Am trying to find the relation between two nominal variables, I did chi-square test for measuring the independence between two variables, but, one of my variables has a massive frequency in one of its options, and for sure that will affect the score.
am trying to find the relation between "Were requirements gathered by using a particular method?" and "Choose what is the methodology was used."
But, how can I interpret this?!
Please see the attached table
Any ideas
I have attached an Open University chi-square helpful document, and a link to a YouTube tutorial (from mid-way the output and its interpretation is discussed). The Google Books link will bring you to a chapter on how to interpret the chi-square in light of parameter violations etc. from an output, and how to write up the statistical test in the text of your dissertation:
https://www.youtube.com/watch?v=wfIfEWMJY3s
https://books.google.co.uk/books?id=-N2OCgAAQBAJ&pg=PA166&lpg=PA166&dq=spss+for+psychologists+chi-square&source=bl&ots=T4yi5HH8XG&sig=a5TXnDR0WCnqhyG7YilusPGbhtY&hl=en&sa=X&ved=0ahUKEwjM5ovGx8LVAhVkFMAKHZouBBEQ6AEIMDAC#v=onepage&q=spss%20for%20psychologists%20chi-square&f=false
You should exclude the category of "none" ---however, of the 112 who selected "none" in your table, 13 of these were "yes" under 'was there a particular method?'. So you might have some data quality issues: if they used some kind of method, then it should perhaps be "other".
You could do a chi-square test to determine if the percentage of 'yes' for each method is statistically different. By the looks of your data, it appears that the spiral model might have significantly fewer people choosing it than the others, but you would have to do the chi-square test to be sure. And again, exclude the "none" category because obviously there were many datapoints in that category and it would not be meaningful in a chi-square analysis of the other categories.
Agree with @Jerry!
It is better to split the sample (S) into two subsamples: S1 - using methods (n1 = 51) and S2 - not using (n2 = 112).
Why?! There is no meaning to search relation b/n "Requirements gathered by using a particular method" and "None" answers. Of course, it should be explained more precisely.
Then apply chi-square on S1. I think there is a relation (looking through the data of S1).
Regards!
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS