2. No interest (D)	.27	.63
3. Nervous (A)	.65	.69
4. Heart or chest ache (S)	.28	.51
5. Lonely (D)	.61	.71
6. Stress (A)	.56	.69
7. Nausea or stomach ache (S)	.34	.62
8. Depressed (D)	.71	.73
9. Scared all of a sudden (A)	.62	.72
10. Work to breathe (S)	.45	.66
11. Useless (D)	.31	.76
12. Fear or panic (A)	.62	.62
13. Pins and needles (S)	.49	.56
14. Without hope for future (D)	.44	.73
15. Restless (A)	.64	.77
16. Weakness in parts of body (S)	.62	.72
17. Suicidal (D)	.45	.62
18. Fearful (A)	.71	.75

Note. Boldface values are factor loadings .40 or greater. The letters in parentheses correspond to each dimension of the BSI; D = Depression, S = Somatic, and A = Anxiety.

Reliability and Convergent Validity Evidence

The internal consistency coefficients for the revised BSI-14 Spanish version with the current sample of Latino migrant day laborers was α = .87, which is the same reliability estimate as the original Spanish BSI-18. Finally, the convergent validity of the BSI-14 with the BSI-18 was examined. Correlation between the one-factor BSI-14 and BSI-18 was r = .985, suggesting a strong correlation between the two versions.

Discussion

To our knowledge, this is one of the first studies to examine the factor structure of a psychological distress measure with Latino migrant day laborers. It is crucial to assess the factor structure and refine the psychometric properties of the revised BSI-14 Spanish version for use with Latino migrant day laborers to develop more accurate measures of assessing psychological distress among this at-risk population, as well as with other subgroups of Latinos. In contrast to the original theorized BSI-18 3 factor structures, the revised BSI-14 Spanish version that was developed in the current study appears to measure a single underlining dimension of general somatic-psychological distress in this sample of male Latino day laborers. Moreover, the internal consistent estimate obtained in the revised BSI-14 was the same compared to the full version (α = .87). These findings are generally consistent with previous studies examining the BSI-18’s factor structure with Latinos in general (Asner-Self et al., 2006; Prelow et al., 2005), however, with some notable differences.

While Asner-Self et al. (2006) and Prelow et al.’s (2005) results also yielded a one-factor model, there were 4 items that did not load (greater than .40) on the general psychological distress factor in the current study. This is in contrast to previous studies in which all of the items generally loaded well on one dimension. Specifically, in Prelow and colleagues’ study, all but one item had a factor loading greater than .50. Related, this explains why only 32% of the variance in the revised BSI-14 items were explained in our study compared to 45% and 50.9% explained in Asner-Self et al. (2006) and Prelow et al.’s (2005) studies, respectively. There are a number of possible reasons that may explicate the discrepancies between our current findings and previous studies. One possible explanation could be that Latino migrant day laborers’ unique characteristics lead them to experience distress differently compared to other Latino groups. As most of the previous studies of the Spanish BSI-18 have been conducted on women, it is likely that some of our results can be explained by gender differences. Specifically, Latino migrant day laborers may endorse traditional masculine norms that could impact and differentiate their experience of distress from Latina women (Iwamoto, Liao, & Liu, 2010). Latino migrant day laborers have also been qualitatively found to indicate that their primary purpose to be in the United States is to work and send remittances to support family in country of origin (Negi, 2011). It is then possible that items that include feelings such as feeling “useless” or feeling “no interest in things” may be in sharp contrast to their singular drive to support family by working. To this end, while men in general tend to underreport depressive symptomatology (Courtenay, 2000), Latino migrant day laborers may be further disinclined to attend to and report depressive symptomatology due to their work context. Latino migrant day laborers may also not be able to attend to psychological distress symptoms due to their life context. Without family and other social support systems, Latino migrant day laborers must continue to work to be able to pay for their rent and other expenses. With family in country of origin relying on their remittances, Latino migrant day laborers may be unable to identify with “feeling useless.” As the study sample was recruited from day labor corners where participants congregated to look for employment, it is also possible that this sample of Latino migrant day laborers were less likely to experience somatic distress such as “heart or chest ache” or “nausea and stomach ache.” Latino migrant day laborers who may be experiencing such significant psychological distress may be in their homes or may even return to their country of origin. Finally, another possible difference between our study’s and previous studies’ findings could be due to the different exploratory factor analysis procedures used. Specifically, all of the studies reviewed used principal component analysis while this study used principal axis factor analysis which is the recommended procedure for identifying latent factors and tends to yield more stable factor loadings (Fabrigar et.al., 1999; Miller et al., 2012; Worthington & Whittaker, 2006). Future BSI-18 validation studies should consider using and testing both types of statistical procedures.

As with any study, there are several limitations that should be attached to our findings. While our sample consists of a hard-to-reach population of Latino migrant day laborers, our sample size was small and relied on purposive sampling, thereby limiting generalizability. However, the purpose of the current study was scale validation and not the description of a sample, which somewhat lessens this concern. In addition, the sample consists of a street recruited sample of Latino migrant day laborers, which could have impacted the representation of this subgroup in the study. Latino migrant day laborers who were present at day labor corners could represent individuals who are more motivated to find work and may be less likely to be distressed.

Regardless of these limitations, we believe that our findings help advance the literature on reliable and valid measures with a hard-to-reach population of migrants, as we recruited from a highly transient population that is underrepresented in research and had sufficient power for analysis. Aligned to previous studies, current findings indicate the revised BSI-14 Spanish version can be used as a general measure of psychological distress. The revised BSI-14 Spanish version may be particularly useful in community studies with Latino migrant day laborers, as it is an even briefer version of the original. Future studies should consider conducting CFA to examine and validate one factor of the revised BSI-14 Spanish inventory, as these methods better account for measurement error and provide more stable estimates compared to exploratory factor analysis.

However, as the revised BSI-14 may not fully capture the distinct sociocultural context of LDLs’ lives, the development of measures that are normed on Latinos and other subgroups is recommended to ensure accuracy in gathering data regarding mental health symptomatology. Specifically, researchers and practitioners need to be mindful of within-group differences that may lead to distinct manifestations of distress and related response patterns. For example, in consideration that a majority of Latino migrant day laborers are undocumented, the item “suddenly scared for no reason” may not have broad applicability to Latino migrant day laborers who are often fearful of being deported due to real and valid reasons. Questions related to deportation fear or workers’ rights abuses may then make conceptual sense for inclusion in a distress measure normed on this population. This may be especially true, as research indicates that deportation fear is a prominent and chronic source of stress for many undocumented Latino immigrants (Arbona et al., 2010). The further development of culturally responsive measures that take Latino migrant day laborers’ work and life circumstances into consideration with attention to gender are then recommended.

Footnotes

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: The first author was partly funded through the National Institute of Drug Abuse, R25 DA030310_01A1.

References

Arbona

Olvera

Rodriguez

Hagan

Linares

Weisner

(2010). Acculturative stress among documented and undocumented Latino immigrants in the United States. Hispanic Journal of Behavioral Sciences, 32, 362–384. doi: 10.1177/0739986310373210

Asner-Self

K. K.

Schreiber

J. B.

Marotta

S. A.

(2006). A cross-cultural analysis of the Brief Symptom Inventory-18. Cultural Diversity and Ethnic Minority Psychology, 12, 367–375. doi: 10.1037/1099-9809.12.2.367

Bartlett

M. S.

(1950). Tests of significance in factor analysis. British Journal of Psychology, 3, 77–85.

Cabassa

L. J.

Zayas

L. H.

Hansen

M. C.

(2006). Latino adults’ access to mental health services: A review of epidemiological studies. Administration and Policy in Mental Health and Mental Health Services Research, 33, 316–330. doi: 10.1007/s10488-006-0040-8

Cepeda

Negi

Nowotny

Arango

Kaplan

Valdez

(2012). Social stressor, special vulnerabilities and violence victimization among Latino immigrant day laborers in post-Katrina New Orleans. In Punishing immigrants: Policy, politics, and injustice (pp. 207–231). New York, NY: New York University Press

Courtenay

W. H.

(2000). Constructions of masculinity and their influence on men’s well-being: A theory of gender and health. Social Science and Medicine, 50, 1395–1401.

de Castro

A. B.

Voss

J. G.

Ruppin

Dominguez

C. F.

Seixas

N. S.

(2010). Stressors among Latino day laborers: A pilot study examining allostatic load. American Association of Occupational Health Nurses, 58, 185–196. doi:10.3928/08910162-20100428-01

Fabrigar

L. R.

Wegener

D. T.

MacCallum

R. C.

Strahan

E. J.

(1999). Evaluating the use of exploratory factor analysis in psychological research. Psychological Methods, 4, 272–299.

Gorsuch

R. L.

(1983). Factor analysis (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

10.

Gorsuch

R. L.

(2003). Factor analysis. In Schinka

J. A

Velicer

W. F.

(Eds.), Handbook of psychology: Research methods in psychology (Vol. 2, pp. 143–164). Hoboken, NJ: John Wiley.

11.

Guillemin

Bombardier

Beaton

(1993). Cross-cultural adaptation of health-related quality of life measures: Literature review and proposed guidelines. Journal of Clinical Epidemiology, 46, 1417–1432.

12.

Horn

J. L.

(1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 30, 179–185.

13.

Iwamoto

D. K.

Liao

Liu

W. M.

(2010). Masculine norms, avoidant coping and depressive symptoms among Asian American men. Psychology of Men and Masculinity, 11, 15–24. doi: 10.1037/a0017874

14.

Lazarus

R.S.

Folkman

(1984). Stress, appraisal, and coping. New York: Springer.

15.

Marin

Sabogal

Marin

B. V.

Otero-Sabogal

Perez-Stable

(1987). Development of short acculturation scale for Hispanics. Hispanic Journal of Behavioral Sciences, 9, 183–205.

16.

Meijer

R. R.

deVries

R. M.

van Bruggen

(2011). An evaluation of the brief symptom inventory-18 using item response theory: Which ttems are most strongly related to psychological distress? Psychological Assessment, 23, 193–202. doi: 10.1037/a0021292

17.

Miller

M. J.

Kim

Chen

G. A.

Alvarez

A. N.

(2012). Exploratory and confirmatory factor analyses of the Asian American Racism-Related Stress Inventory. Assessment, 9, 53–64. doi: 10.1177/1073191110392497

18.

Negi

N. J.

(2011). Identifying psychosocial stressors of well-being and factors related to substance use among Latino day laborers. Journal of Immigrant and Minority Health, 13, 748–755. doi: 10.1007/s10903-010-9413-x

19.

Negi

N. J.

(2013). Battling discrimination and social isolation: Psychological distress among Latino day laborers. American Journal of Community Psychology, 51, 164–174. doi: 10.1007/s10464-012-9548-0

20.

O’Connor

B. P.

(2000). SPSS and SAS programs for determining the number of components using parallel analysis and Velicer’s MAP test. Behavior Research Methods, Instrumentation, and Computers, 32, 396–402.

21.

Organista

K. C.

Kubo

(2005). Pilot survey of HIV risk and contextual problems and issues in Mexican/Latino migrant day laborers. Journal of Immigrant Health, 7, 269–281. doi:10.1007/s10903-005-5124-0

22.

Passel

Cohn

(2011). Unauthorized immigrant population: National and state trends, 2010. Washington, DC: Pew Hispanic Center. Retrieved from http://www.pewhispanic.org/2011/02/01/unauthorized-immigrant-population-brnational-and-state-trends-2010/

23.

Prelow

H. M.

Weaves

S. R.

Swenson

R. R.

Bowman

M. A.

(2005). A preliminary investigation of the validity and reliability of the Brief Symptom Inventory-10 in economically disadvantaged Latino American mothers. Journal of Community Psychology, 33, 139–155. doi: 10.1002/jcop.20041

24.

Quesada

(1999). From Central American warriors to San Francisco Latino day laborers: suffering and exhaustion in a transnational context. Transforming Anthropology, 8, 162–185.

25.

Sentell

Shumway

Snowden

L. R

. (2007). Access to mental health treatment by English language proficiency and race/ethnicity. Journal of General Internal Medicine, 22, 289–293. doi: 10.1007/s11606-007-0345-7

26.

Tabachnick

B. G.

Fidell

L. S.

(2001). Using multivariate statistics (4th ed.). New York, NY: Harper & Row.

27.

Torres

L. T.

Miller

M. J.

Moore

K.M.

(2013). Factorial invariance of the Brief Symptom Inventory-18 (BSI-18) for adults of Mexican descent across nativity status, language format, and gender. Psychological Assessment, 25, 300–305. doi:10.1037/a0030436

28.

Valenzuela

Jr (2003). Day labor work. Annual Review Sociology, 29, 307–333. doi:10.1146/annurev.soc.29.010202.100044

29.

Valenzuela

Theodore

Melendez

Gonzalez

A. L

. (2006). On the corner: Day labor in the United States. Los Angeles, CA: Center for the Study of Urban Poverty. Retrieved from http://www.sscnet.ucla.edu/issr/csup/uploaded_files/Natl_DayLabor-On_the_Corner1.pdf

30.

Valdez

Cepeda

Negi

Kaplan

(2010). Fumando la piedra: Emerging patterns of crack cse among Latino immigrant day laborers in New Orleans. Journal of Immigrant & Minority Health, 12, 737–742. doi: 10.1007/s10903-009-9300-5

31.

Walter

Bourgois

Loinaz

Schillinger

(2002). Social context of work injury among undocumented day laborers in San Francisco. Journal of General Internal Medicine, 17, 221–229. doi:10.1046/j.1525-1497.2002.10501.x

32.

Warner

L. A.

Valdez

Vega

W. A.

de la Rosa

Turner

R. J.

Canino

(2006). Hispanic drug abuse in an evolving cultural context: An agenda for research. Drug and Alcohol Dependence, 84S, S8–S16. doi: 10.1016/j.drugalcdep.2006.05.003

33.

Wiesner

Chen

Windle

Elliott

M. N.

Grunbaum

J.A.

Kanouse

D.E.

Schuster

M.A.

(2010). Factor structure and psychometric properties of the Brief Symptom Inventory18 in women: A MACS approach to testing for invariance across racial/ethnic groups. Psychological Assessment, 22(4), 912–922. doi: 10.1037/a0020704

34.

Worthington

R. L.

Whittaker

T. A.

(2006). Scale development research: A content analysis and recommendations for best practice. The Counseling Psychologist, 34, 806–838. doi: 10.1177/0011000006288127

35.

Zabora

Brintzenhofeszoc

Jacobsen

Curbow

Piantadosi

, … Derogatis L. et al. (2001). A new psychosocial screening instrument for use with cancer patients. Psychosomatics, 43, 241–246.

36.

Zwick

William

Velicer

Wayne

(1986) Comparison of five rules for determining the number of components to retain. Psychological Bulletin, 99(3), 432–442. doi: 10.1037/0033-2909.99.3.432