Ethical considerations and participant selection
The present study followed the British Psychological Society’s code of ethics and ethical standards to ensure participant protection from all foreseeable psychological risks: distress and fears due to questionnaire items were minimized by conveying that there was counseling and debriefing if and when needed. To allay their feelings of deception and mistrust, student participants were ensured that confidentiality and anonymity was never breached; and they had the freedom to withdraw from the study at any time without penalty, and their data destroyed. Student participants were told about the aims and procedures of the study before actual commencement of data collection. More importantly, to dispel any qualms over stigmatization they were informed that the study results would be disseminated and published in an international journal. This required informed consent: none of the student participants dropped out of the study because of this information.
Subsequent to ethical approval by the AUW ethics committee in February 2010, Darfuri student lists were made available from the registrar’s office, which contained contact information including year group, school allocation, class locations and course timetables. At the time of study, there were a total of 209 registered students originating from the state of Darfur. In accordance with research objectives, a set of inclusion criteria were established: Darfuri born and raised; all members of her extended family must be of Darfuri origin; parents, siblings and extended family must still reside in Darfur; she had been in Darfur up to at least 2003 (war broke out in 2003; massive destruction and displacement followed); and she had no prior visits to Omdurman (her first visit to Omdurman was to continue her education). In total, 176 Darfuri students fit the inclusion criteria. Twenty of them already served in a previously conducted exploratory study [
10] and fifteen participated in the pre-testing of the measurement. This resulted in 141 Darfuri students being invited to participate in this study. A total of 123 (88 % response rate) Darfuri female students participated in the current study with 18 (12 % non-response rate) potential participants unavailable at the allocated time because of conflict with their study schedules.
Research team and data collection procedures
The research team was made up of eleven AUW staff members and divided into two sub-teams; six field team members and five members on the instrument translation and adaptation team. Two senior members of staff served in a supervisory capacity overseeing the data collection procedure, providing debriefing sessions, and time-scheduling logistics, while the principal researcher and three staff members were responsible for the actual data collection and meetings with the student participants. The translation and adaptation team included an in-house counsellor, a clinical psychologist from the school of psychology, two bilingual translators and the bilingual/bicultural principal researcher. Initial meetings with all members of the research team began on February 4 2010 and lasted for three weeks. The aim of these meetings was to introduce the research objectives, target population, rationale and history of the questionnaire, give detailed discussions of each item in the questionnaire, and plan for instrument translation, adaptation, pre-testing and data collection within the allocated timeframe, which was scheduled to start in March of that year. Field-team meetings focused on establishing a contextual framework within which the data collectors were trained, which also included increasing an awareness of the ethno-graphic qualities and experiences of the target population, being attentive and vigilant for signs of distress among susceptible Darfuri students, and keeping in mind the sensitive nature of some questions. Each data collector was assigned a school and given a copy of the registrar’s list of student names. Introductory meetings with the students were made to ascertain availability and willingness to participate and to then agree upon exact dates, times and places for subsequent questionnaire administration. Thereafter s/he would meet the students, distribute the questionnaires and provide clarification if needed, and collect the finished questionnaires. Students were again informed of the voluntary nature of their participation, were assured confidentiality and were requested to sign a standard informed consent sheet on the covering page of the booklet. Depending on the actual volume of Darfuri students within a particular school, on average there was a show of 7 students per session, with each session taking approximately 45 minutes to an hour to complete. Data collection concluded on April 17 2010.
Instruments
The questionnaire booklet was comprised of two sections. An ethno-socio-demographic section in which data were collected on participants characteristics, such as place and date of birth, familial geographical origin, name of village and location to nearest town, tribe name, parental main source of livelihood, displacement characteristics (for example, property loss, parental loss of livelihoods, and date since displacement), and date of arrival in Omdurman city.
The Harvard Trauma Questionnaire (HTQ) [
11] made up the second section of the questionnaire booklet. A widely used assessment instrument, it contains characteristics of cross-cultural adaptability which reflects context specific descriptions of war-related traumatic events of the community in which it is administered [
12,
13]. Furthermore, it is a comprehensive screening assessment of PTSD where culturally relevant expressions of PTSD symptoms are accurately measured [
14‐
16]. For the purposes of the present study, the self-report Parts I and IV of the HTQ were adapted and analyzed to become the Darfuri versions of the HTQ.
The HTQ has been validated and tested for reliability within the Sudanese context [
6]. However, as stressed by Mollica and colleagues [
12], each new war-affected population needs to develop a different HTQ. The adaptation of Part I of the HTQ began by undertaking in-depth interviews with twenty AUW Darfuri female undergraduate students in December 2009. A core checklist of 20 context specific war-related experiences was compiled from this reference group of students, which accurately described common acts of violence, combat situations and other war-related traumatic events. These events were: ‘forced to hide’, ‘constant insecurity’, ‘movement in search of shelter’, ‘forced separation from family’, ‘distance from family’, ‘fleeing/running away’, ‘innocent victims being killed’, ‘finding dead bodies and burying them’, ‘living in displaced camps’, ‘burning of home’, ‘not having a home’, ‘owning nothing’, ‘complete change of life’, ‘aerial bombardment’, ‘hearing of atrocities’, ‘loss of health’, ‘increased physical ailments’, ‘suddenly gone missing’, ‘parentless children/teenagers’, and ‘seeing dead bodies’ [
10]. A further twenty items were selected from two existing versions of the HTQ: Indochinese and the Japanese (Mollica et al.., [
12]. These items were added because they were cited by the reference group of Darfuri students and reflected important contextual details. Based on Flaherty [
17] a five step validation process for cross-cultural equivalence was administered to the resultant forty item section, including pre-testing among a group of fifteen AUW Darfuri students not in the original sample of this study but representing similar ethno-demographic characteristics, and matched for age, ethnicity, and locale. No item was misunderstood or needed further restructuring. A total of 40 war-related trauma items (20 Darfuri items added onto the existing 20 items) made up Part I of the Darfuri HTQ version.
In accordance with the guidelines set forth by the HTQ, participants were presented with this list of forty potentially traumatic war-related events and were requested to endorse individual events according to four options (E = experienced, W = witnessed, H = heard about, or N = "no"). Fox and Tang [
18] noted that the ‘experience’ option reflected a person’s direct exposure to war-related trauma. However, greater trauma severity can also be highlighted by proximity to the event as is represented by the ‘witnessed’ option. Therefore, the present study focused on these two experiential levels of the HTQ, i.e., ‘experienced’ and ‘witnessed’, to reflect the intensity and severity of war-related traumatic events.
Part IV of the HTQ is a screening checklist for post-traumatic stress disorder (PTSD) symptoms. The first 16 items measure PTSD according to the DSM-IV criteria [
19]; these remained the same. Only one of the two dissociative symptom items was omitted: ‘feeling as if you are split into two and one of you is watching what the other is doing’ because this was misunderstood repeatedly by the pre-test group and was not endorsed by the reference group [
10]. The other dissociative symptom item; ‘finding out or being told that you have done something that you cannot remember’, remained. This made a total of 17 PTSD items based on the DSM-IV criteria.
In addition, Part IV focused on formulating a contextually relevant refugee-specific symptom checklist by incorporating culturally identifiable idioms of distress [
14,
15]. Based on the interviews previously conducted with the reference group, twenty-two most commonly mentioned, culturally and contextually verifiable expressions were identified. These included: ‘feeling miserable’, ‘feeling hopeless’, ‘only feeling normal among other Darfuris who also suffered the same sort of trauma’, ‘feeling no trust in others’, ‘feeling humiliated by your experiences’ and ‘spending time thinking about why these events happened to you’. An additional seven items were derived from the Japanese version of the refugee-specific symptoms checklist, which included: ‘feeling as if you are going crazy’, ‘feeling isolated because of loss of relationships with family’, ‘feeling overly dependent on others’, and ‘feeling discriminated against because of loss of home’. Part IV of the HTQ was also pre-tested by the same group as before, to ensure that the concepts reflected participants’ understanding. None of the pre-testers suggested that any modifications were required. Participants were asked to mark 1 for ‘not at all’, 2 ‘a little’, 3 ‘quite a bit’, and 4 ‘extremely’ for the resultant number of 46 items on the Darfuri version of Part IV of the HTQ.
Based on the guidelines set forth by Flaherty et al.. [
17] and Guillemin et al.. [
20] the translation and adaptation team undertook the task of translating the HTQ Darfuri version from English to classical Arabic (as opposed to colloquial Sudanese Arabic, as not all Darfuris speak the same dialect while all students are educated and can efficiently read and understand classical Arabic). Semantic equivalence was established using a bilingual probe technique to verify linguistic accuracy and meaning to the original English text. Back-translation of the Arabic version showed high concordance with the original English text when pre-tested: modifications were unnecessary.
Data analysis
All data were analysed using the Statistical Package for the Social Sciences (version 17.0 for Windows). Means and standard deviations have been used to illustrate the severity of war-related exposure dimensions and PTSD sub-scale symptoms. Pearson’s correlations were conducted to test for the strength of association between dimensions of war-related trauma exposures and the four sub-scales of PTSD symptoms. The significance level was set at p < 0.01 and two-tailed test values were reported. Strengths of association between variables, where Pearson’s correlation coefficients are calculated to represent: .10 = weak; .30 = moderate; .50 = strong [
21].
To highlight the experiential level for each of the 40 items of war-related traumatic events the ‘experienced’ and ‘witnessed’ options were re-coded into one value: ‘1’ indicating the experience and/or witness of trauma; and ‘0’ indicating heard about or no trauma experience. A sum score was created for all 40 items known as the experienced and\or witnessed cumulative dimension ‘EW’. To assess the underlying factorial structure of war-related exposures the responses were then subjected to Principal Component analysis extraction. After Varimax factor rotation, a four factor model, known as war exposures dimensions, was chosen based on the Scree test and Eigen values >1. The first dimension, combat situations, included 13-items (α = .91), the second dimension included 7-items (α = .84) referring to the experience of material loss, the third dimension with 12-items (α = .89) was concerned with the loss of relatives through killing, kidnapping or forced separation, and the fourth dimension with 8-items (α = .88) referred to forced displacement and related camp life issues.
Scoring of PTSD symptoms was according to the original instrument guidelines and standards [
11]. Mean scores were calculated to establish ‘checklist positive’ symptoms for PTSD with a critical cut-off score at >2.0 [
6]. Internal consistency was calculated from the mean scores for the complete scale of PTSD symptoms (α = .93), and reliability within each sub-scale also showed a high internal consistency: re-experiencing sub-scale (α = .70), avoidance/numbing sub-scale (α = .73), and psychological arousal sub-scales (α = .81), and the refugee-specific self perception of psychosocial functioning sub-scale (α = .91).