Project Overview
It is known from previous studies that university students in sub-Saharan Africa engage in sexual risk-taking behaviour. However, much less is known on correlates of sexual risk-taking behaviour among university students (emerging adults) at the Kenyan Coast, thus hindering intervention planning. This study seeks to provide an in-depth qualitative understanding of correlates of sexual risk-taking behaviour and their interconnectedness among university students at the Kenyan coast, combining qualitative research with a systems thinking approach. Using the ecological model, and employing in-depth interviews, we explored the perceptions of twenty-one emerging adults and five other stakeholders on what constitutes and influences sexual risk-taking behaviour among emerging adults at a tertiary institution of learning in Coastal Kenya.
Data were analysed using a thematic framework approach. A causal loop diagram was developed to map the interconnectedness of the correlates of sexual risk-taking behaviour. Our findings show that sexual risk-taking behaviour is common. Based on the ecological model and causal loop diagram, most of the reported risk factors were interconnected and operated at the individual level. Our study shows that emerging adults are engaging in sexual risk-taking behaviour. Enhancing sexuality education programs for students in Kenyan universities and strengthening support systems including counselling may help reduce sexual risk-taking behaviour among emerging adults in universities in Kenya.
Data and Data Collection Overview
A qualitative study was conducted at a large public university in Coastal Kenya between October 31st 2019 and March 16th 2020. Undergraduate students aged 18-24 years were eligible. Other stakeholders including the Dean of Students, Student Counsellor, and nurses working at the students’ health unit and the university HIV voluntary counselling and testing centre, were also recruited as key informants. Recruitment of undergraduate students was through snowballing, whereas the other key informants were purposively selected. Initially a meeting was held with undergraduate students to introduce the study. From this meeting, a few students volunteered to participate in the study. Recruited students recommended other students deemed to be knowledgeable on the subject matter. The process was repeated with all new interviewees until saturation was reached. To explore diversity, efforts were made to ensure students recommended for participation were spread across years of study, gender, program of study and region where they came from. Students who were opinion leaders, specifically those holding leadership positions in student organizations were given preference as they were deemed to have more insight on the subject matter. For the other key informants, the Dean of Students and other staff who work closely with the students and on matters around sexual risk-taking behaviour were purposively recruited.
We conducted in-depth interviews with the students and other key informants. All interviews were conducted in English. Each interview lasted about an hour and was conducted at a time convenient to each key informant. Student interviews were conducted in a private room within the voluntary counselling and testing centre. For convenience, stakeholder interviews were conducted in their offices where privacy was assured. Participants were each reimbursed Kenyan shillings 350 (an equivalent to about USD 3.00) to compensate for time spent. Interviews were moderated by Stevenson Chea in English and permission for notes taking and audio recording was sought a-priori. An interview guide was earlier developed following the World Health Organization guidelines on school-based student health surveys. Participants’ perceptions were explored using general open-ended questions followed by additional probing where appropriate. Participants’ sociodemographic data including date of birth, gender and program of study were also collected.
Data Analysis
Audio recordings from in-depth interviews were transcribed. To ensure the study team could not directly associate the transcripts with individual participants, identifying information was not included in the transcripts. A thematic framework analysis approach was applied as follows: Firstly, the transcripts were coded in QSR NVivo 12 (QSR International Ltd, Southport, UK). Initial coding was guided by major themes from the in-depth interviews and socio-ecological model. New codes and themes were developed on the basis of the in-depth interview transcripts. Secondly, excerpts were reviewed to identify common themes and variant views. Codes representing similar themes were collapsed to develop final codes. Finally, illustrative quotes that represent each theme were selected and presented. To develop the causal loop diagram, first a causal structure diagram was prepared based on the data (ex-post development) to identify the causal relationships between the correlates of sexual risk-taking behaviour. The information from the causal structure diagram was then used to develop the causal loop diagram using Vensim software version 9.3.4 (Ventana systems Inc. 2015). Where the data did not suggest presence of a connection between two variables of interest, extra connections were added based on literature, and where necessary variables renamed, to complete the causal system.
Shared Data Organization
This data project consists of one aggregated data file (containing all thematic codes, their definitions and illustrative interview excerpts for each) and five documentation files including an informed consent form, an interview guide, a sociodemographic form, an administrative README file, and this Data Narrative.
