Question by Conni: What does it mean if something is significant in Statistics and how do perform the proper significance test?
I am currently studying Chi-squared in my statistics class and after each question they ask if it is significant and to perform the proper test to support the hypothesis. Ok so what I have in my head is that if something is significant P<0.5 and you accept Ho and if it isn't significant you accept Ha and p> or = to 0.5? By perform the proper test do they mean like z-test? or chi-squared test? and one more thing, If something is indeed significant what does that mean?

Best answer:

Answer by Lost in Translation
In a proper test, first you postulate your hypothesis. You always begin with a passive hypothesis which is your H0 (null hypothesis). After that, your next step is to set up your significance level. In my days, for teaching purposes we set it to 5% or 0,05. This determines the sensitivity of your test.

Then you do your Chi square test, (usually done with software nowadays) and compare your results with the one from your Chi square table. If it is greater than that on the table for the corresponding degree of freedom then the H0 is rejected. From there you can draw your conclusion. If you’re using software, reject it when p value is smaller than significance level set.

Add your own answer in the comments!

Technorati Tags: , , , , , , ,