My PhD explores the training of judges who hear asylum appeals. The project proposes a mixed methods approach. The purpose is to arrive at a well-rounded picture of asylum-related training and its role in supporting judges in adjudicating asylum (in the UK context). I will combine quantitative and qualitative methods, to both trace second-instance evaluations of first-instance judicial decisions (text mining “error of law” findings), and to capture relevant individuals’ views and experiences of training (observation, interviews/ focus groups). The novelty in the proposed approach lies in making a new field fruitful for text mining and mixed methods analysis.