We agree that societal impact should be rated more highly in scientific publishing and research evaluation (Nature 553, 5; 2018). To this end, we suggest that ways to achieve it should be introduced as an important component of curricula at higher-education institutions.
Degree theses and university classes in academic publishing are generally structured with bibliometric output in mind because that is the main driver of tenure and promotion. They rarely touch on the merits of societal impact. This monotheistic evaluation of academic pursuits means that few faculty members make time for public outreach. Even new journals for applied excellence address a symptom of insufficient societal impact in research, not the cause.
Instead of relying solely on papers and citations as proxies for impact, funders and research organizations need to broaden their assessments of scientific output. They should also acknowledge other important outputs, such as developments in scientific products and services, important data sets, platforms and software, as well as their influence on policy.
Nature 554, 300 (2018)