h指数与发文量、篇均引文量关系模型的验证与评价  

Testing and Evaluation of Relational Model of H-index and Papers Quantity as Well as Average Quotation Quantity

作　　者：汪跃春[1] 胡敏[1] 

机构地区：[1]金陵科技学院图书馆,南京210038

出　　处：《情报杂志》2011年第10期49-54,共6页Journal of Intelligence

基　　金：江苏省教育厅高校哲学;社会科学基金项目"h指数与论文被引规律研究"(编号:2011SJD870011)研究成果之一

摘　　要：通过本人统计、数据库系统列表及公开出版文献中提供的有关h指数、被引量、发文量或篇均引文量的统计样本共11个,样本内容涉及机构、学者、期刊、基金与专利权人等,统计数据来源于国内外5个不同引文数据库,并将取得的11个样本合并成一个混合样本,共计12个样本。以这12个样本对目前的4个h指数与被引量、篇均引文量模型分别进行拟合与χ2检验,在此基础上对取得的11个样本进行分类分组验证,结果表明模型1对个人h指数的估算最为精确,模型5对团体h指数在h<102范围内估算最为精确,模型3作为理论模型无论对个人h指数或团体h指数的估算具有普适性。团体h指数对被引量与篇均被引量之间的规律不一致,随着总被引量增加h指数增加的同时,篇均被引量对h指数的影响加强,根据团体h指数样本将模型5修正为模型7,模型7可作为不同h指数范围的估算公式。There are eleven statisticsamples related to h-index and cited quantity as well as average quotation quantity which are supplied by our statistics and DBS＇ listing along with public published literature.The statisticsamples involve institution,scholar,periodical,fund and patentee.Statistical data root in five different reference database and count up to twelve samples through joining the eleven samples to one.According to the twelve samples,the fitting and χ2 of four h-index and cited quantity as well as average quotation quantity are differently tested,on the basis of whch the eleven samples are categorically analyzed,and the result indicates that the estimate of the first model is the most accurate to the personal h-index;the fifth model is the most accurate to the group h-index within h102,and the third model is universal to the personal and group h-index as theoretical model.The discipline of the group h-index is inconformity to cited quantity and average quotation quantity,and the influence of the average quotation quantity on h-index is enhanced as the total cited quantity＇s increase.And based on the group h-index samples,the fifth model is amended to the seventh which can be an estimate formula of h-index in different scope.

关 键 词：H指数 模型拟合 χ2检验 

分 类 号：G350[文化科学—情报学]