I'm doing hearing test on 70 factory workers to see if the noise in the factory affect the workers to have Noise Induced Hearing Loss (NIHL) or not.
I already got the data, and I'm grouping the workers based on how long they already work.
So the duration of working there: 1-5 Years,6-10 Years,11-15 Years,16-20 Years,21-25 Years <<< This kind of data are Interval (Scale) right?
and the result of hearing test, of course it's Yes(he/she got NIHL) and No(he/she didn't got NIHL).
So the Result of Hearing Test : Yes,No <<< This kind of data are Nominal right?
Now I have H0 and H1 like this:
H0: There is no relationship between working duration on the factory and NIHL. (Factory Noise didnt Affect Workers to got NIHL)
H1: There is a relationship between working duration on the factory and NIHL. (Factory Noise Affect Workers to got NIHL)
Based from a book I read, I should use "Independent Sample T-Test" or "One Way Anova" for Scale vs Nominal <<< Is this right?
And it says that "Independent Sample T-Test" only works on small size of sample(<30) and my Sample is more than 30,so I can't use "Independent Sample T-Test".
The only test I can use is only "One Way Anova", but to use "One Way Anova" it says that the Data Must Be Normal and Homogen.
And here come the problem, my data is not normal (Based on Kolmogorov-Smirnov Test, sig: .000(from what I read,if sig<0.05 = Not normal)) and my data also not homogen (Sig: .000 too)
Then what should I do now?