Violeta J Rodriguez1,2,3, Pablo D Radusky4, Mahendra Kumar1, Charles B Nemeroff1, Deborah Jones1. 1. Department of Psychiatry and Behavioral Sciences, University of Miami Miller School of Medicine, Miami, FL, United States. 2. Department of Psychology, University of Georgia, Athens, GA, United States. 3. Ford Foundation Fellowship, National Academies of Sciences, Engineering, and Medicine, Washington, DC, United States. 4. Department of Psychology, University of Buenos Aires, Argentina.
Abstract
BACKGROUND: Assessing traumatic childhood events has important implications for treatment, due to increased high-risk behaviors, treatment nonadherence, and all-cause mortality. As such, it is important to ensure that screening tools used to measure traumatic childhood events are invariant across groups. The focus of this study was to examine measurement invariance across gender, poverty level, and HIV status in a commonly used childhood trauma screening tool, the Childhood Trauma Questionnaire - Short Form (CTQ-SF). METHOD: Participants were N= 473 HIV-infected and uninfected men and women who completed a demographic questionnaire, the CTQ-SF, and underwent HIV testing. RESULTS: Participant age was an average of 36 years (SD= 9.40); 51% of participants were male, and 49% were female. Forty-three percent of participants were below the poverty level, and 36% were HIV-infected. Configural invariance was supported by gender, poverty level, and HIV status; scalar and strict invariance were not supported by gender, poverty level, and HIV status. Neither full nor partial metric invariance could be established by gender and income; however, the scale was invariant at the metric level by HIV status. DISCUSSION: Given the measurement bias identified in gender, poverty level, and HIV, practitioners and researchers must use caution when drawing conclusions regarding childhood trauma when using the CTQ-SF. Findings also suggest that statistical inferences and implications for practice based on comparisons of observed means will be distorted and may be misleading, and as such, established cutoffs may not apply similarly for these groups, suggesting an avenue for further research.
BACKGROUND: Assessing traumatic childhood events has important implications for treatment, due to increased high-risk behaviors, treatment nonadherence, and all-cause mortality. As such, it is important to ensure that screening tools used to measure traumatic childhood events are invariant across groups. The focus of this study was to examine measurement invariance across gender, poverty level, and HIV status in a commonly used childhood trauma screening tool, the Childhood Trauma Questionnaire - Short Form (CTQ-SF). METHOD: Participants were N= 473 HIV-infected and uninfected men and women who completed a demographic questionnaire, the CTQ-SF, and underwent HIV testing. RESULTS: Participant age was an average of 36 years (SD= 9.40); 51% of participants were male, and 49% were female. Forty-three percent of participants were below the poverty level, and 36% were HIV-infected. Configural invariance was supported by gender, poverty level, and HIV status; scalar and strict invariance were not supported by gender, poverty level, and HIV status. Neither full nor partial metric invariance could be established by gender and income; however, the scale was invariant at the metric level by HIV status. DISCUSSION: Given the measurement bias identified in gender, poverty level, and HIV, practitioners and researchers must use caution when drawing conclusions regarding childhood trauma when using the CTQ-SF. Findings also suggest that statistical inferences and implications for practice based on comparisons of observed means will be distorted and may be misleading, and as such, established cutoffs may not apply similarly for these groups, suggesting an avenue for further research.
Authors: Jutta Lindert; Ondine S von Ehrenstein; Rachel Grashow; Gilad Gal; Elmar Braehler; Marc G Weisskopf Journal: Int J Public Health Date: 2014-04 Impact factor: 3.380