Department of General Psychology and Methodology, University of Bamberg, Bamberg, Germany.
PLoS One. 2011;6(12):e28770. doi: 10.1371/journal.pone.0028770. Epub 2011 Dec 14.
Assessing an individual's research impact on the basis of a transparent algorithm is an important task for evaluation and comparison purposes. Besides simple but also inaccurate indices such as counting the mere number of publications or the accumulation of overall citations, and highly complex but also overwhelming full-range publication lists in their raw format, Hirsch (2005) introduced a single figure cleverly combining different approaches. The so-called h-index has undoubtedly become the standard in scientometrics of individuals' research impact (note: in the present paper I will always use the term "research impact" to describe the research performance as the logic of the paper is based on the h-index, which quantifies the specific "impact" of, e.g., researchers, but also because the genuine meaning of impact refers to quality as well). As the h-index reflects the number h of papers a researcher has published with at least h citations, the index is inherently positively biased towards senior level researchers. This might sometimes be problematic when predictive tools are needed for assessing young scientists' potential, especially when recruiting early career positions or equipping young scientists' labs. To be compatible with the standard h-index, the proposed index integrates the scientist's research age (Carbon_h-factor) into the h-index, thus reporting the average gain of h-index per year. Comprehensive calculations of the Carbon_h-factor were made for a broad variety of four research-disciplines (economics, neuroscience, physics and psychology) and for researchers performing on three high levels of research impact (substantial, outstanding and epochal) with ten researchers per category. For all research areas and output levels we obtained linear developments of the h-index demonstrating the validity of predicting one's later impact in terms of research impact already at an early stage of their career with the Carbon_h-factor being approx. 0.4, 0.8, and 1.5 for substantial, outstanding and epochal researchers, respectively.
基于透明算法评估个人的研究影响力是评估和比较的重要任务。除了简单但不准确的指标，如仅仅计数出版物的数量或总引用量的积累，以及以原始格式呈现的高度复杂但也令人不知所措的全范围出版物列表外，Hirsch（2005）引入了一个巧妙地结合了不同方法的单一数字。所谓的 h 指数无疑已成为个人研究影响力的科学计量学标准（注：在本文中，我将始终使用“研究影响力”一词来描述研究表现，因为本文的逻辑基于 h 指数，该指数量化了特定研究人员的具体“影响力”，也因为“影响力”的真正含义既指质量，也指质量）。由于 h 指数反映了研究人员发表的至少有 h 次引用的论文数量 h，因此该指数固有地偏向于高级别研究人员。当需要评估年轻科学家的潜力时，特别是在招聘早期职业职位或为年轻科学家的实验室配备资源时，这有时可能会成为一个问题。为了与标准 h 指数兼容，建议的指数将科学家的研究年龄（Carbon_h 因子）纳入 h 指数，从而报告每年 h 指数的平均增益。为了广泛的四个研究学科（经济学、神经科学、物理学和心理学）和三个高水平的研究影响力（实质性、卓越和划时代）的研究人员，对 Carbon_h 因子进行了全面计算，每个类别有 10 名研究人员。对于所有研究领域和产出水平，我们都获得了 h 指数的线性发展，表明通过 Carbon_h 因子在职业生涯的早期阶段就可以预测一个人的后期影响力，这在研究影响力方面是有效的，对于实质性、卓越和划时代的研究人员，Carbon_h 因子分别约为 0.4、0.8 和 1.5。
