Hello. I want to use the rank sign test to compare ordinal data. My data is conceptions of nature of science and the categories are naive, a mixture of naive and informed, and informed. Am I able to use the test for this? Also, I am comparing a pre, post, delayed post test, so can I do different comparisons between pre and post, pre and delayed post, and post and delayed post?
Thank you for your video. Could you please add how to report the results of the first test (related samples) in APA style? I don't understand what is the "Test Statistic" and whether I should report both that and the "standardized test statistic".
Hi Dr E - Great video. Just wondering if that data has to be set to 'scale' to use this non-parametric test. I believe I have ordinal data. I am comparing the aggultination strenghts of manual vs automated antibody screen results. Can this test still work if i change to ordinal data instead of scale? Faye
Hi Faye, Many thanks for your kind comment. Unfortunately I can't check this out for you. I am retired (nearly 4 years) and no longer have a licence for SPSS. Dr E.
Hi Death ray, I don't use SPSS any more (I'm retired!), but I guess there is somewhere in the settings where you can change the number of decimal places. However, you really don't need to see more decimal places if your result is "0.000". Simply report as "p < 0.001". Hope this helps, Dr E.
Hi Bhoj, You are correct, in the Wilcoxon test the smaller value is the one you choose for reporting. In SPSS it is a good idea to check the order in which you enter your data. In my example I chose August/September which gave a test statistic of 75, if I have done this the other way around (September/August) the result would be a test statistic of 16. Hope this helps, Dr E.
you are amazing man ty for the information
great video! just a question: is that footnote about the test being based on negative ranks useful for the interpretation? what does that tell us?
Hello. I want to use the rank sign test to compare ordinal data. My data is conceptions of nature of science and the categories are naive, a mixture of naive and informed, and informed. Am I able to use the test for this?
Also, I am comparing a pre, post, delayed post test, so can I do different comparisons between pre and post, pre and delayed post, and post and delayed post?
Thank you for your video. Could you please add how to report the results of the first test (related samples) in APA style? I don't understand what is the "Test Statistic" and whether I should report both that and the "standardized test statistic".
Hi Dr E - Great video. Just wondering if that data has to be set to 'scale' to use this non-parametric test. I believe I have ordinal data. I am comparing the aggultination strenghts of manual vs automated antibody screen results. Can this test still work if i change to ordinal data instead of scale? Faye
Hi Faye,
Many thanks for your kind comment.
Unfortunately I can't check this out for you. I am retired (nearly 4 years) and no longer have a licence for SPSS.
Dr E.
@@EugeneOLoughlin No worries at all. Thank you for your response anyway!
How can I increase the decimals shown for p? I have 0.000 but I would like to see the complete number!
Hi Death ray,
I don't use SPSS any more (I'm retired!), but I guess there is somewhere in the settings where you can change the number of decimal places.
However, you really don't need to see more decimal places if your result is "0.000". Simply report as "p < 0.001".
Hope this helps,
Dr E.
Why in two test the test statistics is taken 75 rather than 16, isn't the test statistics is min of Tnegative and Tpositive?
Hi Bhoj,
You are correct, in the Wilcoxon test the smaller value is the one you choose for reporting. In SPSS it is a good idea to check the order in which you enter your data. In my example I chose August/September which gave a test statistic of 75, if I have done this the other way around (September/August) the result would be a test statistic of 16.
Hope this helps,
Dr E.
@@EugeneOLoughlin Hello Eugene, Thanks for clarification.
Sir how can I download dataset of this kind to test in SPSS