This paper is from Session 5H: Statistics Education Across Disciplines: pedagogy and assessment for statistical literacy, critical thinking, and evidence-based practice
which comes under Topic 5: Statistics education in the workplace, government and across disciplines
Paper 5H2 (Friday 13th, 11:00-12:30)
A short classroom-based workshop on Latent Class Analysis
Presenter
- Daniel Green (University College London, United Kingdom)
Co-author
- Eirini Koutoumanou (University College, London, United Kingdom)
Abstract
Latent Class Analysis (LCA) is a statistical technique that differs to standard regression methods by attempting to identify patterns not directly observed within a population. LCA courses in the United Kingdom focus primarily on statistical software whilst detailed underlying theory is given relatively minor coverage. Our aim was to develop a non-computer based LCA course for non-statisticians that encourages a solid understanding of the theory by using lay language, thus providing a better basis to progress to utilising software and an improved overall student learning outcome. The target audience consists of participants that might never apply these techniques themselves in addition to those that may progress to application. In this presentation, I discuss student feedback and attitudes to the classroom-based approach.