Psychometric Properties of the Thai Mental Health Literacy Scale in Sixth-Year Medical Students

Objective: To assess the psychometric properties of the Thai Mental Health Literacy Scale (TMHLS) in sixth-year medical students. Materials and Methods: By using the purposive sampling method, we enrolled 202 participants in this study. Descriptive statistics were used to analyze demographic data. The index of item-objective congruence (IOC) was used to verify content validity. Exploratory factor analysis (EFA) was performed to establish the construct validity of the TMHLS. The internal consistency was estimated by computing Cronbach’s coefficient alpha. Results: The TMHLS had good content validity (IOC=.85) and construct validity. The EFA resulted in five factors, which included 32 of the 35 items and accounted for 46.86% of the variance. The factors were the ability to recognize mental disorders; confidentiality of mental health practitioners; skills of mental health information seeking; beliefs about mental illnesses; and attitudes toward patients with mental illness. The reliability coefficient of the TMHLS total test was .851, and reliability coefficient in subdomains were range from .197 to .872. Individuals who had a mental health professional as an intimate contact and individuals who had a history of seeking help from mental health professional(s) in person showed significantly higher mental health literacy than those who did not. Conclusions: The TMHLS has good psychometric properties. Dynamic knowledge transfer and exchange with a close mental health professional should be applied to promote mental health literacy in medical students.


INTRODUCTION
Mental health problems have been increasing throughout the world 1 , with young adults being the most affected group. Thirty percent of them have mental disorders while the remaining are also at risk. 2 Because of poor mental health literacy, high mental health problems and low engagement in help-seeking behaviors were reported in these individuals. [3][4][5][6][7] Mental health literacy reduces the risk of developing mental disorders along with increasing help-seeking behaviors. 8 People with high mental health literacy will be able to recognize, manage, and prevent mental health problems. Oppositely, people with low mental health literacy may not be able to appropriately manage and often end up with more serious complications. 9 Unfortunately, there is no assessment tool for mental health literacy in Thai at the time.
Sixth-year medical students were targeted in this study because they were young adults at risk of mental disorders 2,10-11 who already gained mental experiences that may affect their mental health literacy. [12][13] Due to the lack of an instrument to measure mental health literacy among Thai people, this study aimed to assess the psychometric properties of the Thai mental Original Article SMJ health literacy scale (TMHLS) in sixth-year medical students who may exemplify the young adults at risk of mental disorder.

Participants
The number of participants in this study was determined by the Cochran formula. 14 We enrolled 250 sixth-year medical students from the Faculty of Medicine Siriraj Hospital in Bangkok who had registered for the first semester in academic year 2017 and voluntarily answered the questionnaires using purposive sampling method.

Tools
A demographic questionnaire was used to collect data from participants including gender, age, sources of mental health experiences, and their mental illness if applicable.

The translation of mental health literacy scale (MHLS)
The MHLS was translated to Thai under the supervision of a language expert. The index of itemobjective congruence (IOC) was used to verify content validity by three mental health experts: one psychiatrist and one licensed clinical psychologist from the Department of Psychiatry, Faculty of Medicine Siriraj Hospital; and one licensed clinical psychologist from the Faculty of Psychology, Chulalongkorn University. All mental health experts discussed the translated version until reaching a consensus. The Thai mental health literacy scale (TMHLS) was finally completed following expert opinion.
The TMHLS is a self-reporting questionnaire with 35 items covering six attributes of mental health literacy: the ability to recognize a disorder; knowledge of where to seek information; knowledge of risk factors and causes; knowledge of self-treatment; knowledge of professional help available and attitudes that promote recognition or appropriate help-seeking behavior. The total score is the summation of all items. Therefore, the maximum score is 160 whereas the minimum score is 35. A higher score means greater mental health literacy.

Statistical analyses
All statistical analyses were performed by PASW 18.0. 16 Descriptive statistics were used to analyze demographic data. The IOC was used to verify content validity. The factor solution was determined based on the number of eigenvalues greater than one. 17 We conducted the exploratory factor analysis (EFA) using .30 as a factor loading criterion 18 , five to ten participants per item 19 , and a minimum sample size of 200. [20][21] The EFA began with an initial analysis run to obtain eigenvalues for each factor in the data. The Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy test and Bartlett's Test of Sphericity were executed to determine construct validity and to confirm those data were appropriate. The KMO test was used to verify the sampling adequacy for the analysis, and Bartlett's Test of Sphericity was used to determine if correlations between items were sufficiently large for EFA. Bartlett's Test of Sphericity should reach a statistical significance of less than .05 in order to conduct an EFA. The reliability of an instrument is concerned with the consistency, stability, and dependability of the scores. 22 For this reason, the internal consistency was tested using Cronbach's alpha for each competency.

The sixth-year medical students
Two-hundred and two of the 250 participants (80.8%) answered the questionnaires. The majority of respondents were female (n=133; 65.8%) aged between 22-24 years (M = 23, SD = 0.46). Psychiatric rotation was the most popular source of their mental health experience (n=190; 94.1%). Thirteen out of 202 medical students had major depressive disorder (6.4%), the most common diagnoses among the samples ( Table 1).

Reliability
Total Cronbach's alpha coefficient of the TMHLS   Table 3). The reliability coefficient in subdomains of TMHLS were range from .197 to .872 (Table 4).

The mental health literacy in sixth-year medical students
The medical students' mean score of mental health literacy was 123.09 (S.D. ± 11.55, 95% CI = 121.49-124.69). Multiple comparisons of our participants' mental health experiences showed having intimate contact with a mental

The sixth-year medical students
Major depressive disorder was the most common diagnosis in this study which was in accordance with previous Thai, Malaysian and Chinese Studies. [23][24][25] The Psychometric properties of the Thai mental health literacy scale (TMHLS) The TMHLS has good validity. The content validity by the IOC in the second-round was .85, and only 4 out of 9 items needed to be revised. According to the original study 15 that stated measurement cannot assess all attributes of mental health literacy when some of the items needed to be removed, all items were used in the scale altogether. Consistent with a previous Persian study 26 , the EFA of data resulted in five meaningful factors that were similar to the original ones 15 , and accounted for 46.86% of the variance. The trivial differences could have been due to cultural diversities of the participants. Socioeconomic status, cultural and language variances interact with health literacy. 27 Total Cronbach's alpha coefficient of the TMHLS was .851 which was considered in a good criterion. The reliability coefficient in subdomains were range from .197 to .872. Still, there were 6 items that did not meet the criterion. The Cronbach's alpha if item deleted for each item was slightly different from the Cronbach's alpha of all items. According to the original study 15 that stated the measurement cannot assess all attributes of mental health literacy when some of the items needed to be removed. Therefore, those 6 items that do not meet the criterion were persevered.

The mental health literacy in sixth-year medical students
The mental health literacy of our medical students was aligned but slightly lower than a prior British study. 13 Our score was marginally inferior than an Australian study exploring university students. 15 This may uncover differences in mental health literacy between developing and developed countries. The necessity of mental health literacy acknowledgement in village health workers was mentioned in a previous Thai study. 28 A South African study urged for mental health education in healthcare professionals. 29 Language deviance and questionnaire format may also be responsible for the different results.
Our participants had already gained mental health experiences that may affect their mental health literacy. Previous works also showed higher mental health literacy in individuals who encountered mental health problems than the individuals who did not. 12,30 The more exposure someone has, the more mentally health literate they are. 12 Consistent with the original study 15 , the mental health literacy of individuals who had a history of seeking help from mental health professional(s) in person was higher than those who did not. Dynamic knowledge transfer and exchange with a close mental health professional, like in family businesses 31 , could be a reason for higher mental health literacy of individuals who had a mental health professional as an intimate partner than those who did not.

The questionnaire comments
The main concern about the TMHLS was the complexity and clarity of the questions. However, the items that should be allocated were not mentioned. A separate version of TMHLS between medical students and general population was advised. Although some participants described the questionnaire as easy and clear to answer, an equal number expressed the overly theoretical concerns. Some of them requested more attitude questions.

Limitations
Information and recall bias may have been presented in this observational descriptive cross-sectional study. Based on purposive sampling method, the results cannot legitimize any generalizations. We did not perform back-translation process; hence the quality assurance of the TMHLS should be concerned. As the EFA is not a sufficient tool to test the theoretical foundations of the instrument, a confirmatory factor analysis (CFA) should be conducted to further the knowledge in this area. Since we used Cronbach's alpha for reliability testing, the interitem covariance and the measurement assumptions error could be considered as the alpha value cannot be equivalent with the reliability of the test score. Additional studies in other population are recommended to validate this instrument to widen its application.

CONCLUSION
The TMHLS has good validity and reliability. Dynamic knowledge transfer and exchange with a close mental health professional should be applied to promote mental health literacy in medical students.

ACKNOWLEDGEMENT
We gratefully acknowledge Matt O'Connor, Ph.D. for his kind permission to allow us to use MHLS in this work; the Postgraduate Education Division and Deputy Dean of Undergraduate Education of Faculty of Medicine Siriraj Hospital for the scholarship and permission to collect data, correspondingly. We also Original Article SMJ thank all expert validators, Boonjira Thungsuk, M.A. (English); Tikumporn Hosiri, MD; Panida Yomaboot, Ph.D. and Kullaya Pisitsungkagarn, Ph.D. for their superb suggestions about the TMHLS. And last but not least, we thank all the participants in this study for their contribution.