Gender disparities in water-related knowledge, perceptions, and governance in rural Ghana: insights from a qualitative study augmented by Natural Language Processing

Gender inequalities remain persistent across water-related activities in labor division and decision-making disparities. Yet, our understanding of these dynamics remains limited to generalized narratives. This study examined gender disparities in water knowledge, perceptions, and governance in rural Ghana, drawing from focus group discussions (FGDs) conducted between January and March 2023. Using both qualitative thematic analysis and Natural Language Processing (NLP), we analyzed FGDs with women (n = 30 FGDs) and men (n = 25 FGDs) covering perceptions, experiences, knowledge, and satisfaction related to drinking water sources and water system governance structures. FGD transcripts were qualitatively analyzed using thematic analysis in Nvivo and coded text was further examined using NLP tools such as sentiment analysis. This approach revealed disparities in experiences with and perspectives on drinking water, in both qualitative and NLP results. Our qualitative results indicate that women had detailed impressions of water quality, closely tracking sensory parameters and health outcomes. In contrast, men tended to adopt a broader perspective, focusing on the local political environment and governance. Furthermore, NLP results suggest women had more negative perceptions of chlorination and a heightened awareness of water treatment than men, indicating active involvement of women in water management could enhance treatment and quality. While both NLP and qualitative results show shared frustrations with governance and a desire to improve water system management, there was no unified demand to increase women's participation. These results highlight nuanced gender perceptions regarding drinking water and reveal a complex set of gendered dynamics. Furthermore, this study introduces an innovative research approach integrating qualitative thematic analysis with NLP tools to enhance the analytical depth and validation of qualitative findings.