美国的犹太人发明家知多少？

武夷山

美国哈佛医学院的Igor Kissin在2011年的《科学计量学》杂志曾发表文章，比较了美国顶级生物医学杂志上犹太人作者（根据典型犹太人姓氏来判断）文章数量占文章总量的比例与犹太人获得生理学与医学诺贝尔奖人数占美国人获得生理学与医学诺贝尔奖人数的比例，发现这两个比例很接近，两个指标实际值与期望值（根据犹太人人口数来估计期望值）之比都高达20％。

2013年11月出版的《科学计量学》杂志上， Igor Kissin和美国阿拉巴马大学生物统计系的Edwin L. Bradley Jr.共同发表论文（A surname-basedpatent-related indicator: the contribution of Jewish inventors to US patents，原文见http://link.springer.com/content/pdf/10.1007%2Fs11192-013-1005-9.pdf），再次分析犹太人的成就，这次是分析犹太人发明家对美国专利总量的贡献。他们统计了每千名同姓者的专利数，其中包括96个典型犹太姓氏（拥有每个姓氏的人数至少为3000人）。结果发现，犹太人获得美国专利数的实际值与期望值（基于犹太姓氏人口数占人口总数的比例，或某些犹太姓氏发明者所获专利数占美国前10大姓发明者所获专利数之比例）之比也高达6.1。另外，他们再次统计美国顶级生物医学杂志上每千名犹太人同姓者的文章数量的实际值与期望值之比，为23.1。比如，最大的犹太姓氏在全美国有87，226人，该姓氏获得的专利有894项，在顶尖生物医学期刊发表论文3079篇。

可见，犹太人发明成就也很突出，但不如论文发表方面的成就突出，反映在实际值与期望值之比，一为6.1，另一为23.1。两位作者推测，这可能与犹太人传统上特别推崇医生职业有关。

该文摘要如下：

A surname-basedpatent-related indicator: the contribution of Jewish inventors to US patents

• Igor     Kissin,

• Edwin     L. Bradley Jr.

• … show all 2 hide

368Downloads 200 Citations 9 Comments

DownloadPDF (218 KB) ViewArticle

Abstract

It was found that the surname-based representation ofJewish authors in the top US biomedical journals corresponds to the representationof Jewish Nobel Laureates in Medicine among US laureates: Both of them arealmost equally disproportionately high, with the ratio of actual to expectednumber close to 20 (Kissin, Scientometrics 89:273–280, 2011). The main aim ofthis study was to determine whether the contribution of Jewish inventors isalso disproportionately high. The number of patents (US Patent and TrademarkOffice database) per thousand persons with the same surname (2000 Census) wasdetermined (index P). Index P was compared with index A, which represents thenumber of the articles in the top US biomedical journals, and index G, which isbased on the representation of a surname in the Google’s option “Discussions”,reflecting a combination of various business and leisure activities (designedas a negative control). The collective contributions of the 96 Jewish surnamegroups for each of the above indices were calculated. The ratio of actual toexpected number of US patents was found to be disproportionately high—6.1 (p < 0.0001).At the same time, this disproportionality was four-fold lower than that relatedto biomedical articles (ratio of 6.1 vs. 23.3, p < 0.0001).There was some degree of correlation between index P and index A (r = 0.407,p < 0.0001), but no significant correlation was foundbetween index P and index G. The role of various factors in the observeddisproportionalities is discussed. The greater degree of disproportionalitywith biomedical research articles might be a consequence of traditional Jewishinclination towards occupations in medicine.