An online survey was conducted from August 2020 to September 2020 via Qualtrics, targeting Sri Lankan young adults living in Australia.
To be included participants had to be aged between 18–30 years, live in Australia and self-identifying as being of Sri Lankan heritage. Participation was not possible for people who were not fluent in written English or did not have access to the internet.
Participants were recruited via two stages. Initially, a purposive sampling strategy was employed, followed by snowball sampling to allow for a large sample size to be collected.
Specifically, young Sri Lankan- Australians were contacted through four methods. First, 27 Sri Lankan Student Unions at Australian Universities across all states and territories were contacted via e-mail. Respondents were provided with the survey link to post on their social media pages. Secondly, International Student Unions of 37 Australian Universities were contacted in a similar manner and respondents were provided with the survey link. Third, the survey link was posted on 23 public Sri Lankan Facebook groups and for two closed groups, the administrative teams were contacted and provided with the survey link. Last, the research team disseminated information about the study via their professional networks including Sri Lankan and CALD specific organisations and community groups.
Data were collected through a structured anonymous online survey consisting of study specific questions and four widely used standardized short questionnaires assessing ethnic identity, attitude towards mental health problems and help-seeking behavior.
Stigma towards mental illnesses
Attitudes towards mental health problems was measured using the devaluation-discrimination subscale of the Perceived Stigma Questionnaire (PSQ) . The PSQ is a 29-item questionnaire created to measure perceived stigma of participants on four subscales. The devaluation-discrimination subscale was employed because it has been used cross-culturally to measures the extent to which participants may discriminate against persons living with mental illness . The secrecy, withdrawal and education subscales measure how individuals cope with mental illness. As this study primarily focuses on attitudes towards MHPs, these subscales were not used. Instead, only the devaluation-discrimination subscale was used in this study to measure how participants may discriminate against persons with mental illness. The wording of the original scale was modified in this study to reflect the opinion of the respondent with regards to what they would do as opposed to what most people would do. Furthermore, the original wording of ‘mental hospital’ was modified to ‘psychiatric unit’ as this is a more commonly used term in Australia. The scale has 12 items scored on a six-point Likert Scale ranging from 1 (Strongly Disagree) to 6 (Strongly Agree). Items 5, 6, 7, 9 and 11 are reversed scored therefore a higher mean scale score indicates more positive attitudes towards mental illness. A lower mean scale score indicates a higher level of stigma.
Attitudes towards seeking mental healthcare
Two standardized measures were used to examine participants’ attitudes towards seeking mental healthcare.
Section 1 of the General Help-Seeking Questionnaire (GHSQ)  was used to measure the likelihood of participants seeking help for personal & emotional problems. Section 2 of the GHSQ was excluded as it relates to participants experiencing suicidal ideations. The scale was initially validated among 218 high school students where it was established as a flexible method in measuring help-seeking intentions from a variety of sources that can be applied to a range of contexts . The 10-item questionnaire has a 7-point Likert scale ranging from 1 (Extremely Unlikely) to 7 (Extremely Likely). Items a-d are categorized as informal sources while items e–h are categorized as formal sources. Item ‘I’, ‘I would not seek help from anyone’, was measured separately and was renamed as item ‘no one’. Average scores are calculated for informal and formal sources as well as for item ‘no one’. A score greater than 3.5 indicated a higher likelihood of seeking help from the particular mental healthcare source.
The Attitudes Towards Seeking Professional Psychological Help- Short Form (ATSPPH-SF)  was also used to ascertain participants’ general attitudes toward seeking professional psychological help. This instrument is divided into two subscales to examine openness to seek treatment for emotional problems and the perceived value and need in seeking treatment. It is a widely used instrument which is supported in its validity and reliability in a study measuring mental health treatment attitudes of university students and medical patients . This measure has 10-items on a 4-point Likert Scale ranging from 0 (Disagree) to 4 (Agree). A higher subscale score is associated with a more positive attitude towards seeking professional psychological help.
Ethnic identity was measured using the Multi Ethnic Identity Measure- Revised (‘MEIM-R’) . The MEIM-R  is a revised version of the original MEIM . The scale measures the extent to which a person identifies with an ethnic group. It comprises 6-items on a 5-point Likert Scale, 3 of which measure exploration of ethnic identity and the others, commitment to ethnic identity. The scale ranges from 1 (Strongly Disagree) to 5 (Strongly Agree). Higher scores on the exploration subscale suggests greater interest in learning and participating in a given culture and its practices. Higher scores in the commitment scale suggests positive affirmations and great commitment to a particular ethnic group.
In addition, fixed-response option study specific questions were used to assess sociodemographic characteristics, the number of years lived in Australia, language spoken at home and association with Sri Lankan and other ethnic groups in the community. Further, the survey included basic questions: place of birth, parents’ place of birth and parents’ level of education completed.
Data management and analysis
The primary outcome of the study was stigma towards mental illness and its impact on help-seeking between Sri Lankan young adults born in Sri Lanka and born in Australia.
Recoding and collapsing
The initial data were recoded and collapsed to form binary variables as follows. Variables were coded as 0 and the italicized reference variables were coded as 1. Gender (other genders, female); educational status (no university education, university education); relationship status (not single, single/never married); household composition (living away from home, living at home); unemployed (no, yes); employed (no, yes); student (no, yes); I was born in (other, Sri Lanka); my father was born in; (Australia, Sri Lanka); my father’s highest level of education completed (no university education, university education); my mother was born in (Australia, Sri Lanka); my mother’s highest level of education completed (no university education, university education); language spoken at home (mostly English, Equally a Sri Lankan Language and English, mostly a Sri Lankan language); association in the community (mostly other ethnic groups, mostly Sri Lankan groups).
Furthermore, the ATSPPH-SF was recoded following the original instruments coding. 0 = disagree, 1 = partly disagree, 2 = partly agree, 3 = agree.
Statistical analyses were conducted in three stages:
- 1.Mean scale and subscale scores were calculated for all standardized data tools using the published protocols. Descriptive statistics were calculated for all sociodemographic variables. Chi-squared tests were used to compare significance between participants’ country of birth and categorical demographic variables such as: gender, educational status, relationship status, household composition, employment status, student status, parental country of birth and education, language spoken at home and association in the community. Independent t-tests for continuous variables such as: age, number of years spent in Australia and MEIM-R score, were used to assess significance between these variables and participants’ country of birth.
- 2.Bivariable analyses were used to examine whether the dependent variable could be explained by the independent variables. In this case, whether country of birth had an effect on sociodemographic characteristics. A bivariate analysis was also conducted on country of birth and mean scale scores of all standardized data tools (PSQ, GHSQ, ATPPHS) to explore whether country of birth influenced attitudes towards MHPs and accessing mental healthcare. ANOVA was used for multiple group mean comparisons. All bivariable analyses included Sri Lankan young adults born in Sri Lanka, born in Australia and those born in other countries who identified as being of Sri Lankan heritage.
- 3.Multivariable analyses were used to investigate multifactorial determinants and mean scale scores of the PSQ, GHSQ and ATSPPH. For this, the ‘born in other countries’, respondents were excluded from the dataset as the analysis determines the effect of sociodemographic factors on the association between being born in Australia/Sri Lanka and mental illness stigma and help-seeking. The binary forms of the variables were recoded to Sri Lanka, Other where ‘Sri Lanka’ indicates participants born in Sri Lanka and ‘Other’ indicates participants born in Australia. The italicized reference variable was coded as 1.
Informed consent was sought from all participants. Participation in the survey was voluntary and participants were informed that consent would be assumed by completing the survey. The study protocol was approved by the Monash University Human Research Ethics Committee (MUHREC 24878).