Can you elaborate how you want to compare them? What do you want to compare, this seems unclear to me.

I am planning to take up dummy data from "School oral health promotion perspective" and illustrate rationale of choosing ANOVA test versus chi square test for different types of variables within that dear Rainer!

Mh, I am still puzzled, since the purpose of those tests is quite different in my opinion. Do you want to generate an example, when to use what test (like in any statistics textbook...) or are you trying to answer the same research question / hypothesis with ANOVA and chi square test? If the latter, please elaborate how this could be done, maybe with a simple example?

Dear Rainer,

The background for this paper would be that I'm working with several postgraduates for their statistical analysis. They often come with very fundamental doubts as and what are the different criteria is for choosing a statistical analysis. Simple doubts like as this one is not addressed with their research papers so uncovering the fundamentals.

This paper compares the Chi-square test and ANOVA to help school-based oral health programs analyze different types of data, like Oral hygiene related health behaviours versus measurable outcomes (e.g., Dental caries counts - based on the type of index used). It guides researchers on when to use each test for accurate analysis and better-informed health interventions.

Key Takeaways from the paper would include the following:

I will also be covering fundamentals of the tools that are commonly used and the approach to use the same with screenshots of the steps where needed.

With regards,

Anitha

Dear Anitha,

while I value your idea and intention to clarify when to use which test, I am also afraid that your endeavor is doomed to fail, if you want to produce one single paper.

There are dozens or even hundreds of statistical textbooks and none of them is able to cover each individual decision, when to select which test. Even worse, although they have textbook length, lots of them reduce statistics to simple cook book strategies, with simple decision criteria and simplified decision trees. This is really not how it works imho and is misleading for students (maybe one of the worst examples you can find everywhere: "If your data is not normally distributed, apply a Mann-Whitney U test, instead of a t-test", this is simply a nonsense advise).

I have already problems with your first "key takeaway": "Chi-square is ideal for categorical data [...] while ANOVA is best for continuous data [...]", since this is already simply not true. There are lots of continuous data, which are clearly not suitable for an OLS ANOVA, at least an ANOVA would not be "optimal".

If you want to cover "fundamentals of the tools" of the tools, there are lots of books and resources available already out there.

So, I am really not sure if it is a realistic goal what you are trying to achieve.

All the best

Raine