非线性期刊声望归一化生产力：定义与度量
Nonlinear Journal Prestige Normalized Productivity: Defining and Measuring It
浏览论文内容
中文总结 AI 辅助
研究提出期刊声望归一化生产力（JPNP）指标，结合期刊声望百分位归一化、凸变换及作者份额划分，能产生现实生产力分布，更准识别顶尖研究者，减少低声望出版物噪声，提高跨领域可比性，是有用的测量工具。
中文摘要 AI 辅助
本文介绍了期刊声望归一化生产力（JPNP），这是一种非线性、百分位归一化且按作者份额划分的研究生产力指标。该指标可汇总到学科、机构和国家层面。基于发表数量且未按期刊声望归一化的传统衡量方法似乎无法体现全球科学出版系统当前的层级结构。JPNP结合了：（a）期刊声望的百分位归一化；（b）强化期刊分布上尾的凸变换；（c）作者份额划分。JPNP产生了一个偏态但现实的生产力分布，比线性度量更准确地识别出科学领域的顶尖研究者，并减少了低声望出版物产生的噪声。该指标提高了生产力的跨领域可比性，反映了当前科学领域的分层情况。JPNP是一个概念连贯且实证稳健的测量工具，在研究评估、机构基准测试和科学政策的汇总层面很有用。
英文摘要
This article introduces Journal Prestige Normalized Productivity (JPNP), a nonlinear, percentile-normalized, and authorship-fractionalized indicator of research productivity. The indicator can be aggregated to the level of disciplines, institutions, and countries. Traditional measures based on publication counts and not normalized by journal prestige do not seem to capture the current hierarchical structure of the global scientific publishing system. JPNP combines: (a) percentile normalization of journal prestige; (b) a convex transformation that strengthens the upper tail of the journal distribution; and (c) authorship fractionalization. JPNP produces a skewed but realistic distribution of productivity, identifies top performers in science more accurately than linear measures, and reduces the noise generated by low-prestige publications. The indicator improves cross-field comparability of productivity and reflects current stratification in science. JPNP is a conceptually coherent and empirically robust measurement tool, useful at the aggregated levels of research evaluation, institutional benchmarking, and science policy.
