Teaching statistical inference at the University level: the Italian experience

The reproducibility crisis in psychology has been attributed to a faulty application of the NHST paradigm. Some researchers have advocated and adopted other methods, which calls for changes in university-level statistics courses. Instructors should de-emphasize NHST and teach new approaches such as effect size estimation and Bayesian statistics. A multi-phase investigation was conducted of statistics courses for psychology students at Italian Universities to determine how these courses are evolving and the problems these changes pose for students. Phase 1: an analysis of the course contents and required texts listed on university websites. Phase 2: a survey of the time devoted to traditional and newer analysis methods in these courses. Phase 3: interviews with instructors regarding the topics that are proving most difficult for students