OZCOTS 2010 Proceedings of the 7 th Australian Conference on Teaching Statistics

In this talk we gaze through the ripple glass of a bathroom window and wander Alice-like down garden paths through a wonderland where what we see is never quite the way it really is. The paths our odyssey leads us along are conceptual pathways that start with conceptualisations of statistical inference that are intended to be accessible to, and operable by, students mid-way through high school and lead us, via a series of connected trails, all the way to plot annotations that better reveal the stories being told by factor variables in generalised linear models. Along the way, both motivating and suggesting ways forward for all of this, we meet novel visualisations of sampling variation, resampling variation and randomisation variation. The talk draws on a paper with Maxine Pfannkuch, Matt Regan and Nicholas Horton entitled, “Towards more accessible conceptions of statistical inference” read to the Royal Statistical Society late in 2010 and on other work on making inference more accessible, particularly via visualisations, with these and other collaborators.