已有 4548 次阅读2012-5-5 01:39|个人分类:科研评价|系统分类:科研笔记|科研人员, literature, 2011, Principles, following
Lutz Bornmann在Scientometrics (2011) 86:173–177发表了一篇题为“Mimicry in science?”的文章。这篇文章涉及了一些科研评价、科研人员的行为等有关话题。正好这几天在与几位前辈探讨中涉及到了这方面的内容。下面对这篇文章的主要内容做了一些翻译,与大家分享,从中也许可以获得一些感兴趣的研究主题。
Abstract: Since bibliometric indicators have obtained a general acceptance in science
policy and attained applied relevance in research evaluation, feedback effects on scientists’
behaviour resulting from the use of these indicators for science funding decisions have
been reported. These adaptation strategies could be called mimicry in science. Scientists
apply strategies that should enable them to comply to bibliometric accountability and to
secure funds to their own research.
Keywords: Research evaluation Mimicry Scientific progress
Bibliometric data are being used by leading and fast-growing countries in science for research
evaluation purposes. In UK the allocation of public funds to the universities will be mainly
carried out according to these data. “The Government has a firm presumption that after the
2008 RAE [Research Assessment Exercise] the system for assessing research quality and
allocating “quality-related” (QR) research funding to universities from the Department for
Education and Skills will be mainly metrics-based (UK Office of Science and Technology
2006)“. Due to this development “the death of peer review” in the allocation of research funds
is being discussed (Gilbert 2006). The reason for the popularity of bibliometrics compared to
peer review is seen in lower costs and criticism of the peer review system (Weingart 2005). In
an ever more complex science system, bibliometric analysis should allow for evaluation of
mass-data no longer understandable to the single reviewer (Butler 2004).
科研人员行为的耦合
Back coupling on scientists’behaviour
Since science funding has become more and more determined by principles of ‘publish or perish,’ the following changes of publication behaviour have been reported in the literature (see here Lawrence 2003; Research Evaluation and Policy Project 2005):
(1) To increase probability of acceptance of their papers by a journal, scientists tend to do research in accordance with the mainstream in their fields and avoid unusual research (e.g. risky, interdisciplinary or long-term);
(2) To be able to come to publishable results more quickly, scientists pursue short-term rather than longterm research;
(3) Scientists attempt to provide their paper to low-quality journals as long as these journals are indexed by literature data bases used for bibliometric analyses in research evaluation;
(4) To boost the number of publications, scientists slice their findings as thin as salami and submit these to different journals even though findings could be presented in a single paper.
Fraser and Martin (2009)研究发现,在科学论文中经常会出现关键的、决定性的、重要的(e.g. pivotal, crucial, and essential)等词汇。这个发现是科研人员想通过这些词汇的运用来增加其论文被录用的可能性。Merton (1938)认为,在这种极端的“仿生”会导致一个混乱的状态。Merton (1938)研究了社会结构如何影响对特定的人群,这种影响是鼓励这些人群去遵守规则,还是不守规范。如果“赢得比赛”超过了“在一定规则下赢得比赛”,那么破坏规则的现象就会发生(Martinson et al. 2006)。如果科研人员去迎合基于评价指标的目标,科学不端行为的现象就会出现。很多情况下,发文的压力会导致学术不端行为的出现。中国一些大学的高质量成果与奖金、房子、津贴等挂钩的同时,也在面临着不少学术失范行为(Qiu 2010)。
Increase in productivity as an effect of national research evaluation systems
发文量增加对国家科研评价系统的影响
在过去一些年,除英国外,象澳大利亚、芬兰等国家利用评价系统来分配科研资金的做法被不断完善 (see an overview in Macilwain 2010; Schneider 2009)。一些调杳表明,这些评价系统对科研人员的行为产生了一定的影响。在澳大利亚(Butler 2003, 2004),自从“发文等于基金”(formula based funding)的制度实施之后,科研人员每年的发文量有了很大的提升 (如发文的与基金分配的关系),虽然这些文章被发表在WOS收录的期刊上,但多数是在低影响因子的期刊上。
Glaser and Laudel (2007)的5位访谈者表示,他们改变了他们的发文策略:发更多的论文、独立发表论文、在高影响力的期刊上发文。而且被采访对象更倾向于一些应用性和热点研究主题的论文。在西班牙,科研人员根据the National Commission for the Evaluation of Research Activity (CNEAI, Madrid)的规定在高水平的期刊发表其成果。一项对科研政策影响的测度表明:首先,这种政策导致了西班牙科研人员发文习惯的变化,其次,科研人员在ISI收录的西班牙期刊上发文的数据在增加;第三,西班牙文献计量数据库的source items趋于稳定(Jimenez-Contreras et al. 2002)。同样的影响也其它的研究中被发现(Rey et al. 1998)。UK RAE Moed (2008) 发现,三种明显的文献计量模式可以来解释科研人员对RAE标准的适应。RAE1992强调发文总数时,英国科研人员的发文量就出现一个确定的增长。当RAE1996注重由数量向质量转变时,英国科研人员的发文和被引次数都有了提高。在1997–2000年之间,机构增加了其科研人员的数量,并鼓励研究人员之间紧密合作,即使论文并不是真正合作的产物。
Possibilities of increasing citation impact
提高引用影响的可能性
与发文量相比,被引次数对科研人员似乎是不敏感的。但是Bornmann and Daniel (2007)研究结论正好相反。......只有少数研究人员能够看较多的所在领域的文献,因此,写越多的论文,其被引用的可能性越大。
Do behavioural changes contribute to scientific advancements in science?
这种行为的适应性也可以对科技进行有正面的影响。一方面,科研人员在SCI收录的低影响因子上发文是负面的,另一方面,这些期刊能够作为来源刊,它们也是要达到一定标准的。低影响因子期刊比高影响因子期刊的质量相对较差,但还是要好于那些没有被作为来源刊的期刊。一方面,好多研究人员都倾向于主流的研究是有问题的,但这些主流是一个领域最重要的研究内容。追求短期影响就真的比长期的研究会影响科学的进步吗((Laloe¨ and Mosseri 2009)?)在大科学时代,短期影响会不会影响规则?要回答这些类似的“科学仿生问题“,要深入研究评价系统、发文行为和引用行为的适应性和科学知识生产需要之间的联系是非常必要的。
(In times of big science (de Solla Price 1965), isn’t short-term impact the rule? To be able to answer these and similar questions concerning mimicry in science satisfyingly, the results of in-depth analysis on the relationship between an evaluation-based research system, adaptation of the publication and citation behaviour and scientific knowledge production are needed.)