Estimating scientific coherence using population-level indicators and research production data: a longitudinal...

The rapid expansion of scientific research is frequently assumed to translate into improvements in population health and development outcomes, yet methods to empirically evaluate this alignment remain limited. Existing bibliometric and impact-based approaches describe scientific activity but rarely examine its longitudinal relationship with population-level indicators. We introduce the concepts of scientific coherence and development coherence, referring to measurable associations between research production, population indicators, and structural determinants over time. To operationalize these concepts, we propose the Data-driven Analysis and Inference of Longitudinal population indicators and research production (DAIL) framework, a three-step analytical pipeline integrating regression models, hierarchical mixed-effects analyses, and moderator screening. A proof-of-concept application illustrates how longitudinal associations between research production and global indicators can be quantified using widely available data. While our approach quantifies these longitudinal patterns, we explicitly acknowledge the inherent potential for reverse causality, recognizing that favorable socioeconomic conditions and structural development may act as prerequisites for sustaining a functioning academic research infrastructure, rather than acting strictly as outcomes of expanded research. This framework provides a methodological basis for studying the co-evolution and alignment between scientific activity and population dynamics in epidemiology.