Journal Article
The use of linear models in quantitative research
Quantitative Science Studies, 1–32 (2024)
Abstract
The diversity of analysis frameworks used in different fields of quantitative research is understudied. Using bibliometric data from the Web of Science (WOS), we conduct a large-scale and cross-disciplinary assessment of the proportion of articles that use linear models in comparison to other analysis frameworks from 1990 to 2022 and investigate the spatial and citation patterns. We found that, in absolute terms, linear models are widely used across all fields of science. In relative terms, three patterns suggest that linear-model-based research is a dominant analysis framework in Social Sciences. First, almost two-thirds of research articles reporting statistical analysis framework reported linear models. Second, research articles from underrepresented countries in the WOS data displayed the highest proportions of articles reporting linear models. Third, there was a citation premium to articles reporting linear models in terms of being cited at least once for the entire period, and for the average number of citations until 2012. The confluence of these patterns may not be beneficial to the Social Sciences as it could marginalize theories incompatible with the linear models’ framework. Our results have implications on quantitative research practices, including teaching and education of the next generations of scholars.
Keywords: World, computational social science, statistical analysis