Personal social networks have a profound impact on health, but collecting personal network data for use in health communication, behavior change, or translation and dissemination interventions has proven to be challenging. Recent advances in social network data collection software have reduced the burden of network studies on researchers and respondents alike, but little testing has occurred to discover whether these methods are: (1) acceptable to a variety of target populations, including those who may have limited experience with technology or limited literacy; and (2) practical in the field, specifically in areas that are geographically and technologically disconnected, such as rural Appalachian Kentucky.
To fill this knowledge gap, investigators from the University of Kentucky College of Public Health explored the early-stage feasibility (Acceptability, Demand, Implementation, and Practicality) of using innovative, interactive, tablet-based network data collection and visualization software, OpenEddi, in field collection of personal network data in Appalachian Kentucky. The resulting article appears in JMIR Research Protocols.
A total of 168 rural Appalachian women who had previously participated in a study on the use of a self-collected vaginal swab (SCVS) for human papillomavirus testing were recruited by community-based nurse interviewers between September 2013 and August 2014. Participants completed egocentric network surveys via OpenEddi, which captured social and communication network influences on recruitment and participation for the SCVS study.
After completion of the network study, investigators conducted a qualitative group interview with four nurse interviewers and two participants. Using this qualitative data along with quantitative data collected through the network study, they applied guidelines from Bowen et al to assess feasibility in four areas of early-stage development of OpenEddi: Acceptability, Demand, Implementation, and Practicality. Basic descriptive network statistics (size, edges, density) were analyzed using RStudio.
Investigators found that the OpenEddi software was perceived as fun, novel, and superior to other data collection methods or tools. Respondents enjoyed the social network survey component, and visualizing social networks produced thoughtful responses from participants about leveraging or changing network content and structure for specific health-promoting purposes. Areas for improved literacy and functionality of the tool were identified. However, technical issues led to substantial (50%) data loss, limiting the success of its implementation from a researcher’s perspective, and hindering practicality in the field.
The conclusion drawn from the study is that OpenEddi is a promising data collection tool for use in geographically isolated and socioeconomically disadvantaged populations. It is predicted that future development will mitigate technical problems, improve usability and literacy, and test new methods of data collection. These changes will support goals for use of this tool in the delivery of network-based health communication and social support interventions to socioeconomically disadvantaged populations.
Dr. Kate Eddens, assistant professor of Health, Behavior & Society, is the lead author of the study. Tom Collins, Associate Director of the Rural Cancer Prevention Center at the UK College of Public Health, is a co-author.