Eleanor M Pullenayegum1,2, Robert W Platt3, Melanie Barwick4, Brian M Feldman1,5, Martin Offringa5, Lehana Thabane6. 1. Child Health Evaluative Sciences, Hospital for Sick Children, ON, Toronto, Canada. 2. Dalla Lana School of Public Health, University of Toronto, ON, Toronto, Canada. 3. Department of Epidemiology & Biostatistics, McGill University, Westmount, QC, Canada. 4. Community Health Systems Resource Group, Hospital for Sick Children, ON, Toronto, Canada. 5. Institute of Health Policy, Management and Evaluation, University of Toronto, ON, Toronto, Canada. 6. Department of Clinical Epidemiology and Biostatistics, McMaster University, ON, Hamilton, Canada.
Abstract
BACKGROUND: The use of standard statistical methods in the medical literature has been studied extensively; however, the adoption of new methods has received less attention. We sought to understand (i) whether there is a perception that new methods are underused, (ii) what the barriers to use of new methods are, (iii) what dissemination activities are used, and (iv) user preferences for learning about new methods. METHODS: We conducted a cross-sectional survey of members of the Statistical Society of Canada (SSC) and of principal investigators (knowledge-users) funded by the Canadian Institutes of Health Research (CIHR). RESULTS: There were 157 CIHR respondents (14% response rate), and 39 respondents were statisticians from the Statistical Society of Canada. Seventy percent of CIHR respondents and 82% of statisticians felt that new developments were under-used. Barriers to use of new methods included lack of access to the necessary expertise (selected by over 90% of respondents), lack of suitable software (selected by 81% of statisticians), and lack of time to implement new methods (selected by 78% of statisticians). Greater access to statistical colleagues with an interest in collaboration and availability of software to implement new methods were the top-rated preferences among knowledge-users. CONCLUSIONS: There was a clear perception among all respondents that new statistical methods are underused. Encouraging statistical methodologists to develop a knowledge translation plan for improved dissemination and uptake, placing greater value on the role of the statistical collaborator in research, and providing software alongside new methods may improve the use of newly developed statistical methods.
BACKGROUND: The use of standard statistical methods in the medical literature has been studied extensively; however, the adoption of new methods has received less attention. We sought to understand (i) whether there is a perception that new methods are underused, (ii) what the barriers to use of new methods are, (iii) what dissemination activities are used, and (iv) user preferences for learning about new methods. METHODS: We conducted a cross-sectional survey of members of the Statistical Society of Canada (SSC) and of principal investigators (knowledge-users) funded by the Canadian Institutes of Health Research (CIHR). RESULTS: There were 157 CIHR respondents (14% response rate), and 39 respondents were statisticians from the Statistical Society of Canada. Seventy percent of CIHR respondents and 82% of statisticians felt that new developments were under-used. Barriers to use of new methods included lack of access to the necessary expertise (selected by over 90% of respondents), lack of suitable software (selected by 81% of statisticians), and lack of time to implement new methods (selected by 78% of statisticians). Greater access to statistical colleagues with an interest in collaboration and availability of software to implement new methods were the top-rated preferences among knowledge-users. CONCLUSIONS: There was a clear perception among all respondents that new statistical methods are underused. Encouraging statistical methodologists to develop a knowledge translation plan for improved dissemination and uptake, placing greater value on the role of the statistical collaborator in research, and providing software alongside new methods may improve the use of newly developed statistical methods.
Authors: Adam J Streeter; Nan Xuan Lin; Louise Crathorne; Marcela Haasova; Christopher Hyde; David Melzer; William E Henley Journal: J Clin Epidemiol Date: 2017-04-28 Impact factor: 6.437
Authors: Amy E Wahlquist; Lutfiyya N Muhammad; Teri Lynn Herbert; Viswanathan Ramakrishnan; Paul J Nietert Journal: PLoS One Date: 2018-08-01 Impact factor: 3.240
Authors: Willi Sauerbrei; Aris Perperoglou; Matthias Schmid; Michal Abrahamowicz; Heiko Becher; Harald Binder; Daniela Dunkler; Frank E Harrell; Patrick Royston; Georg Heinze Journal: Diagn Progn Res Date: 2020-04-02