Divergent Institutional Logics of Quantification in Research Evaluation: Evidence from RPT Policies in Chinese and U.S. Research Universities
Authors/Creators
Description
This repository contains the institutional documents, coded dataset, and analysis scripts used in the study:
"Divergent Institutional Logics of Quantification in Research Evaluation: Evidence from RPT Policies in Chinese and U.S. Research Universities".
The repository consists of four components:
(1) Original RPT policy documents from 100 research-intensive universities, including 50 institutions in China and 50 in the United States. These materials comprise publicly available promotion and tenure (RPT) policies, faculty appointment rules, and performance evaluation guidelines collected from official university websites.
(2) A consolidated coded dataset (Excel/CSV format) summarizing institutional design features for all 100 universities. The dataset includes the following variables:
ID, Country, University, Authority Structure, Degree of Delegation, Outcome-based Thresholds, Process-based Thresholds, Administrative Conversion, and Peer Benchmarking.
Each observation represents one university. Binary indicators ("Yes/No") and ordinal scales are used to capture how quantification is institutionally embedded within RPT regimes across the two national contexts.
(3)All documents were coded using a standardized protocol with explicit definitions and decision rules. Inter-coder reliability was assessed on a subset of the data independently coded by three trained coders using a back-to-back procedure. Agreement was evaluated using Krippendorff’s alpha (implemented in "Krippendorff_alpha.py"), appropriate for nominal and ordinal data. Disagreements were reviewed to refine coding rules and establish consensus.The final dataset used for analysis is based on these reconciled coding decisions.
(4) A Python script for statistical analysis and figure generation: "analysis_and_visualization_RPT_quantification.py". This script uses the finalized dataset to reproduce all descriptive statistics, statistical tests, tables, and figures reported in the article.
(5) Supplementary outputs, including generated figures and tables used in the manuscript.
This repository is intended to enhance transparency, reproducibility, and comparative research on research evaluation, quantification, and higher education governance.