My approach to stats relies on three things: (1) visualizations, (2) estimation, and (3) the general linear model approach. The last of these is the most controversial. Here I explain why it's necessary.
*Correction: you currently cannot do chi square in visual modeling. Sorry!
Literature cited:
* visuals allow people to make sophisticated inferential decisions without stats training: https://graphics.cs.wisc.edu/Papers/2014/CG14/Preprint.pdf
* standard stats training makes people worse at some decision-making: https://pubsonline.informs.org/doi/10.1287/mnsc.2015.2212
And here's a paper I wrote about my eight step approach to data analysis: https://psyarxiv.com/r8g7c/
Undergraduate curriculum playlist (GLM-based approach): https://www.youtube.com/playlist?list=PL8F480DgtpW_v1fmBauNMPF9Gqdoaa8zJ
Graduate curriculum playlist (also GLM-based approach): https://www.youtube.com/playlist?list=PL8F480DgtpW8gi4XtViwqBYbA4nIjAoUe
Exonerating EDA paper: https://psyarxiv.com/5vfq6/
Download JASP (and visual modeling module): www.jasp-stat.org