‘Citizen identification’: online learning supports highly accurate species identification for insect‐focussed citizen science

Citizen science is widely used in ecological research. Data verification of citizen collected data continues to be an issue, and confirming accurate species identification reported by citizens can be especially difficult. Here, we determine the efficacy of using remote learning to identify UK social wasp (Vespinae) species. Citizen scientists (N = 559) collected wasps and identified specimens to species level using a series of online videos and support material. A pre‐ and post‐identification questionnaire, and a post‐identification assessment test, obtained both qualitative and quantitative data for engagement and changes in identification skills. Some (13.5%) of the participants sent their samples in for expert verification of species identification. Self‐assessed skill ratings increased from 2.2/5 pre‐identification to 3.5/5 post‐identification process. Identification accuracy was high, with 85.6% of assessment test images and 96% of the verified specimens being identified correctly. In previous years, face‐to‐face public ID workshops with expert trainers yielded an identification accuracy of 91.3%. Eighty‐seven percent of participants reported enjoying the experience and would take part again. Remote learning of identification skills in non‐specialists can produce greater identification accuracy than face‐to‐face expert‐led workshops, with lower resource requirements and enhanced engagement. Remote learning of identification skills in non‐specialists can produce greater identification accuracy than face‐to‐face expert‐led workshops. Online learning is effective for educating and engaging the public; 87.6% of participants said they enjoyed the experience and wanted to do it again. Self‐identification in identification‐based projects could greatly reduce time and resource requirements for such projects and thus improve their long‐term sustainability.