A survey of Canadian computer science faculty was carried out in April-May 2018 to identify which funding programs are most useful in supporting their research. There were 261 respondents (28% response rate). The results of the survey show clearly that the NSERC Discovery Grant program is the most important funding source for Computer Science researchers. It was rated “Vital” or “Very Useful” by 244 (93.5%) of the respondents and has been used by 97.7% of the respondents. The second most important is the NSERC Scholarships programs, rated “Vital” or “Very Useful” by 67.8% of the researchers and used by 93.5% of them. Innovation support programs (such as NSERC Partneships and Mitacs) were appreciated highly by 35-40% of participants rating them as “Vital” or “Very Useful”, and 70-80% of the researchers having held at least one such grant.
The recommendation resulting from this survey is that more funding needs to flow into the NSERC Discovery Grant Program for Computer Science, since it is the most important funding source for our discipline.
Participants expressed concern in their comments that there was a tremendous growth in the Computer Science area, with many new brilliant researchers joining faculty ranks, but the funds allocated to the Computer Science Evaluation Group within the Discovery Grant program have not changed proportionally. This leads to very high competitiveness in the grant allocation process, leading to lower grant amounts, insufficient to attract good students and maintain a competitive research program.
In addition to evaluating the funding programs, the participants were asked to evaluate the online services for grant applications. The results show that the majority of the respondents find the CommonCV as very hard to use, suggesting that the attempt to create a universal format for CV across all government funding agencies does not serve well computer science researchers. For example, conference proceeding papers need to be reported under the category “Presentations”, generating excessively long lists, where it is hard to highlight publications of particular importance or competitiveness. A resulting recommendation is to replace this category with more differentiated categories for the NSERC version of CCV, such as “peer-reviewed conference publications”, “peer-reviewed workshop publications”, “posters and other presentations”, which fit better the venues preferred by computer science researchers as outputs for their research results.
Detailed results from the study are available here.