• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Measurement >Citation Measures as Criterion Variables in Predicting Scientific Eminence
【24h】

Citation Measures as Criterion Variables in Predicting Scientific Eminence

机译：引用度量作为预测科学卓越性的标准变量

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

The target article seems to provide yet another illustration of the classic "It don't make no nevermind" principle in statistical analysis (cf. Wainer, 1976). In particular, relatively simple measures appear to do approximately as well as more complex measures do, even including indicators that represent nonlinear transformations of the simpler measures. Especially striking is how a rather straightforward indicator such as the h index seems to do an excellent job integrating quantity and quality in scholarly impact. To be sure, it is always easy to second-guess any empirical study that attempts to make exhaustive comparisons. For example, I was surprised that the authors did not scrutinize log-transformed citation counts, a measure used quite frequently in the scientometric literature (e.g., Feist, 1993, 1994; Helmreich, Spence, Beane, Lucker, & Matthews, 1980; Simonton, 1992, 2000). That said, a logarithmic transformation would be very closely approximated by the square-root transformation (viz., sqrt-C), so perhaps here, too, it really does not make any difference one way or another. Nevertheless, the former transformation may have more justification than the latter because theoretical models often predict that the cross-sectional distributions of output and impact should be approximately lognormal (e.g., Simonton, 1999, 2003). Given two measures that are empirically similar but not equivalent, the measure that features greater theoretical justification should be preferred.
机译：目标文章似乎提供了统计分析中经典的“ It's not make no mindd”原则的另一个例证（参见Wainer，1976）。特别是，相对简单的度量似乎和复杂的度量一样有效，甚至包括代表较简单度量的非线性变换的指标。尤其引人注目的是，诸如h指数之类的直接指标似乎在将数量和质量整合到学术影响力方面表现出色。可以肯定的是，总是容易对任何试图进行详尽比较的经验研究进行猜测。例如，令我惊讶的是，作者没有仔细检查对数转换后的引文计数，这是科学计量文献中经常使用的一种度量标准（例如，Feist，1993，1994； Helmreich，Spence，Beane，Lucker和Matthews，1980； Simonton ，1992，2000）。就是说，对数变换将非常接近平方根变换（即sqrt-C），所以也许在这里，它实际上也并没有任何区别。尽管如此，前者的转换可能比后者更合理，因为理论模型经常预测产出和影响的横截面分布应近似为对数正态分布（例如，Simonton，1999，2003）。给定两个在经验上相似但不相同的度量，应首选具有更大理论依据的度量。

著录项

  • 来源
    《Measurement》 |2012年第4期|170-171|共2页
  • 作者

    Dean Keith Simonton;

  • 作者单位

    Department of Psychology, One Shields Avenue, University of California, Davis, Davis, CA 95616;

  • 收录信息 美国《科学引文索引》(SCI);
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号