CHAPTER FOUR METHODOLOGY
Study Site and Sampling
The study was carried out in the capital city of Ethiopia, Addis Ababa, where high HIV prevalence rates (5.2%) have been estimated in the Ethiopia Demographic and Health Survey report, compared to other regions in the country (Central Statistical Agency of Ethiopia and ICF International, 2012). In 2012, the Ethiopian Health and Nutrition Research Institute (EHNRI) had estimated that there were 50,000-250,000 female sex workers working and living in Addis Ababa. The EHNRI (2012) had further estimated that 35% females in the reproductive age of 15 – 49 years have involved in some form of sex work at some point in their lives in Addis Ababa at the time of the study. Considering the growing numbers of such workers and types of sex work involved in the city, it was recommended that this be used with caution as regards generalisation. Another recent study undertaken by DKT Ethiopia (2012) has indicated that number of female sex workers working in Addis Ababa is about ten times bigger than the size estimation of sex workers residing in seventeen major cities/towns of the country where sex workers were residing. Furthermore, Addis Ababa is considered to be a representative stratum of the study population, for the capital city is much larger than other cities in the country, incorporating all types and categories of sex workers who have migrated from all corners of Ethiopia, as it is located at the centre of the country (EHNRI, 2012).
Five female sex workers’ associations were used as cluster groups as an entry point from which participants were selected. The associations were established by a national level project called Wise Up that focuses on sex workers, through the financial and technical assistance of DKT-Ethiopia in collaboration with FHAPCO. These associations were located in major areas where large populations of female sex workers were residing, and they were also distributed in the ten sub-cities of Addis Ababa (one association located more-or-less adjacent to two sub-cities). The guidance and counselling rooms of these associations were used to undertake the interviews as they were confidential and private settings.
The researcher has been actively involved in the Wise Up project since its inception in 2007. He is one of the primary technical experts involved in the designing and appraisal of the project. Major activities undertaken by the researcher include, inter alia,: a programme document for the establishment, at national level, of these drop-in centres for sex workers’ associations (Wise Up, 2009) and a rapid assessment on condom sells and promotion as an alternative income generating scheme for female sex workers (Wise Up, 2010). The researcher has also made major contributions in the campaigning for continuous inculcation of 100% condom use among female sex workers, by tailoring various information, education and communication (IEC) materials and by providing behavioural change communication (BCC) training on safer sex to many female sex workers. Using his recognition in the establishment of these associations and his acceptance among female sex workers, the researcher has played major roles in arranging the study sites in these associations, in recruiting and training data collectors, and in closely supervising the sampling and data collection processes.
A Respondent-Driven Sampling (RDS) method was used to recruit participants. RDS was an adaptation of chain-referral sampling and a suitable sampling method for hidden populations, which provided relatively unbiased and representative population-based estimates. In this sampling method, participants were asked to bring their peers to participate in the study. However, not all who participated in the study were willing or interested to bring their peers for the study. Those participants who did bring additional peers to the study were known as seeds, while the others were delineated simply as participants (Douglas, 2002).
The selection criteria for eligible seeds and participants followed the recommendation enlisted in the EHNRI (2012) formative assessment report: a female over the age of 18 who has exchanged sex for money; who had lived in Addis Ababa in the past three months; who was willing to participate in the study and who was sober at the time of the interview.
Five seeds were initially selected from members of sex workers’ associations found in the city. One seed was selected from each association. Each selected seed was briefed on the purpose, approaches and processes of the study and was informed on how to refer other eligible female sex workers using coded invitation cards designed as coupons. Each recruited and willing seed was asked to recruit three peers so that they also participated in the study and was given three uniquely coded coupons to refer her peers to the study. Coupons were given to participants until a target sample size of 350 were recruited.
According to the EHNRI’s (2012) formative assessment the prevalence of female sex workers who were in the reproductive age of 15 – 49 years was 35% which implied the expected proportion (P) value to be 0.35. Using the Without Finite Population Correction of the Sample Size Estimation Formula of Naing, Winn and Rusli (2006) the sample size of 350 was generated with level of confidence (Z) of 95% and precision (d) of 0.05%, respectively.
In line with the EHNRI’s (2012) formative assessment recommendations in administrating RDS, seeds and their recruits were given a transportation and incentive package, including Ethiopian Birr 150 (U.S. $ 8.33/ZAR 68.8), HIV prevention pamphlets and four boxes of condoms for their time spent in the study. Furthermore, an additional Ethiopian Birr 20 (U.S. $ 1.11/ZAR 9.17) was provided to each selected participant when recruiting and referring a peer to participate in the study.
Instrumentation and Processes
A structured interview questionnaire that gauged data on basic demographics and constructs of the IMB model that reflected the eight determinants of consistent condom use and the EET personality assessment tool by Daniels and Price (2009) was used to collect data from individuals. The structured questionnaire and the personality measurement had been translated to the local main language (Amharic) and retranslated back to English to minimise losses in translation. After undergoing refinements, the instruments was translated back to Amharic and was pilot tested on 10 female sex workers and then revised accordingly, before administering it on a large scale. Although both instruments were consecutively presented to participants with a single five minute break in-between, all questions that were modified were from the background and IMB questionnaire after the pilot test.
On the first part of the questionnaire, background information question item 02, the lists for this question were increased from seven to twelve to differentiate the currently attending and dropouts in the different stages of schooling. Similarly on background information, question item 05, one item called cohabiting adults (personal sexual partner) was included for it was found to be important to see their relationship in the realm of condom use. Moreover on background information, question item 06 asked about types of sex work they are currently engaged. During the pilot testing it was learnt that a sex worker might be engaged in different forms of sex work at a time or may change also seasonally (during fasting season, tourist seasons etc.). So the question was adjusted to “Mostly, where do you access clients?” to see the common type of sex work that the participant is engaged in.
On section 2, regarding questions on HIV information, especially in items 2, 3, 8 and 13 the phrase without condom was included for more clarity of the questions. Similarly, in item 7 of the HIV information questions “if a partner has HIV” was standardized as “a partner living with HIV” for being culturally insensitive.
For section 4, 5, 6, and 8 of the IMB constructs, the Likert’s scale response items were presented in ‘1 = Extremely low; 5= Extremely high’. However, it was not affable for use for both the interviewer and interviewee thus it was listed in detail in a friendly manner to secure correct information: 1 = very low; 2 = low; 3 = average; 4 = high; 5 = very high)….And the same has been done for others respectively. In section 4 and 5 (items on social referent groups) question items that ask about establishment owners were found to be irrelevant for street based and phone based sex workers for they do not function in establishments. Similarly, in section 7, item 5 “to keep the sperm inside the condom” was included for better clarification of the question.
In section 9, Moreover, the translation version of the Amharic questionnaire was corrected for saying how many times did you drink alcohol in the past week’ to ‘How many times have you been drunk in the past week’ for drinking one bottle of beer, or a glass of wine does not necessarily indicate being drank. In the pilot test it was confirmed that the most common substances used by female sex workers were alcohol, Khat and Cannabis, thus were specifically listed accordingly.
Structured interview questionnaire
A structured coded interview questionnaire with general instructions for the interviewer, information regarding the interviewing and supervision processes and queries of basic socio-demographics of the interviewee was prepared along with the constructs of the IMB model. The socio-demographic data of the interviewee was included in the questionnaire in order to determine the background of the participants. This data included age, place of origin, religion, education, income level, marital status and type and years of sex work in which they were engaged. The individual level determinants for consistent condom use that were assessed in the interview questionnaire were information on HIV preventions, perceptions of social referent support, motivation to comply with social referent support group, perceptions of personal vulnerability, attitude towards condom use, condom use skills, condom negotiation skills, and substance use. The measurement approaches of each determinant are presented below.
Information on HIV
Using the eighteen items (scaled 0-1) questionnaire of Volpe, Nelson, Kraus and Morrison’s (2007) which they adapted and refined for females from the original HIV knowledge questionnaire of Carey and Schroder (2002), participants were asked about their knowledge relating to HIV. Eleven items are related to HIV transmission and seven to HIV prevention. Participants were asked to state whether the statements are true or false, or unknown. ‘Do not know’ responses were considered as incorrect answers. The composite score of this assessment ranged from 0 to 18. The number of questions correctly answered indicated the extent of knowledge of the participant regarding HIV/AIDS transmission and prevention. Cronbach’s Alpha coefficients for the eleven HIV transmission related items and seven HIV prevention related items were indicated to be 0.78, and 0.81, respectively (Zhang et al., 2011). Similarly, in this study the Cronbach’s Alpha coefficients for the eleven HIV transmission related items and seven HIV prevention related items were found to be 0.78, and 0.92, respectively.
Attitudes towards Condoms
Experiences with and attitudes toward condoms were assessed with seven true or false questions. The typical items were: ‘Condom use causes too much trouble’; ‘The westerners have deliberately injected HIV viruses into condoms that they export to developing countries’; ‘Condoms that are supplied freely have HIV virus injected in them’; ‘Some condoms protect from HIV more than others’; ‘Using condoms with regular client is suggestive of filth and disease’; ‘Using condom with personal sexual partner indicated infidelity or mistrust’, and; ‘Condom use is safer than other methods’. The number of answers indicating a positive attitude served as an indicator of experiences with and attitudes toward condoms. ‘Do not know’ responses were considered as incorrect answers. The composite score ranges from 0 to 7. The summation of the rating was used to describe the level of attitude towards condoms. These statements were used by Yang et al. (2005), and Zhang et al., (2011) and both indicated that the 7 items had a Cronbach’s alpha coefficient of 0.80. In this study the Cronbach’s Alpha coefficients for the seven items of experiences with and attitudes toward condoms were found to be 0.78.
Perceived Social Referents Support
Perception of the social referents’ support was assessed with six items used by Yang et al. (2005), Kerrigan, Moreno, Rosario, Gomez, Jerez, Barrington, Weiss and Sweat, (2006), Wang, Li, McGuire, Kamali, Fang and Stanton (2009), and Zhang et al. (2011). In the current study, peers, employers and clients were regarded as social referents. Typical items assessed the participants’ perceptions regarding the approval levels of their employers, peers, clients and others in the vicinity concerning consistent condom use. Each item was scored 1 (disapprove strongly) to 5 (approve strongly). The mean scores of the responses were indicated as an illustration of the participant’s perception of social referents support. Cronbach’s alpha coefficient was indicated to range from 0.75 to 0.83 for the six items. In this study however the Cronbach’s Alpha coefficients for the six items of perceived social referents support were found to be 0.73.
Motivation to Comply
Motivation to comply was measured by three items, indicating the importance of the approval of peers/employers/clients in using condoms. Responses ranged from 1 (extremely unimportant) to 5 (extremely important). The composite mean score indicated the level of motivation to comply with social referent groups. The three items were indicated to have a Cronbach’s alpha coefficient of 0.68 (Zhang et al.,, 2011). In this study the Cronbach’s Alpha coefficients for the three items of motivation to comply were found to be 0.65.
Perceived Personal Vulnerability
Perceived personal risks in the questionnaire were measured by two items that queried the extent to which the participant was worried about contracting HIV, on a scale ranging from 1 (not at all worried) to 5 (extremely worried); and her risk of actually contacting HIV on a scale ranging from 1 (very low) to 5 (very high). The composite mean score indicated the level of perceived susceptibility. These items were indicated to have a Cronbach’s alpha coefficient of 0.66 (Morisky, Stein & Chi, 2006). In this study the Cronbach’s Alpha coefficients for the two items were found to be 0.51.
Condom use Skills
To assess condom use skills, a condom was provided to each participant and they were asked to indicate, using a wooden penis model, how they were using condoms from start to finish. The important steps that needed to be performed for appropriate condom use were assessed based on the work of Yang et al., (2005); Stanton, Deveaux, Lunn, Yu, Brathwaite, Li, Cottrell, Harris, Clemens and Marshall, (2009); Wang et al.,, (2009) and Zhang et al., (2011). A checklist measuring these steps was used to evaluate the condom skills of participants when they were performing their habitual condom use steps on the model-penis. The most important items began with checking the expiration date and ended with disposing of the tied condom by putting it in the condom cover. The composite score ranged from 0 to 6. The scores were used to describe the level of condom use skill of the participant. In this study the Cronbach’s Alpha coefficients for the 6 items of condom use skills were found to be 0.45.
CHAPTER ONE INTRODUCTION
1.2 Research Aim
1.3 Research Objectives
1.4 Research Questions
1.5 Chapter Contents
CHAPTER TWO PERSONALITY THEORIES AND STUDIES.
2.2 Psychodynamic Theories
2.3 Behavioural Theories
2.4 Humanist Theories
2.5 Biological Theories
2.6 Trait Theories
CHAPTER THREE THE ENNEAGRAM THEORY OF PERSONALITY
3.1 Historical Background
3.2 Contributors to the system
3.3 Structure of the Enneagram
3.4 The Dynamics of the Personality Types
3.5 Enneagram Personality Types
3.6 Empirical Research on the Enneagram
CHAPTER FOUR METHODOLOGY
4.1 Study Site and Sampling
4.2 Instrumentation and Processes
4.3 Data Collection and Sample Size
4.4 Data Analysis and Quality Assurance
4.5 Informed Consent and Confidentiality
4.6 Limitations and Delimitations
4.7 Significance of the Study
CHAPTER 5 RESULTS
5.2 Socio-Demographic Background
5.3 Relationship Analysis
CHAPTER SIX Discussion and Conclusion
GET THE COMPLETE PROJECT
The Enneagram: Predicting Consistent Condom Use among Female Sex Workers