Experiments in Upskilling Researchers

Abstract

It is often part of our job as RSEs to train researchers in computational skills, ultimately improving the quality and robustness of research software. However, delivering training effectively and efficiently to a large group of diverse researchers, with different technical skillsets, interests and research areas is definitely a challenge. At the National Oceanography Centre (NOC), we (the Software Engineering and Artificial Intelligence team) have been part of an effort to develop and improve digital skills throughout our organisation. Therefore, over the last two years, we have been delivering frequent training in a wide variety of computational skills, largely (though not exclusively) aimed at NOC researchers. We have run approximately monthly training workshops and code surgeries, curated self-serve material and are starting to offer very short, one topic training sessions. Much of the workshop materials have been based on Carpentries and Carpentries Incubator lessons, though often adapted to have a more oceanographic or geoscience flavour, alongside some bespoke material. We have run training online, in person and in a hybrid fashion. Over the course of this time, we have accumulated a lot of experience in creating and adapting training materials and delivering training in a variety of forms. This talk will discuss the lessons learnt from that experience: what has worked well, what has not, what we plan to try next and what isn’t worth repeating. This work was supported by the Computational Science Centre for Research Communities (CoSeC) Fellowship programme.