The International Journal of Biostatistics

The International Journal of Biostatistics is a peer-reviewed journal issued from Berlin by Walter de Gruyter, started in 2005 to publish biostatistical models, statistical theory and applications with a declared emphasis on semiparametric methods.《国际生物统计学杂志》由柏林的 Walter de Gruyter 出版，实行同行评议，2005 年创办，发表生物统计模型、统计理论及其应用，明确以半参数方法为重。

Its 2025 Journal Impact Factor stands at 1.9, which puts the title in the first quartile of statistics and probability and in the third quartile of mathematical and computational biology.2025 年影响因子为 1.9，该刊在统计学与概率论类目属 Q1，在数学与计算生物学类目为 Q3。

Scope

• The scope statement promises the entire range of biostatistics, from theoretical advances to the sensible translation of a practical problem into a statistical framework.刊物把选题范围界定为生物统计学的全部疆域，从理论推进，到把实际难题稳妥地转写进统计框架。
• Semiparametric modelling is named as the editorial centre of gravity, and every manuscript is expected to say plainly which inferential question it answers.半参数建模被点名为编辑重心，每篇来稿都须直白交代自己所回答的推断问题。
• Estimation strategies built on machine learning and other data-adaptive tools are welcomed, a position the editors set out against a crowded field of biostatistics outlets.基于机器学习与其他数据自适应工具的估计策略受到欢迎，编辑部把这一取向放在生物统计学刊物林立的格局中陈述。
• Applied work is invited from the biological, medical and public health sciences as well as agriculture, with clinical trials and epidemiological studies among the recurring subjects.应用类稿件来自生物学、医学、公共卫生与农业各领域，临床试验与流行病学研究是其中反复出现的题材。
• Causal inference holds a steady share of the pages, including targeted learning estimators that combine machine learning with an explicitly chosen statistical target.因果推断在版面中占据稳定份额，把机器学习与明确选定的统计目标结合起来的目标学习估计量即属此类。
• Time-to-event analysis and longitudinal designs appear often, since cohorts followed over years are where semiparametric estimators earn their keep.生存分析与纵向设计屡屡出现，随访多年的队列正是半参数估计量显出身手的场合。
• Missing data, measurement error and confounding by indication are treated as first-class problems rather than nuisances to be patched at the end of an analysis.缺失数据、测量误差与适应证混杂被当作首要问题处理，并不留到分析末尾再打补丁。
• High-dimensional and omics settings supply much of the raw material, where the number of covariates runs well past the number of subjects.高维与组学场景提供了大量素材，协变量个数远超受试者人数的情形在其中十分常见。
• Nothing in the statement excludes Bayesian work, though the emphasis named there falls on semiparametric and data-adaptive estimation.范围陈述并不排斥贝叶斯工作，其中点明的重心仍落在半参数与数据自适应估计上。
• Papers on study design belong in the remit as well, on the reasoning that an analysis can only be as good as the sampling scheme beneath it.研究设计类论文同属范围，其理由在于分析的品质受制于其下的抽样方案。
• Electronic publication lets authors append data sets and software code, an arrangement the journal presents as an opening for reproducible research.电子出版允许作者附上数据集与软件代码，刊物把这一安排称作通向可复现研究的大门。
• Original research and review articles are both solicited, and papers that apply sound statistical methods to practical problems are explicitly welcomed.原创研究与综述均在征求之列，用可靠统计方法处理实际问题的论文也明确受到欢迎。
• Theory papers belong here too, provided the mathematics stays tied to a question that arose outside the seminar room.理论文章同样在此范围内，前提是数学推导始终系于研讨室之外提出的问题。
• Subgroup analysis and precision medicine raise questions about heterogeneity that the journal's inferential bent suits rather well.亚组分析与精准医学带来的异质性问题，与刊物的推断取向颇为契合。

History

• Volume 1, number 1 appeared on 1 May 2005, the work of Berkeley Electronic Press, a California publisher founded in 1999 and known for its born-digital lists.第 1 卷第 1 期于 2005 年 5 月 1 日面世，出版方是 1999 年创办于加利福尼亚的 Berkeley Electronic Press，以纯电子期刊起家。
• The journal was born digital, with no printed edition to carry, so the publishing model was built around continuous online release instead of a fixed press cycle.刊物自始即为纯电子刊，没有纸本需要承载，出版模式围绕连续在线发布搭建，而不按固定的印刷周期排期。
• Its first editorial home was in Berkeley, California, and the early volumes drew heavily on the semiparametric school that had grown up in that department.最早的编辑部设在加利福尼亚州伯克利，最初几卷的稿件多出自当地成长起来的半参数学派。
• De Gruyter bought the Berkeley Electronic Press journal list in 2011, and the imprint travelled from California to Berlin.2011 年，De Gruyter 收购了 Berkeley Electronic Press 的期刊资产，出版地也随之从加利福尼亚转到柏林。
• The transfer brought a second ISSN, 2194-573X, alongside the original electronic number 1557-4679, and the DOI prefix moved from 10.2202 to 10.1515.转手之后，除原有的电子刊号 1557-4679 之外，又添了一个刊号 2194-573X，DOI 前缀也从 10.2202 换成 10.1515。
• MEDLINE and PubMed have carried the journal from the first volume of 2005 onward, a stretch of indexing that began before the change of owner.自 2005 年第 1 卷起，MEDLINE 与 PubMed 就持续收录该刊，这段收录史比出版方更替更早开始。
• Each calendar year now forms a single volume numbered in two issues, and both are filled as papers clear production rather than on a fixed press date.如今每个日历年合成一卷，卷内编两个期号，两期都随论文完成制作陆续填满，不按固定付印日期排定。
• Accepted papers go online ahead of the issue, a practice the publisher advertises as rapid online publication of contributions ahead of print.录用论文先于整期上线，出版方把这一做法称作录稿先行在线发布。
• Impact factors have been modest and uneven, running from 0.968 in 2021 through 1.829 in 2022 and 1.0 in 2023 to 1.2 in 2024 and 1.9 in 2025.影响因子的走势不算平稳，2021 年为 0.968，2022 年升至 1.829，2023 年回落到 1.0，2024 年为 1.2，2025 年又回到 1.9。
• Total citations passed 1,200 while the cited half-life runs beyond thirteen years, which suits a methods journal whose papers age slowly.总被引次数超过 1200 次，半衰期长于 13 年，这与论文老化缓慢的方法类刊物相称。
• Output stays small, around 27 articles in the latest year, a figure in keeping with the narrow pool of specialised methods work.年发文量不大，最近一年约 27 篇，这与专门化方法研究有限的稿件来源相称。
• The web platform carrying the journal now shows the joint De Gruyter Brill branding, following the combination of the two German houses.如今承载该刊的网站已改用 De Gruyter Brill 的联合品牌，两家德国出版社合并之后即是如此。

Editors

• Three editors share the running of the journal, Antoine Chambaz, Alan E. Hubbard and Mark J. van der Laan, all identified with semiparametric estimation.刊物由三位编辑共同主持，即 Antoine Chambaz、Alan E. Hubbard 与 Mark J. van der Laan，三人都与半参数估计一脉相承。
• Chambaz works at Université Paris Cité, where his research covers causal inference, model selection and the theory behind targeted learning.Chambaz 任教于巴黎西岱大学，研究涵盖因果推断、模型选择与目标学习的理论。
• Hubbard and van der Laan are both at the University of California, Berkeley, in the biostatistics division that supplied much of the early authorship.Hubbard 与 van der Laan 同在加利福尼亚大学伯克利分校，所在的生物统计部门曾为早期各卷大量供稿。
• van der Laan has spent his career on asymptotically linear estimators and on the conditions under which they behave well in finite samples.van der Laan 长期研究渐近线性估计量，以及这类估计量在有限样本中表现良好的条件。
• Two of the three are based in California and one in Paris, an arrangement that runs back to the journal's Berkeley origins.三人之中两位在加利福尼亚，一位在巴黎，这一格局可追溯到刊物的伯克利渊源。
• The editorial listing names the three editors without the country-by-country board roster that larger titles print.编辑名单列出的是这三位编辑，没有大型刊物那种按国别铺开的长编委表。
• Every manuscript goes through external peer review before acceptance, and the journal describes all of its contributions as refereed.所有稿件在录用前都经过外部同行评议，刊物把所刊论文一律视为经审读之作。
• Manuscripts arrive through the publisher's online system, and authors are asked to state the inferential problem at the outset.稿件经出版方的在线系统投递，作者须在开篇处点明所处理的统计推断问题。
• Review articles enter the same referee track as research papers, with the same demand for a stated target parameter.综述文章与研究论文走同一条审稿流程，同样要交代清楚目标参数。
• What the editors ask of a paper is candour about its target, which they treat as the discipline that keeps data-adaptive methods honest.编辑对论文的要求是把目标讲清楚，他们认为这条纪律能让数据自适应方法保持诚实。
• Papers on targeted maximum likelihood by the editors themselves sit in the earlier volumes, next to work from the same school, and the methodological kinship shows.几位编辑研究目标极大似然的论文收在较早的卷次里，与同一学派的成果比邻而列，方法上的亲缘关系一望可知。