Sample: This was a cross-sectional study of 420 Syrian refugees living in Ankara, using face to face interviews administered at the respondents’ home by trained, Arabic-speaking interviewers. The interviews were conducted in October–November 2016 in two neighborhoods of Ankara, populated mostly by Syrian refugees.
Although our measures were developed to be filled-in by the respondent (i.e. self-report), the interviewers were instructed to read out the items and record the responses. This was necessary, since a large majority of the respondents had either no formal education or minimal (i.e. primary school) education.
Demographics: This section was developed by the researchers for the purposes of the current study. It included items assessing sociodemographic characteristics of the participants, their perceived physical and mental health status, utilization of and access to general and mental health services, and factors preventing access to those services. Social network variables such as having Turkish friends and speaking the language (i.e., Turkish). The services use section was quite detailed and inquired about the presence of physical or mental conditions diagnosed by a doctor, subjective assessment of physical and mental status by the respondent, alcohol/drug use history, current psychotropic use, contact with health and mental health services in the past year, the subjective need to contact mental health services, and finally reasons for non-contact, if the respondent reported need for contact and did not contact.
Harvard Trauma Questionnaire (HTQ): The HTQ is a widely-used self-report measure of traumatic stress . We used the Arabic version that assesses DSM-IV PTSD symptoms . The participant is asked to rate each item on a four-point Likert scale (1=not at all, 2=a little, 3=quite a bit, and 4=extremely). The first 16 items such as “recurrent thoughts or memories of the most hurtful or terrifying events” or “feeling as though the event is happening again” assesses the presence of probable PTSD. The validation study of the original version was conducted with 91 patients originally from Cambodia, Laos, and Vietnam, using a semi-structured interview. Probable PTSD is computed following the algorithm suggested by the authors, since there is no established cutoff for the Arabic version. Cronbach’s alpha for the 16 PTSD symptoms in the current study was 0.82.
Beck Depression Inventory (BDI): This widely used, 21-item, self-report questionnaire measures depressive symptomatology for the last week . All items are coded between 0 and 3; higher scores suggest more severe depression. Validity and reliability in Arabic were established by West  and Abdel-Khalek . In the present study, both the BDI total score (range: 0-63), and the probable depression score (using a cutoff of 19) were used in analyses. Cronbach’s alpha for the current sample was 0.85.
Variable selection and transformations:
We used both categorical and continuous variables in reporting our results. We recoded some variables before analyses. For example, marital status (married, single, divorced) was recoded as married/not married. Command of Turkish language, which was coded on a 5-point scale (0=none, to 4 =very good), was recoded as 0=no, 1=yes, where yes=good, very good. Education (0=no education, 4=university graduate) and effect of symptoms on social functioning (1=none, 4=very much) were used as continuous variables. Since BDI and HTQ scores were highly correlated, we decided to include HTQ only as a measure of current psychopathology in our predictor analyses.
Procedure: Convenience sampling method was used to recruit participants into the study. The houses in the two neighborhoods, populated mostly by Syrian refugees, were visited by our six interviewers who were fluent in Turkish and Arabic. Each household was visited by at least two interviewers. Of the households visited, 15 declined to participate and so were not included in the study. Figure 1 shows the recruitment of participants to the study. SPSS 23.0 statistical software package was used for data analysis.