I agree, ranking everything give you a 0.9 rho every time. I followed along with your tutorial and I got a very weird outcome that wasnt realistic at all to my assignment - and it was the wrong answer.
But what if i have a lot of the same numbers in X, the RANK.AVG will come up with many of the same numbers in the rank(x) all the way down. What can be done?
Hello, I am trying to use Spearman´s correlation for my likert scale survey. Your video is super useful, thanks so much so I guess I am able to do it but I still have one question, the values of X and Y (I understand they are the studied variables), where do they come from? are they the median of the answers? I am so confused, I find everywhere how to run the analysis but not how can use the data from the likert scale (1 to 5) in the actual analysis. I would really appreciate any kind of help.
Excellent tutorial. As a novice to stats, may i just ask ...what would be the best correlation test for surveys using Likert scales? I tried using Cronbach's alpha using Google sheets Toolpak. However a lot of my MS error values are coming up = 0. So it is impossible to compute alpha via = 1 - (MS error/ MS rows) because the calculation is coming up to infinity. What shall I do? Please help. TIA
Sorry for the late reply! Spearman's would be an appropriate correlation to use, because it makes no assumption about the underlying relationship between the variables (i.e the relationship need not be linear, as is required by Pearson's correlation). Spearman's is also not affected by how you code the scale(s).
I believe you get a slightly different answer from Pearson's because, according to Wikipedia, the formula you are using for Spearman's does not work if there are ties in the data, as there are in your example.
When you ordered the data, you messed everything up because the data are no longer paired. It’s all wrong.
Hi Deana - I don't believe so. Because there are no blank columns, excel moves rows around, not just cells in a single column
I agree, ranking everything give you a 0.9 rho every time. I followed along with your tutorial and I got a very weird outcome that wasnt realistic at all to my assignment - and it was the wrong answer.
@@QUAN-ul8py it did move around cells for me, hm.
If you look carefully, whenever he orders the data, cells in the same row remain together afterwards so everything is fine
But what if i have a lot of the same numbers in X, the RANK.AVG will come up with many of the same numbers in the rank(x) all the way down. What can be done?
why in my excel there is no function "=rank.avg(...)" ? its only "=rank(...)" im using excel 2007. can you help me please? thank you before
How did you copy that down?
Hello, I am trying to use Spearman´s correlation for my likert scale survey. Your video is super useful, thanks so much so I guess I am able to do it but I still have one question, the values of X and Y (I understand they are the studied variables), where do they come from? are they the median of the answers? I am so confused, I find everywhere how to run the analysis but not how can use the data from the likert scale (1 to 5) in the actual analysis. I would really appreciate any kind of help.
why is it become 13^3? please enlighten me
I cannot find rank.avg function of my excel please sir help me how to add rank.avg function
Excellent tutorial.
As a novice to stats, may i just ask ...what would be the best correlation test for surveys using Likert scales?
I tried using Cronbach's alpha using Google sheets Toolpak.
However a lot of my MS error values are coming up = 0.
So it is impossible to compute alpha via = 1 - (MS error/ MS rows) because the calculation is coming up to infinity. What shall I do? Please help. TIA
Sorry for the late reply! Spearman's would be an appropriate correlation to use, because it makes no assumption about the underlying relationship between the variables (i.e the relationship need not be linear, as is required by Pearson's correlation). Spearman's is also not affected by how you code the scale(s).
Informative video.
Finally!!! Thank you
thank you thank you thank you!!!
Seems to be giving me the wrong answers
great help thanks
Cheers! Very helpfull!
I believe you get a slightly different answer from Pearson's because, according to Wikipedia, the formula you are using for Spearman's does not work if there are ties in the data, as there are in your example.