Customizing a Tool to Collect Complex Network Data among HIV-positive Youth in South Africa (2017-2018)


HIV-positive youth repeatedly demonstrate lower levels of retention in HIV care and adherence to antiretroviral therapy compared to older adults. Developing evidence-based interventions to improve HIV care and engagement in this population will require an in-depth understanding of factors shaping retention and adherence behaviors from multiple levels of influence.

Developmental theories suggest that youth are particularly sensitive to and motivated by the social networks in which they are embedded. However, no studies have assessed the social networks of HIV-positive youth in South Africa to understand the structure and composition of these networks. This is likely because of the enormous methodological challenges historically involved in collecting and analyzing such complex data. Recent technological innovations facilitate the collection and analysis of complex network data, though customizing newly created tools to collect network data among HIV-positive youth in South Africa will require an interdisciplinary approach.

Project Description

The goal of this Bass Connections project is to customize and pilot a novel network data capture tool to collect data on the social networks of HIV-positive youth in South Africa.

With guidance from the Duke University Network Analysis Center, this project team will customize an existing open-source network data capture tool called NetworkCanvas for use in the South African context. NetworkCanvas was developed by researchers at Northwestern University to study the social networks of young men who have sex with men. The tool has many benefits including an interactive touchscreen interface and a user-centered design. However, it has not yet been utilized in international settings and is built to handle only certain types of traditional data. Thus, the team will customize the tool to capture complex data on the various social networks of HIV-positive youth.

In order to inform the development of the data capture tool, the project team will also aim to understand the potential types of social networks (e.g., social support networks, disclosure networks, sexual networks) that should be assessed. Team members will explore and discuss literature on adolescent/youth health, gender and culture in South Africa, social network influence on health behaviors and stigma and barriers of adherence to antiretroviral therapy among youth.

After customizing the tool, the team will pre-pilot an English version of the instrument with youth in Durham. Pre-piloting the tool will allow team members to immediately implement changes to the tool based on feedback solicited from the participants.

Subsequently, Dr. Mulawa (team leader) will travel to Cape Town, South Africa, to meet with the team’s South African collaborators, deliver the tablets and train a local data collector in using the instrument with participants. They will pilot the tool with youth in Cape Town until they have collected data from approximately 15 HIV-positive youth. This will allow the team to collect complex network data of up to 40 network members (and the relationships between them) for each participant. Once the pilot is complete, the team will then analyze the network data collected from these participants to describe the structure and composition of their networks.

Anticipated Outcomes

Published paper describing the adaptation of this network data capture tool, the process of pre-piloting the instrument with youth in Durham and the feasibility of using the tool with youth in South Africa as well as the findings of the pilot; adapted network data capture tool for use in future longitudinal research in South Africa

Student Opportunities

We seek to engage at least one doctoral student, one master’s student and two undergraduate students.

The two undergraduate students (optimally, computer science majors or other undergraduates with familiarity and expertise with Javascript, HTML and CSS) will have the opportunity to apply computer programming skills to customize a network data capture tool for research in a global setting. They will have the opportunity to learn about issues of gender and culture in the South African context and consider how those issues are related to HIV-related health issues among youth. These students will have the opportunity to be involved in the pre-pilot with youth in Durham. They will also be exposed to social network analysis methods and will learn how to construct network measures and to visualize multiplex (overlapping) networks. The students will share their research experience and findings at the Visible Thinking poster session.

The master’s student (optimally, MSc student in global health) will work closely with the team leader to prepare materials for review and discussion in the weekly team meetings. This student will play an important role in pre-piloting the tool with youth in Durham and leading the focus group discussion.

The doctoral student (optimally, PhD student in computer science with expertise in Javascript, HTML and CSS) will gain leadership and management skills by serving as project manager and mentor to the undergraduate students. S/he will provide direct guidance to the undergraduates and be responsible for managing the shared repository (code review, bug tracking and integration). This student will receive additional guidance and mentorship from the team consultant, Dr. Basant Singh. Communication with Dr. Singh will take place through email and video conferencing.


Fall 2017 – Summer 2018

  • Fall 2017: Begin customization of network data capture tool
  • Spring 2018: Complete customization of tool; pre-pilot with youth in Durham; pilot with youth in Cape Town; analyze data
  • Summer 2018: Complete data analysis by June 2018


Independent study credit available for fall and spring semesters; summer funding

Faculty/Staff Team Members

James Moody, Trinity - Sociology
Marta Mulawa, Duke Global Health Institute
Kathleen Sikkema, School of Medicine - Psychology and Neuroscience; Duke Global Health Institute

Community Team Members

John Joska, University of Cape Town
Basant Singh,

* denotes team leader