Psychometric Properties of the CES-D, PCL-5, and DERS in a Honduran Adult Sample
Abstract
Abstract:Objective: The objective of this study was to evaluate the psychometric properties of Latin American Spanish-language versions of the Center for Epidemiologic Studies Depression Scale (CES-D), PTSD Checklist for the DSM-5 (PCL-5), and Difficulties in Emotion Regulation Scale (DERS). Methods: Measures were translated, backtranslated, and adapted for use with a Honduran sample (N = 970). Results: Support for the construct validity of the PCL-5 and moderate support for the CES-D were evident. Internal consistency, correlations, and mean differences across groups demonstrated good evidence of internal consistency and convergent validity. The original factor structure for the DERS did not show evidence of construct validity, but internal consistency for the overall scale was good, and there was evidence of convergent validity. Conclusions: The findings provide support for these versions of the PCL-5 but suggest that future work should examine exploratory factor analyses of the CES-D and DERS.
Posttraumatic stress and depression are being examined with increasing frequency in research in Latin America, with studies of the prevalence and impact of common psychiatric disorders recently conducted in Mexico, Honduras, Chile, and Colombia (Barrera-Valencia et al., 2017; Farfán Chávez et al., 2017; Chirinos-Flores et al., 2002; Medina-Mora Icaza et al., 2005). Despite important strides made in several studies in Spain and elsewhere in Latin America on depression and emotion regulation (Gómez-Simón et al., 2014; Soler et al., 1997), the availability of psychometric data on Spanish-language versions of symptom checklists, especially in high-violence contexts of Latin America, is limited. This represents a substantial gap, since Spanish is one of the most spoken native languages in the world and has substantial regional and contextual variation (Lipski, 2012). Thus, the existence of a Spanish translation does not ensure validity in a particular Spanish-speaking context. The current study addresses an important gap regarding the availability of psychometric data for common psychological assessments used in research on individuals exposed to violence in Latin America.
Following exposure to traumatic events, posttraumatic stress and depression are among the most studied constructs and are frequently comorbid (e.g., Goenjian et al., 2001; McFarlane & Kaplan, 2012). Posttraumatic stress refers to a cluster of symptoms experienced in the aftermath a significant adverse event, characterized by actual or threatened death or serious injury (i.e., trauma). These symptoms can include re-experiencing (e.g., nightmares, flashbacks), avoidance (e.g., of places or people associated with the trauma), hyperarousal (e.g., exaggerated startle response), and negative mood and cognition (e.g., persistent, distorted, negative beliefs about oneself or the world; American Psychiatric Association [APA], 2022). Depression is a highly heterogeneous disorder and refers to persistent difficulties with poor mood, including feelings of sadness, irritability, and/or diminished interest or pleasure, in addition to a range of associated symptoms (e.g., weight gain/loss, feelings of worthlessness; APA, 2022). In recent years, emotion regulation – the ability for individuals to effectively modulate emotional responses to internal and external stimuli – has been proposed as a consistent transdiagnostic factor in the development of psychopathology, including both depression and posttraumatic stress (Aldao et al., 2016), making it a highly relevant construct to assess in the context of ongoing violence.
Across studies, there is strong evidence for the cross-cultural validity and relevance of posttraumatic stress, emotion regulation, and depression in understanding individual functioning following exposure to violence. Importantly, while the core sets of disorder-specific symptoms appear to be relatively consistent across contexts, the prevalence of symptom clusters, the presentation of symptoms, and the social implications of symptoms may be variable. For example, one multinational study of emotion regulation showed that cultural values related to the maintenance of social order affected emotion regulation (Matsumoto et al., 2008). Similarly, the general dimensions of posttraumatic stress symptoms seem to be common across cultures, but the prevalence and endorsement of particular symptom groups seems to vary across ethnoracial groups (Hinton & Lewis-Fernández, 2011).
The importance of psychometrically sound assessments, while relevant in all contexts, takes on particular importance in settings where the mental health needs are high. The extraordinarily high rates of crime and homicide in the Northern Triangle of Central America (including Honduras, El Salvador, and Guatemala) paired with the dearth of widely available mental health care has created a pressing need for both empirically sound assessments and evidence-based mental health care (Cuneo et al., 2021; Hiskey et al., 2014). Honduras perennially ranks as one of the most violent countries in the world with an intentional homicide rate 10 times the global average (United Nations Office on Drugs and Crime, 2019). Crimes such as extortion, gender-based violence, and assaults are also prevalent. These levels of crime and violence are largely the result of rival gangs, drug traffickers, and an “inadequate judicial response that fuels impunity, corruption, and high levels of poverty and inequality” (Inter-American Commission on Human Rights, 2015). In Honduras, 18.9% of individuals were directly victimized by reported crime in 2012 alone (Pérez et al., 2013). Whether directly victimized or not, 34% of the population reports feeling insecure in their own neighborhood, nearly three times the rate of neighborhood insecurity reported in the United States (Latin American Public Opinion Project, 2014). Women in Honduras report high rates of exposure to gender-based violence; nearly a third of women experiencing physical or sexual intimate partner violence. In this context, there are limited criminal penalties for this and other forms of gender-based violence that occurs (World Bank, 2023). Despite endemic rates of violence exposure, little is understood about the effects of this systemic violence on either community members or incarcerated offenders in Honduras. Assessing the psychological effects of pervasive violence is limited, in part, by the lack of validated assessments in these contexts. Given that understanding the psychological effects of violence in Honduras is particularly important for policies and programs that seek to reduce risks associated with victimization and offending, the psychometric validation of assessments in the high-violence context is a pressing priority.
Several assessments exist that assess depression, posttraumatic stress, and emotion regulation. However, psychometric data available for Spanish-language mental health assessments in Latin America, and Central America in particular, are not widely available. The goal of the current study was therefore to conduct a psychometric evaluation of often-used measures for evaluating depression (CES-D; Radloff, 1977), posttraumatic stress (PCL-5; Weathers et al., 2013), and difficulties in emotion regulation (DERS; Gratz & Roemer, 2004). For the data collection effort, we selected the CES-D, PCL-5, and DERS as measures used to assess the presence of depression, posttraumatic stress, and difficulties in emotion regulation, respectively.
In the United States, evaluating posttraumatic stress symptoms via the PCL-5 has been evaluated as a four-factor model, consistent with the DSM-5 framework (i.e., symptoms of re-experiencing, hyperarousal, avoidance, and negative mood and cognition) and has achieved adequate fit and have also demonstrated good internal consistency, test–retest reliability, and convergent validity (Blevins et al., 2015). The PCL-5 has also demonstrated evidence of measurement invariance across samples, including configural, metric, and scalar invariance (Caldas et al., 2020). Although no record of the use of the PCL-5 could be found in Latin America, the previous version of the measure (PTSD Checklist-Civilian Version) has been widely used in international research and has established measurement equivalency across English-language and Spanish-language versions (Miles et al., 2008). Given this past work, the current study evaluates the PCL-5 using its four-factor DSM-5 structure (Blevins et al., 2015), which is most consistent with its applied clinical use, the subscale framework also being supported by past work on measurement invariance.
The CES-D has been used in many large-scale studies on depression (Andresen et al., 1994) and has high internal consistency, retest reliability, and validity (Radloff, 1977). The CES-D has been highlighted as one of the best available tools for screening for depression in Spanish-speaking individuals (Reuland et al., 2009), but no record of CES-D validation was found in Latin America. Of note, the CES-D is typically interpreted as a full-scale total score, but the original measure indicated the presence of four factors (Radloff, 1977). Generally speaking, subscales of the CES-D have not demonstrated strong invariance across languages and contexts, suggesting poor invariance at the subscale level (e.g., Leykin et al., 2011). As such, the current study evaluates the CES-D unidimensionally, which is most consistent with its applied clinical use across contexts, including Spanish versions (Soler et al., 1997).
The DERS has been used in numerous studies and has established internal consistency, construct validity, and predictive validity (Gratz & Roemer, 2004). It has six subscales, which include nonacceptance of emotional responses, difficulty engaging in goal-directed behavior, impulse control difficulties, lack of emotional awareness, limited access to emotion regulation strategies, and lack of emotional clarity (Gratz & Roemer, 2004). Although no studies of its use in Latin America were able to be located, the DERS has been validated for use in Spanish in other contexts (Gómez-Simón et al., 2014). These analyses have supported the six-factor scale structure with demonstrated factor, metric and scalar invariance across sexes (Gómez-Simón et al., 2014). Thus, consistent with past work, the current study evaluates the DERS using its six-factor structure.
In addition to evidence of measurement invariance, the current study will also consider convergent validity and evidence for expected group differences in assessments. Previous work has shown that the severity of depression, posttraumatic stress, and difficulties in emotion regulation are all significantly higher in those with a history of violent victimization (Contractor et al., 2017; Lorenzo-Blanco et al., 2016; Iverson et al., 2017). There is also ample literature suggesting that beginning in adolescence, women report significantly more internalizing problems than do men, including greater vulnerability to depression and posttraumatic stress (Nolen-Hoeksema, 2001; Olff et al., 2007). Differences in emotion regulation skills between men and women have been more mixed; the original validation study of the DERS did not find differences across men and women (Gratz & Roemer, 2004). Consequent studies have suggested that there may be gender differences in the item functioning and the reported use of emotion regulation skills, possibly due to extant scales better capturing women’s experiences of emotion than men’s (Anderson et al., 2016). Finally, previous work has identified a relationship between unemployment and poorer mental health (Paul & Moser, 2009). Previous examinations of the convergence of assessments of depression, emotion regulation, and posttraumatic stress and emotion regulation difficulties have also shown high correlations across studies (r ≈ .30–.60; Chang et al., 2018; Goldsmith et al., 2013).
Current Study
The high rates of victimization and perpetration in Honduras have motivated international organizations to develop programs that are specifically designed to benefit the positive development of violence-involved persons (i.e., victims, perpetrators, and potential victims/perpetrators). The data analyzed in the current study were collected as part of the Workforce Development (WFD) activity in Honduras, funded by the US Agency for International Development (USAID). The WFD program is an integral part of a broader portfolio directed to achieve an increase in citizen security among vulnerable populations in urban, high-crime areas (US Agency for International Development, 2015b). To achieve an increase in security, USAID has defined a set of strategically integrated activities that will lead to reductions in crime and violence in a subset of communities in a targeted group of municipalities. Thus, the overall goal of the WFD activity is to increase employment and protective factors for at-risk young adults (ages 16–30 years) living in targeted high crime municipalities in Honduras. A beneficiary of the WFD activity will receive life skills training, cognitive behavioral therapy, basic labor competencies training, vocational training, and job placement. USAID needed a method to measure the impact of its programing in preventing at-risk youth to be victims or perpetrators of violence. For this reason, depression, posttraumatic stress, and emotional regulation are part of an integrated approach to capture the risks associated with violence and target services for violence-involved individuals.
Shorter versions of these assessments have also been used in the context of a multidimensional tool for risk assessment (Hare et al., 2018), but the current study uniquely aims to include the most number of items possible to maintain the maximal consistency with the original measure. The psychometric qualities of the scales were evaluated via three objectives: (1) Conduct a confirmatory factor analysis of the CES-D, PCL-5, and DERS; (2) evaluate measurement invariance across gender, employment, and victimization history (i.e., victimized in the last 6 months or not); and (3) examine scale consistency (internal) and validity (convergent).
Methods
Participants
Participants included 970 adult community members. All participants were interviewed in nine neighborhoods around Tegucigalpa, the capital of Honduras. The sample were 54.17% men. The age of participants in the study ranged from 16 to 34 (M = 22.71, SD = 4.25). In terms of marital status, 64.91% of the sample was single, 34.02% were married, and 1.08% were widowed or divorced. The majority of the sample (78.51%) was not currently employed.
Procedures
Data were collected in April and May, 2017, in Tegucigalpa. For the community sample, we worked with two local NGOs to contact leaders in each community and organize data collection. After obtaining approval from local officials and community leaders to work in each community, we asked those officials and leaders to invite eligible individuals to participate according to the WFD program target population, those aged 16–30 who are currently neither working nor studying (having completed at least 6th grade). The minimum age of 16 was the only exclusionary criteria, while maximum age, working, and studying were guidelines for program recruitment, but not exclusionary, resulting a few individuals older than 30 years in the current sample. Eligible individuals received information on where and when the interview would take place and did not receive incentives for participation. Spaces were selected where both participants and interviewers would feel safe such as a school or a community building. Following each day of data collection, the sample was monitored for gender and age balance. All interviews lasted an average of 40 min.
Measures
An experienced translator from Central America translated measures from English to Spanish. The Spanish version was then backtranslated to English. This process ensured both local contextual understanding and accuracy. The translated measures were then subjected to an item-by-item review during a workshop conducted with 13 individuals matching the target demographic characteristics. Workshop participants were provided with a copy of the measures, and each item was read aloud. Participants were then asked if they understood the item and to describe what the item meant to them. Minor changes were made to the translation during the workshop to ensure understanding and accuracy. For example, in the CES-D, item three refers to not being able to shake off the blues, a phrase that is not used in Spanish. Workshop participants suggested replacing quitarse las tristezas (shed sadness) or with evitar las tristezas (avoid sadness). The former translation was ultimately used as it best reflected the meaning of the question. A pilot of 200 interviews and a final review were conducted with enumerators to ensure enumerator understanding and ability to convey the meaning of each item if further explanation were necessary.
Demographics
Participants provided basic sociodemographic information, including age, gender, and employment status.
Posttraumatic Stress
Symptoms of posttraumatic stress were evaluated using the PTSD Checklist for DSM-5 (PCL-5; Weathers et al., 2013). This 20-item assessment evaluates the severity of symptoms of posttraumatic stress in the four key domains included in the DSM-5: re-experiencing, avoidance, negative mood and cognition, and hyperarousal. For each item, individuals report the extent to which they experienced the given symptom in the past month using a Likert scale of 1–5 (1 = not at all; 5 = extremely).
Depression
Symptoms of depression were evaluated using the Center for Epidemiological Studies Depression Scale (CES-D; Radloff, 1977). This 20-item measure assesses the severity of different aspects of depressed mood, including loneliness, poor appetite, sadness, and restless sleep. For each item, individuals report the extent to which they experienced the item in the past week (0 = rarely or none of the time; 3 = most or all of the time). Four items of the scale are reverse scored (e.g., hopefulness) and items are summed to create a total score.
Difficulties in Emotion Regulation
Emotion regulation problems were evaluated using the Difficulties in Emotion Regulation Scale (DERS; Gratz & Roemer, 2004). This 36-item scale evaluates skills in emotion regulation in six domains: nonacceptance of emotional responses, difficulty in engaging in goal directed behavior, impulse control difficulties, lack of emotional awareness, limited access to emotion regulation strategies, and lack of emotional clarity. For each item, individuals rate the extent to what extent the description is true of them (1 = almost never; 5 = almost always).
Analytic Plan
Given that the selected scales have well-established psychometric properties in English and have been successfully used in other Spanish-speaking contexts, we elected to conduct a confirmatory factor analysis (CFA) of each scale to establish its psychometric properties in Honduras. Missing data were handled using full information maximum likelihood estimation (Enders, 2010). Consistent with previous research and diagnostic conceptualizations of symptom domains, posttraumatic stress symptoms were modeled as a single-level, four-factor model (Blevins et al., 2015). Symptoms of depression were modeled as a single-level, one-factor model (Radloff, 1977), and difficulties in emotion regulation were modeled as a single-level, six-factor model (Gómez-Simón et al., 2014; Gratz & Roemer, 2004). Model fit was evaluated using RMSEA (≤ .01 excellent; ≤ .06 good, ≤ .08, fair; Hu & Bentler, 1999; MacCallum et al., 1996), TLI, and CFI (≥ .95; Hu & Bentler, 1999). Each model was then examined for parameter invariance across groups. Specifically, we evaluated nested models with increasingly strict equivalency: configural invariance, metric invariance, scalar invariance, and strict invariance. Fit indices for each model were examined; decreases of CFI/TLI ≤ .01 and increases in RMSEA of ≤. 015 were considered to indicate invariance across groups (Chen, 2007; Cheung & Rensvold, 2002; Dimitrov, 2010). χ2 difference tests were also evaluated. Finally, internal consistency and convergent validity were examined. Internal consistency was evaluated using Cronbach’s α and McDonald’s ω. Convergent validity was assessed by examining the relationships of the scales to one another. Differences in the scales by gender and victimization history were also examined, to determine whether or not effects were in the expected direction.
Results
Psychometric Properties of the PCL-5
Overall, 20.9% of the sample reported clinically significant symptoms of posttraumatic stress (M = 20.30, SD = 14.80). The overall fit of the four-factor model for the PCL-5 was good (χ2 = 545.80 p < .001; RMSEA = .05, CFI = .95, TLI = .94). All factor loadings showed evidence of good contribution to the subscales (factor loadings > .50; see Table E1 in Electronic Supplementary Materials 1 [ESM 1]). When examining nested models for configural, metric, and scalar invariance using the four-factor model, the PCL-5 demonstrated strong evidence of invariance across gender, employment status, and victimization history (see Table 1). Internal consistency for the PCL-5 total scale was α = .93 and ω = .93. Subscale reliabilities were re-experiencing (α = .82, ω = .82), avoidance (α = .75, ω not calculated since scale has only two items), negative mood and cognition (α = .84, ω = .84), and hyperarousal (α = .78, ω = .79). Those who reported recent victimization had significantly higher reports of total posttraumatic stress [t(939) = −5.90, p < .001, d = 0.47]. Women reported higher levels of total posttraumatic stress than did men [t(945) = −3.57, p < .001, d = 0.37]. There no differences in total posttraumatic stress [t(944) = 0.74, p = .458, d = 0.06] by employment status.
Psychometric Properties of the CES-D
Overall, 37.0% of the sample reported clinically significant levels of depressed mood (M = 14.19, SD = 10.04). Depression symptoms reported on the CES-D were modeled as a one-factor model, consistent with previous work (Radloff, 1977). The overall fit of the model for the full sample was adequate (χ2 = 605.13 p < .001; RMSEA = .05, CFI = .91, TLI = .90). Some items (n = 5), however, exhibited poor factor loadings (< .40; see Table E2 in ESM 1). Invariance testing of the one-factor CES-D model by gender indicated that the configural invariance showed adequate (but not strong) fit, but support for measurement invariance across gender, employment status, and victimization history. Total scale consistency for the CES-D was α = .86 and ω = .87. Those who reported recent victimization had significantly higher reports of depression [t(943) = −5.72, p < .001, d = 0.44]. Women reported significantly more symptoms of depression than did men [t(951) = −4.48, p < .001, d = 0.29]. Individuals who were unemployed had significantly higher levels of depressed mood [t(966) = 2.69, p = .017, d = 0.19] than those who were working.
Psychometric Properties of the DERS
On average, individuals in the sample reported low levels of emotion regulation difficulties (M = 2.15, SD = 0.56). Difficulties in emotion regulation as reported on the DERS were models as a six-factor model, consistent with previous psychometric evaluations of the measure (Gratz & Roemer, 2004). The overall fit of the six-factor model for full sample was not strong (χ2 = 2,321.64, p < .001, RMSEA = .06, CFI = .80, TLI = .78). Ten items showed poor contribution to subscales (≤ .40; see Table E3 in ESM 1). Because model fit was not strong, invariance testing was not conducted. Internal consistency for the DERS total scale was α = .86 and ω = .86. Subscale reliability were nonacceptance of emotional responses (α = .78, ω = .78), difficulty engaging in goal-directed behavior (α = .67, ω = .69), impulse control difficulties (α = .69, ω = .74), lack of emotional awareness (α = .65, ω = .65), limited access to emotion regulation strategies (α = .68, ω = .69), and lack of emotional clarity (α = .46, ω = .47). Those who reported recent victimization had significantly more total emotion regulation difficulties [t(958) = −2.61, p = .009, d = −0.22]. There were no large differences between men and women in total emotion regulation difficulties [t(967) = −1.94, p = .052, d = −0.13], and there were no differences in total emotion regulation difficulties by employment status [t(966) = 1.10, p = .272, d = 0.09].
Convergent Validity
Posttraumatic stress, depressed mood, and emotion regulation difficulties were highly correlated at the expected magnitude (PCL-5 & CES-D: r = .65, p < .001; PCL-5 & DERS: r = .52, p < .001; CES-D & DERS: r = .62, p < .001).
Discussion
Rates of violence exposure are high. Yet programs seeking to provide services to violence-involved groups are met with a dearth of validated Spanish-language mental health assessments in local dialect that can be used to establish empirical evidence for programs tailored to individual symptoms and the success or failure of programs overall. The goal of the current study was therefore to conduct a psychometric evaluation of widely used measures for evaluating depression (CES-D; Radloff, 1977), post-traumatic stress (PCL-5; Weathers et al., 2013), and difficulties in emotion regulation (DERS;Gratz & Roemer, 2004). Data were collected on a sample of community adults (N = 970). Overall, the psychometric support for the measures in these data was strong for the PCL-5 in this context and moderate for the CES-D. The six-factor structure for the DERS was not supported, although tests of internal consistency and convergent validity, however, suggested good support and group differences in the measure were in the expected direction. As such, future work should consider evaluating alternative factor structures and item-level modifications for the DERS in this context.
The overall fit for the CFA evaluating the construct validity of the four-factor structure of the PCL-5 was good. In addition, there was strong evidence for parameter invariance across groups by gender, employment, and victimization history (see Table 2). Consistent with past work (Contractor et al., 2017; Lorenzo-Blanco et al., 2016), the PCL-5 was higher for those with a recent history of violent victimization as well as for women (Olff et al., 2007). Together, with evidence of strong interrelations between symptoms of posttraumatic stress and depression and emotion regulation, the PCL-5 demonstrated high psychometric quality on all dimensions assessed.
The overall fit for the CFA evaluating the construct validity of the CES-D demonstrated adequate fit, but model fit was not further disadvantaged by metric, scalar, and strict invariances (see Table 2). Thus, while future work should consider possible alternative scale structures for the CES-D that consider alternative factor structures or items, there is evidence that the CES-D items demonstrate good evidence of invariance across groups. Moreover, the CES-D demonstrated evidence of good internal consistency and convergent validity, in this context with relations with other assessments of mental health and across groups (i.e., victims/nonvictims, employed/unemployed, men/women) all in the expected direction, based on previous work. Moreover, group differences were in the expected direction.
The overall fit for the six-factor structure of the DERS was not strong, and a large percentage of items (10 items; see Table E3 in ESM 1) had poor factor loadings. It is originally hypothesized factor structure of the DERS may have been affected by a lack of semantic equivalence between Spanish and English in the translation on some items. Although the scale was forward and backtranslated, consistent with best practice, and the translation was consistent with the Spanish DERS in other studies (Gómez-Simón et al., 2014), there is no semantic equivalent for the word upset in Spanish, which is a critical component of the stem of many DERS items. The closest matching word in Latin American Spanish, molestar (to be bothered by), does not generally capture the same degree of emotional distress or the range of negative emotions that is captured by the word upset. As such, it is likely that the Spanish version of the questionnaire is tapping into a somewhat different emotional trigger for regulatory strategies than is the English version of the questionnaire.
Of note, there was substantial variability the performance of the DERS subscales. That is, nonacceptance of emotional responses overall performed well, with good internal evidence of internal consistency and with all items achieving sufficiently high factor loadings. The dimensions of difficulty engaging in goal-directed behavior, impulse control difficulties, lack of emotional awareness, and limited access to emotion regulation strategies all demonstrated evidence of internal consistency that approaches cut-offs for acceptability, and an examination of factor loadings suggested that all items with the exception of 1–2 on each subscale had adequate factor loadings. As such, these subscales may demonstrate acceptable psychometric properties with some additional specific attention to these specific items (e.g., re-evaluating wording, semantic equivalence, considering alternative factor structures). The only subscale that was far below acceptable levels for internal consistency and for which the majority of items failed to load was lack of emotional clarity. In sum, while additional work is required to consider possible additional modifications to the DERS for use in this context, these analyses are promising.
Future work might also consider building on these scales by evaluating culture-specific manifestations of distress that might be meaningfully incorporated into measurements of these constructs. Previous work has suggested that while many of the core symptoms and experiences of difficulties across context are shared, scales that are developed external to a context may be lacking other context-specific and culture-specific manifestations of distress (Pincay & Guarnaccia, 2007). Although the translation of scales across contexts is important insomuch as it facilitates the rapid availability of measurement resources, qualitative inquiry regarding the expression of symptoms rising from ethnographic studies and other forms of emic information should be used to further modify and update measurements of depression, post-traumatic stress, and emotion regulation.
Although the current study had a number of strengths and focuses on an understudied geographic area, several limitations must also be noted. First, because these scales were implemented as part of a large-scale agency effort that was not directed at research as its primary goal, not all types of validity and reliability were able to be assessed. In addition, although the sample was drawn from an understudied region, it is important to note that all participants were from metropolitan Tegucigalpa and were not a random population. In addition, the sample is not generalizable in terms of age and employment status. For these reasons, the psychometric characteristics observed in the current study may not be broadly generalizable; future evaluation in Honduras that draws from more diverse geographies is therefore recommended. In addition, sensitivity analyses that consider Likert scale responses as ordinal variables rather than as dimensional variables using WLSMV estimation is an important next step for future work. Despite these limitations, the study nonetheless is an important initial contribution to psychometric work and measures development in this region.
Electronic Supplementary Materials
The electronic supplementary material is available with the online version of the article at https://doi.org/10.1027/2698-1866/a000055
References
(2016). Emotion regulation as a transdiagnostic factor in the development of internalizing and externalizing psychopathology: Current and future directions. Development and Psychopathology, 28(4pt1), 927–946. 10.1017/S0954579416000638
(2022). Diagnostic and statistical manual of mental disorders (5th ed., text rev.). 10.1176/appi.books.9780890425787
(2016). Gender-based differential item function for the difficulties in emotion regulation scale. Personality and Individual Differences, 92, 87–91. 10.1016/j.paid.2015.12.016
, (1994). Screening for depression in well older adults: Evaluation of a short form of the CES-D. http://psycnet.apa.org/psycinfo/1994-43563-001
(2017). Cognitive profiles of post-traumatic stress disorder and depression in children and adolescents. International Journal of Clinical and Health Psychology: IJCHP, 17(3), 242–250. 10.1016/j.ijchp.2017.05.001
(2015). The posttraumatic stress disorder checklist for DSM-5 (PCL-5): Development and initial psychometric evaluation. Journal of Traumatic Stress, 28(6), 489–498. 10.1002/jts.22059
(2020). Factor structure and multi-group measurement invariance of posttraumatic stress disorder symptoms assessed by the PCL-5. Journal of Psychopathology and Behavioral Assessment, 42, 364–376. 10.1007/s10862-020-09800-z
(2018). Emotion regulation is associated with PTSD and depression among female adolescent survivors of childhood sexual abuse. Psychological Trauma: Theory, Research, Practice and Policy, 10(3), 319–326. 10.1037/tra0000306
(2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. Structural Equation Modeling, 14(3), 464–504. 10.1080/10705510701301834
(2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modeling, 9(2), 233–255. 10.1207/S15328007SEM0902_5
(2002). Prevalencia de trastornos mentales en la población mayor de 18 años en 29 comunidades urbanas de Honduras, 2001 [Prevalence of mental disorders in the adult population in 29 urban communities in Honduras]. Revista Médica de los Post Grados de Medicina Universidad Nacional Autónoma de Honduras (UNAH), 7(1), 42–48. http://www.bvs.hn/RMP/pdf/2002/pdf/Vol7-1-2002-10.pdf
(2017). Heterogeneity in patterns of DSM-5 posttraumatic stress disorder and depression symptoms: Latent profile analyses. Journal of Affective Disorders, 212(1), 17–24. 10.1016/j.jad.2017.01.029
(2021). What counts as “safe?”: Exposure to trauma and violence among asylum seekers from the northern triangle. Health Affairs, 40(7), 1135–1144.
(2010). Testing for factorial invariance in the context of construct validation. Measurement and Evaluation in Counseling and Development, 43(2), 121–149. 10.1177/0748175610373459
(2010). Applied missing data analysis. Guilford Press.
(2017). Exposición a la violencia y su relación con la salud mental en estudiantes de educación media superior en ciudad Juárez, México [Violence exposure and its relationship with the mental health of higher education students in Juarez, Mexico]. European Scientific Journal, ESJ, 13(10), 1857–7881. 10.19044/esj.2017.v13n10p%25p
(2001). Posttraumatic stress and depressive reactions among Nicaraguan adolescents after Hurricane Mitch. The American Journal of Psychiatry, 158(5), 788–794. 10.1176/appi.ajp.158.5.788
(2013). Emotion regulation difficulties mediate associations between betrayal trauma and symptoms of posttraumatic stress, depression, and anxiety. Journal of Traumatic Stress, 26(3), 376–384. 10.1002/jts.21819.
(2014). Factor structure and measurement invariance of the difficulties Emotion Regulation Scale (DERS) in Spanish adolescents. Psicothema, 26(3), 401–408, 10.7334/psicothema2013.324
(2004). Multidimensional assessment of emotion regulation and dysregulation: Development, factor structure, and initial validation of the difficulties in emotion regulation scale. Journal of Psychopathology and Behavioral Assessment, 26(1), 41–54. 10.1023/B:JOBA.0000007455.08539.94
(2018). Identifying high-risk young adults for violence prevention: A validation of psychometric and social scales in Honduras. Journal of Crime and Justice, 41(5), 627–642.
(2011). The cross-cultural validity of posttraumatic stress disorder: Implications for DSM-5. Depression and Anxiety, 28(9), 783–801. 10.1002/da.20753
(2014). Violence and migration in Central America. Americas Barometer Insights, 101, 1–8. https://www.vanderbilt.edu/lapop/insights/IO901en.pdf
(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. 10.1080/10705519909540118
. (2015). Situation of Human Rights in Honduras. Organization of American States. http://www.oas.org/en/iachr/reports/pdfs/Honduras-en-2015.pdf
(2017). Intimate partner violence victimization and associated implications for health and functioning among male and female post-9/11 veterans. Medical Care, 55, S78–S84. 10.1097/MLR.0000000000000741
. (2014). AmericasBarometer. Vanderbilt University. http://www.vanderbilt.edu/lapop
(2011). Factor structure of the CES-D in a sample of Spanish-and English-speaking smokers on the Internet. Psychiatry Research, 185(1-2), 269–274. 10.1016/j.psychres.2010.04.056
(2012). Geographical and social varieties of Spanish: An overview. In J. I. HualdeA. OlarreaE. O'Rourke (Eds.), The Handbook of Hispanic Linguistics (pp. 1–26). Wiley.
(2016). Profiles of bullying victimization, discrimination, social support, and school safety: Links with Latino/a youth acculturation, gender, depressive symptoms, and cigarette use. The American Journal of orthopsychiatry, 86(1), 37–48. 10.1037/ort0000113
(1996). Power analysis and determination of sample size for covariance structure modeling. Psychological Methods, 1(2), 130–149. 10.1037/1082-989X.1.2.130
(2008). Culture, emotion regulation, and adjustment. Journal of Personality and Social Psychology, 94(6), 925–937. 10.1037/0022-3514.94.6.925
(2012). Evidence-based psychological interventions for adult survivors of torture and trauma: A 30-year review. Transcultural Psychiatry, 49(3-4), 539–567. 10.1177/1363461512447608
(2005). Prevalencia de sucesos violentos y de trastorno por estrés postraumático en la población mexicana [Prevalence of violent events and posttraumatic stress disorders in the Mexican population]. Salud pública de México, 47(1), 8–21. http://www.insp.mx/salud/47/eng
(2008). Spanish and English versions of the PTSD Checklist – Civilian version (PCL-C): Testing for differential item functioning. Journal of Traumatic Stress, 21(4), 369–376. 10.1002/jts.20349
(2001). Gender differences in depression. Current Directions in Psychological Science, 10(5), 173–176. 10.1111/1467-8721.00142
(2007). Gender differences in posttraumatic stress disorder. Psychological Bulletin, 133(2), 183–204. 10.1037/0033-2909.133.2.183
(2009). Unemployment impairs mental health: Meta-analyses. Journal of Vocational Behavior, 74(3), 264–282. 10.1016/j.jvb.2009.01.001
(2013). Political culture of democracy in Honduras and in the Americas, 2012: Towards equality of opportunity. Vanderbilt University. https://pdf.usaid.gov/pdf_docs/PA00K88B.pdf
(2007). “It’s like going through an earthquake”: Anthropological perspectives on depression among Latino immigrants. Journal of Immigrant and Minority Health, 9(1), 17–28. 10.1007/s10903-006-9011-0
(1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1(3), 385–401. 10.1177/014662167700100306
(2009). Diagnostic accuracy of Spanish language depression-screening instruments. The Annals of Family Medicine, 7(5), 455–462. 10.1370/afm.981
(1997). Validation study of the Center for Epidemiological Studies-Depression of a Spanish population of patients with affective disorders. Actas Luso-Espanolas de Neurologia, Psiquiatria y Ciencias Afines, 25(4), 243–249. https://europepmc.org/article/med/9412163
. (2019). Global study on homicide. https://www.unodc.org/unodc/en/data-and-analysis/global-study-on-homicide.html
. (2015a). Gender-based violence analysis for USAID/Honduras: Final report. https://encompassworld.com/resources/gender-based-violence-analysis-usaidhonduras-final-report
. (2015b). Country Development Cooperation Strategy 2015–2019. https://www.usaid.gov/honduras/cdcs
(2013). The PTSD checklist for DSM-5 (PCL-5). National Center for PTSD. https://www.ptsd.va.gov
(2023). Gender-based violence country profile: Honduras. https://openknowledge.worldbank.org/bitstreams/0b30d540-24de-4b70-944b-ddf111d33eb3/download#:~:text=GBV%20Country%20Profile%3A%20HONDURAS,-Page%204&text=rank%20(67th).-,Honduras%20has%20one%20of%20the%20highest%20rates%20of%20violent%20deaths,one%20murder%20every%2027%20hours
