Understanding Reproducibility in the Canadian Research Ecosystem: A cross-sectional survey

We conducted a cross-sectional survey of Canadian-based researchers and academic librarians to examine perceptions and behaviors related to research reproducibility. We randomly selected 2700 researchers funded by the three Canadian federal research funders (the Canadian Institute of Health Research (CIHR), Natural Sciences and Engineering Research Council of Canada (NSERC), and Social Sciences and Humanities Research Council (SSHRC)) in 2023 and 2024 and a random subset of 300 academic librarians identified from a publicly available dataset1. Participants were invited to complete a purpose built anonymous online survey, which collected demographic information, perceptions of reproducibility crisis, perceived contributors to irreproducible research, experiences with institutional training on reproducibility, and perceptions and experiences with open science and research security. A total of 487 (16.2% of total invitees) participants responded to our survey with 337 (11.2% of total invitees) participants completing it. Overall, 65.6% of participants (n=221, 65.6%) perceived some form of reproducibility crisis, with 20.2% (n=68, 20.2%) reporting a significant crisis. The leading perceived contributors to irreproducibility were pressure to publish (n=205, 60.8%), selective reporting of published literature (n=173, 51.3%), and lack of training or resources to support reproducible research (n=133, 39.5%). Only a minority of participants had completed formal training in reproducibility (n=27, 8.0%), or in conceptually related areas of open science (n=75, 22.3%), or research security (n=55, 16.3%).Most participants indicated their institution either did not provide training or that they were unsure of its availability for enhancing the reproducibility of research (n=289, 85.8%), open science practices (n=207, 61.5% )and research security (n=222, 65.9%) . Participants reported a lack of incentives (n=148, 43.9%), insufficient funding (n=107, 31.8%), and institutional prioritization of novel research over replication studies (n=228, 67.7%), with replication studies perceived as harder to fund (n=285, 84.6%), as barriers to reproducibility. These results may inform training, policy, and interventions to strengthen research reproducibility,as well as open science and research security best practices, within the Canadian research ecosystem.