Investigating change in subjectivity: The analysis of Q-sorts in longitudinal research

A growing number of studies in applied linguistics have used Q methodology to systematically explore language learners’ and teachers’ subjectivity (e.g., opinions, beliefs, identity, emotions). However, very few studies have used Q methodology and its technique to investigate change over time or after an intervention, partly because of a lack of guidelines on how to analyse Q-sorts completed by the same subjects at different time points. This article aims to provide guidance on how paired Q-sorts can be analysed and to discuss the opportunities and challenges in using the Q-sort method within repeated-measures research designs. To achieve this aim, a systematic review of longitudinal studies (observational and experimental) across research fields was conducted. Twenty-five studies that met the inclusion criteria were selected and scrutinised. Although the review revealed a high degree of heterogeneity in the approaches used to analyse paired Q-sorts, these could be grouped into three broad categories, labelled “Q-factor analysis”, “descriptive statistics” and “inferential tests”, which are summarised and critically discussed. Based on the findings of the review and in line with the principle of holistic inquiry characterising Q methodology, a mixed-methods analytical approach is then proposed and exemplified using a dataset from a quasi-experimental study on pre-service teachers’ beliefs about multilingualism.