Female students and those with a low socioeconomic status (SES) typically score lower in assessments of self-efficacy and ability in science, technology, engineering, and mathematics (STEM). In this study, a cohort of over 200 UK students attended an intensive, active learning, physics workshop, with pre- and postassessments to measure both physics self-efficacy and physics ability before and after the workshop. Our control took the form of material that was closely related but not covered during the workshop. Students benefited from attending the workshop, as self-efficacy and ability increased significantly in the post-test, with the material not covered showing the smallest increase as expected. A significant socioeconomic attainment gap in ability was completely alleviated for questions on material covered at both secondary and upper secondary level, but not for questions on material seen at upper secondary only. In contrast, although no overall significant initial gender gap in ability was found, despite female students having a lower mean score than male students, a gender gap was alleviated for material seen only at upper secondary level. Female and low SES students' physics ability improved more than male and high SES students' physics ability, respectively. The workshop particularly benefited students from a mildly underperforming demographic tackling the hardest questions, or students from a significantly underperforming demographic tackling intermediate questions but not the hardest questions. The already high levels of confidence in their abilities felt by the cohort (which was boosted further by the workshop) meant that none of the demographics considered were less self-efficacious than their peers; however, the self-efficacy of female students improved more than male students, but of high SES students more than low SES students.