I want to show you an example of some software designed to help you be more productive and consistent in applying statistical tests. It’s designed to work along side a oneway analysis of variance and will automatically assess the assumptions associated with the analysis.
I call it a oneway advisor – which is a slight misnomer – it checks the assumptions but doesn’t give advice (so hopefully no angry emails from your in-house statisticians!). The point of showing you this software is to inspire you to write it yourself! Think of this as step zero – an overview of what the end product will look like – then together we will start with a blank piece of “paper” and write the code step-by-step. Associated with a oneway analysis of variance are the “standard” statistical assumptions:
- Errors are normally distributed
- Errors have uniform variance
- Errors are independent
The oneway advisor checks these assumptions and provides the following summary:
Four checks – the first assumption can be tested by looking at both the within-group data as well as the overall data.
The icons provide a simple visual status of these assumptions and hovering will provide a summary of the test.
Over the coming days and weeks I will be providing updates to this blog that describe how to write the JSL to build this Oneway Advisor. I hope you are ready for an exciting journey. And if you know anyone else who would be interested in a step-by-step guide to writing JSL please let them know about this blog!