Gender differences in grant and personnel award funding rates at the Canadian Institutes of Health Research based on research content area: A retrospective analysis
Gender differences in grant and personnel award funding rates at the Canadian Institutes of Health Research based on research content area: A retrospective analysis
复制标题
DOI：
10.1371/journal.pmed.1002935
复制
发表时间：
2019-10-01
期刊：
影响因子：
15.8
通讯作者：
Guyatt, Gordon
中科院分区：
文献类型：
--
作者：
Burns, Karen E. A.;Straus, Sharon E.;Guyatt, Gordon
背景虽然在所有职业生涯阶段的女性比男性更有可能离开学术界，但职业生涯早期的女性是一个特别高的风险群体。研究表明，妇女获得研究资金的可能性低于男子;然而，资助成功率是否因研究内容而异尚不清楚。我们解决了在15年的时间里，代表研究社区的13个研究所中的一个或多个研究所的资助成功率的性别差异。方法和结果我们回顾性地审查了55，700补助金和4，087人员奖申请提交给加拿大卫生研究院。我们根据性别和申请人选择的主要研究所分析了申请成功率，使用随机效应模型汇总了成功率的性别差异，并拟合泊松回归模型以评估性别，时间和研究所的影响。我们注意到，针对选定研究所的赠款申请的成功率各不相同，而且随着时间的推移，成功率不断下降。妇女分别提交了31.1%和44.7%的赠款和人员奖励申请。在汇总的估计中，与男性相比，女性的资助成功率显著较低（风险比[RR] 0.89，95%置信区间[CI] 0.84-0.94; p < 0.001;绝对差异3.2%），具有显著异质性（I-2 = 58%）。与男性相比，向癌症研究所提供赠款的女性（RR 0.86，95% CI 0.78-0.96），循环和呼吸健康（RR 0.74，95% CI 0.66-0.84），卫生服务和政策研究（RR 0.78，95% CI 0.68-0.90），肌肉骨骼健康和关节炎（RR 0.80，95%CI 0.69-0.93）获得资助的可能性明显较低，而那些向土著人民健康研究所提供赠款的人（RR 1.67，95% CI 1.0-2.7）更有可能获得资助。总体而言，女性的人事奖励成功率也显著较低（RR 0.75，95% CI 0.65-0.86; p < 0.001;绝对差异6.6%）。回归模型确定，性别对补助金成功率的影响因机构而异，而不是时间。研究的局限性包括使用机构作为替代标识符，主要机构的指定的可变性，以及缺乏反映申请人，共同申请人，同行评审员和同行评审过程的指标。结论性别差异存在整体的赠款和人员奖励的成功率，特别是针对选定的研究社区的赠款。资助机构应监测性别差异的赠款成功率整体和研究内容。研究表明，女性申请人获得研究资金的可能性低于男性。在各个职业阶段，女性比男性更有可能离开学术界，尤其是博士后培训生和职业生涯早期的女性。我们回顾性地审查了加拿大卫生研究院根据研究内容领域提交的赠款和人员奖励，以评估资助成功率的性别差异，这对于确定资助中是否存在性别差异非常重要。这一信息对于解决审评过程中可能存在的偏见十分重要。研究人员做了什么，发现了什么？我们回顾性地审查了15年来提交给加拿大卫生研究院的55，700份赠款和4，087份人员奖励申请。我们发现，与男性相比，女性获得赠款和新研究人员奖励的可能性明显降低。此外，我们发现在特定的研究内容领域，包括癌症，循环和呼吸系统健康，卫生服务和政策研究，以及有利于男性科学家的肌肉骨骼健康和关节炎，资助成功的梯度存在显着差异。回归模型确定，性别对赠款成功率的影响因研究界而异，而不是时间。这些发现意味着什么？这些研究结果要求提高认识的申请人，同行评审员和资助机构的潜在偏见存在于赠款和人员奖励申请的评估。我们的研究结果强调，资助机构需要监测整体和研究内容领域的赠款成功率的性别差异，并在确定性别差异时探索可能的解释。迫切需要进行更多的研究，以解释成功率方面存在性别差异的原因，并确定同行审查过程中造成偏见的条件。
Background Although women at all career stages are more likely to leave academia than men, early-career women are a particularly high-risk group. Research supports that women are less likely than men to receive research funding; however, whether funding success rates vary based on research content is unknown. We addressed gender differences in funding success rates for applications directed to one or more of 13 institutes, representing research communities, over a 15-year period. Methods and findings We retrospectively reviewed 55,700 grant and 4,087 personnel award applications submitted to the Canadian Institutes of Health Research. We analyzed application success rates according to gender and the primary institute selected by applicants, pooled gender differences in success rates using random effects models, and fitted Poisson regression models to assess the effects of gender, time, and institute. We noted variable success rates among grant applications directed to selected institutes and declining success rates over time. Women submitted 31.1% and 44.7% of grant and personnel award applications, respectively. In the pooled estimate, women had significantly lower grant success (risk ratio [RR] 0.89, 95% confidence interval [CI] 0.84-0.94; p < 0.001; absolute difference 3.2%) compared with men, with substantial heterogeneity (I-2 = 58%). Compared with men, women who directed grants to the Institutes of Cancer Research (RR 0.86, 95% CI 0.78-0.96), Circulatory and Respiratory Health (RR 0.74, 95% CI 0.66-0.84), Health Services and Policy Research (RR 0.78, 95% CI 0.68-0.90), and Musculoskeletal Health and Arthritis (RR 0.80, 95% CI 0.69-0.93) were significantly less likely to be funded, and those who directed grants to the Institute of Aboriginal People's Health (RR 1.67, 95% CI 1.0-2.7) were more likely to be funded. Overall, women also had significantly lower personnel award success (RR 0.75, 95% CI 0.65-0.86; p < 0.001; absolute difference 6.6%). Regression modelling identified that the effect of gender on grant success rates differed by institute and not time. Study limitations include use of institutes as a surrogate identifier, variability in designation of primary institute, and lack of access to metrics reflecting applicants, coapplicants, peer reviewers, and the peer-review process. Conclusions Gender disparity existed overall in grant and personnel award success rates, especially for grants directed to selected research communities. Funding agencies should monitor for gender differences in grant success rates overall and by research content.Author summaryWhy was this study done? Research supports that woman applicants are less likely to receive research funding than men. Women at all career stages are more likely to leave academia than men, especially women who are postdoctoral trainees and early in their careers. We retrospectively reviewed grant and personnel award submissions based at the Canadian Institutes of Health Research based on research content area to assess for gender differences in funding success rates, which are important to identify whether gender disparity existed in funding. This information is important to address potential sources of bias in the review process. What did the researchers do and find? We retrospectively reviewed 55,700 grant and 4,087 personnel award applications submitted to the Canadian Institutes of Health Research over a 15-year period. We found that compared with men, women were significantly less likely to be awarded grants and New Investigator personnel awards. Additionally, we identified significant differences in the gradients of grant success in specific research content areas including Cancer, Circulatory and Respiratory Health, Health Services and Policy Research, and Musculoskeletal Health and Arthritis favoring men scientists. Regression modelling identified that the effect of gender on grant success rates differed by research community and not time. What do these findings mean? These findings call for heightened awareness in applicants, peer reviewers, and funding agencies to the potential for bias to exist in the evaluation of grant and personnel award applications. Our findings highlight the need for funding agencies to monitor gender differences in grant success rates overall and by research content area and to explore possible explanations for gender disparity when identified. Additional research is urgently needed to explicate the reasons for gender differences in success rates and identify bias-enhancing conditions in the peer-review process.