For Chi-Square test we present results, for example: Chi (1, n=436)=0,34 p=0,56 But how we sholud present results in Fisher's Exact Test bearing in mind that there's not calculated Value, only Exact. Sig. (which is p)?
ah, of course. my error. I am comparing frequencies among 7 groups, 14 cells total. 5 of the cells have low expected frequencies, so I wasn't sure if the Pearson value was the appropriate p to report.
It looks like it assumes 2/2 tables.... is there any test I can use for bigger tables when assumption of expected values is violated (more cells less than 5)
I actually calculated every single table that is possible (factorials, odds ratios, and EVERYTHING) when I should have used SPSS to do this instead.... FML.
So if I have a 4x6 table with me, with two variables, say income & age, then I will have 24 comparisons...yes? And my alpha will be = 0.05/24=0.002. Is this correct?
thank for your reply. can the likelihood ratio make up for it in case of a bigger table ? A video by someone who claims to cite andy field says yes, but it s the only video that claims we can use likelihood ration forbigger tables.. pliz help me on this. thank you ver much
+Saber Khelifi You could use the likelihood ratio in that case but this is best used with smaller sample sizes. With larger sample sizes the result will approximate with Pearsons Chi-Square.
When I run this in SPSS 20, Fischers Exact does not show up automatically the way it does in your demo- only Pearson Chi-Square and Likelihood Ratio show up. any thoughts?
Hey, i got a questión related to the P-value. if its less than 0.05 its been said that we reject the Hypothesis of no relation of the variables, but that 0.05, what does it mean? the % of risk level? So what about if its 0.07, we could say the Hypothesis of no relation its rejected with a 10% level of risk? thanks btw.
+0maniac0 An example could look something like this: A Fisher's Exact Test was calculated comparing the frequency of heart disease in men and women. A significant interaction was found (X2(1)= 23.80, p
+TheRMUoHP Biostatistics Resource Channel Thanks so much for the quick response. But what do the 'X', the '2' and the '(1)' refer to? Also, what does the value 23.80 correspond to? Thank you and cheers.
+0maniac0 X2 is meant to be the symbol for "Chi Square", the 1 is the degrees of freedom for the analysis and the 23.80 would be the actual chi square statistic produced by SPSS.
+TheRMUoHP Biostatistics Resource Channel Oh, I didn't know you could use the "Chi Square" value when presenting the results of a "Fisher's Exact Test of Independence"... So basically, the presentation is exactly the same as the "Chi-Square Test for Independence", except for the p value, correct? Thanks.
Hello,
You forgot to say that you need to go to "Exact" option and choose, Exact Test.
If it doesn't show automatically you have to click on the Exact button (in the crosstabs window) and check exact again.
+Sacbé A. Ibarra Escobedo
Thanks for this revelation
For Chi-Square test we present results, for example:
Chi (1, n=436)=0,34 p=0,56
But how we sholud present results in Fisher's Exact Test bearing in mind that there's not calculated Value, only Exact. Sig. (which is p)?
Good explanation, but I cannot find the Fisher's Test either in the statistics cell
Very nice, simple explanation. thank you!
ah, of course. my error. I am comparing frequencies among 7 groups, 14 cells total. 5 of the cells have low expected frequencies, so I wasn't sure if the Pearson value was the appropriate p to report.
This was so helpful!
then if the value I found is ,000 what does that mean? have done a mistake through the process? Thank you for answering .
very explicit thank you
It looks like it assumes 2/2 tables.... is there any test I can use for bigger tables when assumption of expected values is violated (more cells less than 5)
+S. Khelifi It does assume that. I am not aware of an alternative for tables larger than 2x2.
how about likelihood ratio?
I can't find Fishers' Exact test on my corss-tab window in SPSS. please help!!
Thank you all
+S. Khelifi You should find it in the Chi Square table in your analysis output. You will only see this if you have a 2x2 table.
Thanks!!!
You're Welcome!
I actually calculated every single table that is possible (factorials, odds ratios, and EVERYTHING) when I should have used SPSS to do this instead.... FML.
thank you
Nice! one question please: in crosstabs when we click on exact what should we check?
So for a table with more than 2 equal/unequal rows-coloums, How should we go about it??
Which version SPSS is this? As I am using version 20, I dont find the Fisher Exact test in the output.
So if I have a 4x6 table with me, with two variables, say income & age, then I will have 24 comparisons...yes?
And my alpha will be = 0.05/24=0.002. Is this correct?
Tnx a lot! :)
thank for your reply. can the likelihood ratio make up for it in case of a bigger table ? A video by someone who claims to cite andy field says yes, but it s the only video that claims we can use likelihood ration forbigger tables.. pliz help me on this. thank you ver much
+Saber Khelifi You could use the likelihood ratio in that case but this is best used with smaller sample sizes. With larger sample sizes the result will approximate with Pearsons Chi-Square.
When I run this in SPSS 20, Fischers Exact does not show up automatically the way it does in your demo- only Pearson Chi-Square and Likelihood Ratio show up. any thoughts?
Hey, i got a questión related to the P-value. if its less than 0.05 its been said that we reject the Hypothesis of no relation of the variables, but that 0.05, what does it mean? the % of risk level? So what about if its 0.07, we could say the Hypothesis of no relation its rejected with a 10% level of risk?
thanks btw.
That value refers to the probability that the results we see are due to chance, in the case of .05 a less than 5% probability.
And this is good when I have some cells
Can we use Fisher's Test with 3x3?
I am using v 19 can't get fisher
+zaeset zaesetv I also would like to know how to report the results from a Fisher's exact test. Could anyone enlighten me?
+0maniac0 An example could look something like this: A Fisher's Exact Test was calculated comparing the frequency of heart disease in men and women. A significant interaction was found (X2(1)= 23.80, p
+TheRMUoHP Biostatistics Resource Channel Thanks so much for the quick response. But what do the 'X', the '2' and the '(1)' refer to? Also, what does the value 23.80 correspond to?
Thank you and cheers.
+0maniac0 X2 is meant to be the symbol for "Chi Square", the 1 is the degrees of freedom for the analysis and the 23.80 would be the actual chi square statistic produced by SPSS.
+TheRMUoHP Biostatistics Resource Channel Oh, I didn't know you could use the "Chi Square" value when presenting the results of a "Fisher's Exact Test of Independence"...
So basically, the presentation is exactly the same as the "Chi-Square Test for Independence", except for the p value, correct?
Thanks.
+0maniac0 Yes, that is true.