Abstract
This study evaluated the psychometric properties of the Spanish version of the God Locus of Health Control scale, a measure of the extent to which an individual believes God has control over one’s health, among a sample of churchgoing Latinas (N = 398). Confirmatory factor analysis showed support for a one-factor structure and internal consistency reliability, as measured by Cronbach’s coefficient alpha, was good. Evidence for convergent validity was demonstrated by significant correlations in the expected magnitudes and directions with two measures of perceived religious involvement in health. These results suggest that the God Locus of Health Control scale can be used to examine the extent to which God is perceived to control an individual’s health among Latinas.
Approximately 80% of Latinos in the United States report a religious affiliation (Pew Research Center, 2014a). Most identify as Catholic, followed by Protestant and Other Christian. In addition, Latinos show high rates of religious practice with 59% of Latinos reporting praying at least daily and 40% attending a worship service weekly (Pew Research Center, 2014b). Religion has been shown to play an important role in health among several groups, including Latinos (Koenig, 2012; Lujan & Campbell, 2006). Reviews examining this relationship suggest that religion is often protective of mental and physical health (Robinson, Bolton, Rasic, & Sareen, 2012; Williams & Sternthal, 2007). Religion is believed to affect health, in part, through cognitive mechanisms. One such mechanism is health locus of control or the beliefs individuals hold about control over their health (Wallston, 2005).
Different types of health-related locus of control have been studied including internal, chance, and powerful others (e.g., health care professionals; Luszczynska & Schwarzer, 2005; Wallston, Wallston, & DeVellis, 1978). One type of health locus of control that has received less attention is God health locus of control (Wallston et al., 1999; Welton, Adkins, Ingle, & Dixon, 1996). Among individuals who are religious, health may be attributed to God (Welton et al., 1996). Having a strong God health locus of control reflects the belief that God is responsible for one’s health. The God Locus of Health Control (GLHC) scale was developed to examine the extent to which one’s health or disease state is attributed to God’s will, specifically among individuals with acute and chronic medical conditions (Wallston et al., 1999). The GLHC scale was first validated among three samples, one of patients with systemic sclerosis (n = 93) and two of patients with rheumatoid arthritis (n = 308). Internal consistency reliability was excellent in all three samples (α = .87-.94). Convergent validity was demonstrated for the measure via significant correlations in the expected directions with measures of religiosity (Wallston et al., 1999). A general health-related version of the measure was also developed for individuals without an acute or chronic medical condition. In the general health-related version, “condition” is replaced with “health” for all items (e.g., “whether or not my [condition] improves is up to God”). The general health version of the GLHC scale was used in the present study.
Although the GLHC scale has been used in several studies, little information is available on the psychometric properties of the measure among Latinos. Among a sample of 436 English- and Spanish-speaking participants, the general health version of the GLHC scale demonstrated good internal consistency reliability in the total sample (α = .86) and when examined separately by language preference (English: α = .91, Spanish: α = .81; Champagne, Fox, Mills, Sadler, & Malcarne, 2015; Fox, Malcarne, Roesch, & Sadler, 2014). In addition, GLHC scores were generally significantly associated with health attributions, as measured by the Cultural Health Attributions Questionnaire (Murguía, Zea, Reisen, & Peterson, 2000), in the expected directions in the full sample, and when examined separately by language preference. To date, no studies have examined the structural validity of the GLHC scale among a sample of Latinos, and internal consistency reliability and convergent validity data are limited. A measure should be demonstrated to be valid and reliable in a population prior to its use. Given the growing Latino population in the United States (U.S. Census Bureau, 2015), of whom many report a religious affiliation and speak Spanish, psychometric evaluation of the Spanish language version of the GLHC scale among Latinos is needed. The present study focused on Latina women. Studies have shown that religion plays an important role in health among Latina women, influencing coping mechanisms and health beliefs (Jurkowski, Kurlanska, & Ramos, 2010). In addition, Latina women have been shown to ascribe more importance to religion and engage more in religious services and prayer as compared with Latino men (Pew Research Center, 2016). Thus, there may be sex differences in the relationship between religion and health among Latinos. Examining the GLHC scale among Latina women only will prevent study results from being potentially biased by differences in the role of religion in health across gender.
The first aim of the present study was to examine the structural validity of the GLHC scale in a sample of Latinas. Based on previous studies showing good internal consistency reliability (Champagne et al., 2015; Fox et al., 2014; Wallston et al., 1999), a one-factor structure was hypothesized. The structural validity of the measure has not been previously evaluated using factor analysis in Latinos. The second aim was to examine the internal consistency reliability of the measure. Based on previous studies (Champagne et al., 2015; Fox et al., 2014; Wallston et al., 1999), internal consistency reliability was expected to be excellent. The third aim was to examine the convergent validity of the GLHC scale with the Perceived Religious Influence on Health Behavior scale and the Illness as Punishment for Sin scale, two measures of perceived religious involvement in health (Holt et al., 2009). The perceived religious involvement in health measures examine the degree to which participants believe religion affects their health (e.g., “God helps me to avoid bad health habits”) and whether illness is a result of punishment for sin from God (e.g., “God uses sickness to send a message to people”). The GLHC scale was expected to be moderately positively associated with the two measures of perceived religious involvement in health because both measures, in part, query about God’s involvement in one’s health behaviors and health status. The final aim was to examine how GLHC scores were associated with sociodemographic variables to provide insight into the extent to which participants reported God health locus of control in the present study.
Method
Participants and Procedures
This study used baseline data collected among a sample of Latinas recruited from 16 churches that had at least one Spanish-speaking service and a minimum of 200 Latino families. All churches were participating in the Fe en Acción/Faith in Action physical activity promotion intervention in San Diego, California. To be eligible for inclusion, individuals had to self-identify as a Latina woman, be between 18 and 65 years old, attend church at least 4 times a month for any reason, live within a 15-minute drive from their church, have reliable transportation to get to their church, have no barriers to attend church services, and plan on attending their church for the next 24 months. To recruit low active participants, who would potentially benefit the most from the intervention, women also had to report low levels of physical activity and engage in low levels of accelerometer-measured physical activity. Additional information on the design of the parent study can be found elsewhere (Arredondo et al., 2015). Participants completed baseline study questionnaires in their preferred language of either English or Spanish. Only participants who completed study questionnaires in Spanish and completed all items of the GLHC scale were included in the present study (N = 398). English preference participants were excluded from the present study to provide psychometric data on the Spanish version of the measure. Participants provided informed consent prior to study participation. The San Diego State University Institutional Review Board approved study procedures and materials prior to human subject enrollment.
Measures
Sociodemographic variables were collected at baseline as part of the parent study. All measures reported in the current study were translated and back translated to Spanish and English according to standard protocols (Harkness, 2003).
God Locus of Health Control scale
The GLHC (Wallston et al., 1999) scale is a six-item measure that assesses the extent to which an individual believes God has control over one’s health. Response options are on a 6-point scale ranging from 1 = strongly disagree to 6 = strongly agree. Total scores are calculated by summing individual items and range from 6 to 36. Higher scores indicate a greater belief that God has control over one’s health.
Perceived Religious Influence on Health Behavior scale
The Perceived Religious Influence on Health Behavior scale (Holt et al., 2009) is a seven-item measure that evaluates the extent to which an individual’s religion is perceived to influence their health behaviors. Response options are on a 4-point scale and range from 1 = strongly disagree to 4 = strongly agree. Total scores are calculated by summing individual items and range from 7 to 28. Higher scores indicate greater endorsement that an individual’s religion influences their health behaviors. Internal consistency reliability in the present study was good (α = .87).
Illness as Punishment for Sin scale
The Illness as Punishment for Sin scale (Holt et al., 2009) is an eight-item measure that examines the degree to which illness is believed to be a punishment for sin. Response options are on a 4-point scale and range from 1 = strongly disagree to 4 = strongly agree. Total scores are calculated by summing individual items and range from 8 to 32. Higher scores indicate greater belief that illness is a punishment for sin. Internal consistency reliability in the present study was good (α = .81).
Data Analysis
Descriptive statistics for the sample were calculated in SPSS version 23 (IBM, 2015). For the first aim, the structural validity of the GLHC scale was evaluated using confirmatory factor analysis in Mplus version 7.2 (Muthén & Muthén, 2012). As recommended by Bentler (2007), overall model fit was determined by consulting descriptive fit indices: (a) the root mean square error of approximation (RMSEA; Steiger, 1990), (b) the standardized root mean square residual (SRMR; Hu & Bentler, 1999), and (c) the robust comparative fit index (CFI; Bentler, 1990). For RMSEA and SRMR indices, values less than .08 were considered acceptable fit and values less than .05 were considered good fit. For CFI, values greater than .90 were considered acceptable fit and values greater than .95 were considered good fit. Models were determined to have acceptable model fit if values for at least two of the descriptive fit indices indicated at least acceptable model fit. The chi-square test of model fit was also reported for completeness, but not used as a primary indicator of model fit because it is sensitive to sample size and almost always significant, and thus not a good index of fit (Bentler, 1990). Factor loadings greater than .30 were considered acceptable, according to recommendations by Hair, Black, Babin, and Anderson (2010). For the second aim, internal consistency reliability was examined using Cronbach’s coefficient alpha. For the third aim, Pearson’s product-moment correlation coefficients were evaluated to examine the relationships between the GLHC scale and (a) the Perceived Religious Influence on Health Behavior scale and (b) the Illness as Punishment for Sin scale. For the final aim, associations of sociodemographic variables with GLHC scores were evaluated using Pearson’s product-moment correlation coefficients and independent-samples t-tests.
Results
Descriptive Statistics
Participants (N = 398) had a mean age of 44.95 years (SD = 8.98). The majority of participants had less than a high school degree, were employed and married. Approximately half (56%) of the sample had a monthly income of less than US$2,000. Most participants (94%) were born in Mexico. Few were born in the United States (4%) or another country (2%). The mean number of years lived in the United States was 20 years (SD = 9).
The mean GLHC scale score was 15.92 (SD = 7.83). Response options for individual items ranged from a mean score of 1.50 (Item 2: “La mayoría de las cosas que afectan a mi salud sucede debido a Dios” [“Most things that affect my health happen because of God”]) to 3.88 (Item 1: “Si empeora mi salud, se depende a Dios para determinar si yo sentiría mejor de nuevo” [“If my health worsens, it is up to God to determine whether I will feel better again”]). Except for Item 1, all individual items were positively skewed with participants most often reporting that they “strongly disagree” in response to the item. The response options to Item 1 showed a bimodal distribution with 34.4% of the sample reporting that they “strongly agree” and 25.9% of participants reporting that they “strongly disagree,” followed by “moderately agree” (17.8%), “slightly agree” (9.0%), “moderately disagree” (8.0%), and “slightly disagree” (4.8%). Correlations among individual items can be found in Table 1.
Pearson Correlations Among Individual Items of the God Locus of Health Control Scale.
Note. All correlations are significant at p < .01.
Structural Validity, Reliability, and Convergent Validity
Confirmatory factor analysis showed support for a one-factor model (see Table 2). The one-factor model had acceptable model fit (CFI = .96; SRMR = .04; RMSEA = .10; χ2(9) = 42.64, p < .01). In addition, individual item standardized factor loadings ranged from .35 to .81 and were all-significant, providing further support for the one-factor model. The internal consistency reliability of the measure was good (α = .81). GLHC scores were significantly, positively associated with the Perceived Religious Influence on Health Behavior scale (r = .45, p < .01) and the Illness as Punishment for Sin scale (r = .39, p < .01).
Means, Standard Deviations, and Factor Loadings for the God Locus of Health Control Scale.
Note. * p < .001.
Sociodemographics
GLHC scores were higher among participants with less than a high school education (M = 17.09, SD = 8.06) as compared with individuals with a high school education or more (M = 14.28, SD = 7.26; t(394) = 3.57, p < .01). There were no significant differences (p > .05) in GLHC scores, however, across marital status (married vs. single), employment status (employed vs. not employed), monthly household income (less than US$2,000 vs. US$2,000 or more), and country of birth (United States vs. foreign born). In addition, GLHC scores were not significantly correlated with age or the number of years a participant lived in the United States.
Discussion
The present study evaluated the psychometric properties of the Spanish version of the GLHC scale in a sample of Latina women. Confirmatory factor analysis provided support for a one-factor structure and internal consistency reliability was good. Correlations with convergent validity measures were significant, moderate, and in the expected directions. The GLHC scale was moderately, positively associated with two measures of religious involvement in health. Findings from this study suggest that the Spanish version of the GLHC scale can be used among Spanish-speaking Latinas.
Mean GLHC scores in the present study were lower than a previous sample of Presbyterians who also completed the general health version of the GLHC scale (Benjamins, Ellison, Krause, & Marcum, 2011). GLHC scores were expected to be higher in the present study where all participants reported high levels of religious activity (i.e., attending church at least 4 times a month) and the majority of participants reported less than a high school education. Religious practice and less education have previously been associated with higher GLHC scores among majority European American samples of patients with rheumatoid arthritis and systemic sclerosis (Wallston et al., 1999). On the other hand, religious activity may not be as closely associated with beliefs about God’s control over health because high levels of religious activity are so prevalent among Latinas. Other aspects of religion, besides religious involvement, may need to be explored to identify variables associated with GLHC scores. For example, fatalism may be more closely associated with GLHC scores among Latinas. Although mean scores were lower than expected, there was large variation in scores. Scores ranged from 6 to 36. In the present study, except for education, GLHC scores were not significantly associated with sociodemographic variables. This may be a result of the limited sociodemographic variation among the study participants. Exploration of sociodemographic correlates of GLHC scores in a more diverse sample may provide insight into the low GLHC scores reported in the present study.
God health locus of control has been associated with health behaviors and outcomes, yet few studies have examined this relationship in Latinos. This may be a result of the limited psychometric information about the measure available in this population. One study used latent profile analysis to examine how different types of health locus of control (i.e., God, chance, powerful others, and internal) relate to health behaviors and outcomes, among other variables, in a sample of 436 English- and Spanish-speaking Latinos (Champagne et al., 2015). There were no differences in health behaviors (i.e., tobacco and alcohol consumption) and outcomes (i.e., colorectal cancer screening, anxiety, depression) among participants who reported more God health locus of control, along with other forms of external health loci of control (e.g., chance), as compared with participants reporting more internally oriented health locus of control. In other ethnic minority samples, however, lower reports of God health locus of control have been associated with positive health behaviors. For example, lower reports of God health locus of control were associated with greater adherence to clinical breast examination and mammography recommendations in African American women 25 years and older (Kinney, Emery, Dudley, & Croyle, 2002) and greater adherence to antiretroviral therapy among a predominantly African American sample of HIV-infected individuals (Finocchario-Kessler et al., 2011). Future studies should use the GLHC scale to examine how this construct relates to health behaviors and outcomes among Latinas.
The present findings should be interpreted while considering study limitations. Participants were recruited from a single, metropolitan border city in the southwestern United States, and all participants were Catholic and frequently attended church, limiting generalizability. In addition, all participants in the present study were female and spoke Spanish, limiting external validity to Latino men and those with an English-language preference. Future studies should examine whether GLHC scale scores can be meaningfully compared across English- and Spanish-speaking Latinos.
In summary, findings from the present study suggest that the GLHC scale is a reliable and valid measure of God health locus of control in churchgoing Latinas. The six-item measure is brief, minimizing participant burden. Future studies should use the GLHC scale to examine the extent to which God is believed to control one’s health in Latinas.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the National Cancer Institute (Grants R01CA138894, R01CA138894-04S1, R01CA138894-04S2, U54CA132384, F31CA206334).