Representation and societal alignment as facets of research evaluation: an analysis of Tanzanian forest research across different databases

Our study focuses on research representation and societal alignment as two facets of research evaluation. We investigated how Tanzanian forest research is represented in bibliographic databases and the extent to which these representations align with national societal demands. To address the former, we conducted a representation analysis of forestry research publications by Tanzanian-affiliated authors in the period 2005–2021. We used four databases: two commercial (Scopus and Web of Science), one semi-commercial (Dimensions), and one fully open (OpenAlex). We also added a manually curated dataset of articles from local Tanzanian journals, resulting in a total of 1,496 unique articles. While both OpenAlex and the local journal dataset posed challenges, both were necessary to provide the most comprehensive representation of forest research in Tanzania. Together, they account for 90% of the ‘total’ set of publications. For the second question, we selected articles from 2017 to 2021 only, downloading their full texts. We conducted a quantitative content analysis by coding these 728 articles in terms of four Tanzanian socio-ecological elements that we identified as potential markers of societal demand: national research priorities in forestry, forest tenure arrangements, forest types, and forest regions. Our findings suggest that the issue of alignment is not simply binary (with publications in local journals being better aligned with national interests and those in mainstream databases being less aligned), but rather one of convergence and divergence depending on the context and measures used.