A Second Edition of ITNS? Here’s the Latest

Our first blog post about a possible second edition of ITNS is here. All the comments I made there, and the questions I asked, remain relevant. We’ve had some very useful feedback and suggestions, but we’d love more. You could even tell us about aspects of ITNS that you think work well. Thanks!

Meanwhile Routledge has been collecting survey responses from users, adopters, and potential adopters of ITNS. We are promised the collated responses in the next few weeks.

Back to that first post. Scroll down below the post to see some valuable comments, including two substantial contributions from Peter Baumgartner.

Peter has also been in touch with us to discuss various possibilities and, most notably, to send a link to some goodies he has sketched. One of Peter’s interests is effective didactical methods, and he has made positive comments about some of the teaching strategies and exercises in ITNS–especially those that prompt students to take initiative, explore, and discover. He’s keen to see development of these and so, as I mentioned, he has sketched some possibilities.

Peter’s Goodies

Peter’s sketches start here. Then he has a whole sequence of goodies that start here. He asks me to emphasise that he doesn’t claim to be an expert programmer and that he’s largely using tools (including shiny on top of R, and the R tutorial package learnr) that are rather new to him. (So one of the lessons for us is that it may (!) be comparatively easy to use such tools to build new components for ITNS.)

The figure below is from a shinyapp that is one of the last of Peter’s goodies, and shows how a stacked dot plot and a histogram can be linked together, as the user changes the number of bins. ESCI has a version of this, and also allows clicking to display mean, median, and various other aspects of the data set. Peter emphasises that it should be comparatively easy to build such displays in shiny, with the advantage of the full power and flexibility of R.

Thanks Peter! For sure, finding a good way to do more in R is a priority for us. Should we try to replace ESCI entirely? Retain ESCI for some simulations that are useful for learners, while transferring more of the data analysis to R? Is jamovi the way to go, perhaps alongside shiny apps? What’s the best way in R to build detailed, appealing, graphics–with onscreen user controls–such as ESCI attempts to provide?

Please, let’s have your thoughts, either below, or via email to Bob or me. Thanks!

Geoff

Geoff g.cumming@latrobe.edu.au
Bob rcalinjageman@dom.edu