Thanks a lot for sharing. Could you please let me know if you have videos that demonstrate how to run macro developed by Andrew F. Hayes and Preacher for mediation on SPSS? Thank you very much!
Hi, thanks alot for all the explanations and fixes. No way I could've figured this one out! One question: is it plausible to report both Spearman correlation coef. and bootstrapping results (as described here) for non-normally distributed data? Both are significant, however, Spearman's coefficients are weak. Will bootstrapping support the relationship "stronger"? Or is this not the way to go...
Short Answer: Go with the bootstrapped Pearson correlation. Long Answer: Bootstrapping a Spearman correlation is not really in the spirit of bootstrapping, because the Spearman correlation does not assume normally distributed data. Also, keep in mind that the Spearman correlation answers a slightly different question to the Pearson correlation. The Spearman correlation tests the hypothesis of an association between ranked data. Most people are not really as interested in that hypothesis as the raw data hypothesis association tested by the Pearson correlation.
how2stats thanks. Again, any sense in reporting both non-bootstrapped Spearman and bootstrapped Pearsons? We are talking about clinical (healthcare) manuscripts. I think I never saw bootstrapping being reported. However, it seems perfect for my study: 87 participants, 4 variables, non-normal distribution, significant Spearman for 3 variables, but all 4 significant in bootstrapped Pearson. I think I should report both.
Hello thanks for the information. I use SPSS 25 which has the bug. can you tell me please, which syntax I should write to perform non-parametric bootstrap?
Hi, thank you for the informative video! Do you also have a syntax command for a two-way ANOVA? I have the "bug" Version of SPSS 25 and should, in theory, be able to bootstrap, but it does not work via the command window.
I don't think SPSS can (yet) bootstrap the results of of a two-way ANOVA through the GLM procedure. However, if you check out my fairly recent video on conducting a contrast interaction analysis in SPSS, you should be able to bootstrap that (assuming it's a 2x2 ANOVA).
Finding this fix was extremely helpful!
Had the same Bug, it worked. HEUREKA! Thank you very much!
This was diving me crazy, thanks a lot!
Thanks a lot for sharing. Could you please let me know if you have videos that demonstrate how to run macro developed by Andrew F. Hayes and Preacher for mediation on SPSS? Thank you very much!
WHICH VERSION OF SPSS ARE YOU USING? BECAUSE I HAVE 25 AND SYNTAX PASTE IS NOT HELPING. I WANT TO CHANGE IT. PLEASE LET ME KNOW.
thanks help a lot, i just need to change your variables to mine. running smooth
Hi, I have a question about how to report z scores and mean in a results section
Hi, thanks alot for all the explanations and fixes. No way I could've figured this one out! One question: is it plausible to report both Spearman correlation coef. and bootstrapping results (as described here) for non-normally distributed data? Both are significant, however, Spearman's coefficients are weak. Will bootstrapping support the relationship "stronger"? Or is this not the way to go...
Short Answer: Go with the bootstrapped Pearson correlation. Long Answer: Bootstrapping a Spearman correlation is not really in the spirit of bootstrapping, because the Spearman correlation does not assume normally distributed data. Also, keep in mind that the Spearman correlation answers a slightly different question to the Pearson correlation. The Spearman correlation tests the hypothesis of an association between ranked data. Most people are not really as interested in that hypothesis as the raw data hypothesis association tested by the Pearson correlation.
how2stats thanks. Again, any sense in reporting both non-bootstrapped Spearman and bootstrapped Pearsons? We are talking about clinical (healthcare) manuscripts. I think I never saw bootstrapping being reported. However, it seems perfect for my study: 87 participants, 4 variables, non-normal distribution, significant Spearman for 3 variables, but all 4 significant in bootstrapped Pearson. I think I should report both.
@@how2stats WHICH VERSION OF SPSS ARE YOU USING? BECAUSE I HAVE 25 AND SYNTAX PASTE IS NOT HELPING. I WANT TO CHANGE IT. PLEASE LET ME KNOW.
Hello thanks for the information. I use SPSS 25 which has the bug. can you tell me please, which syntax I should write to perform non-parametric bootstrap?
Hi, thank you for the informative video! Do you also have a syntax command for a two-way ANOVA? I have the "bug" Version of SPSS 25 and should, in theory, be able to bootstrap, but it does not work via the command window.
I don't think SPSS can (yet) bootstrap the results of of a two-way ANOVA through the GLM procedure. However, if you check out my fairly recent video on conducting a contrast interaction analysis in SPSS, you should be able to bootstrap that (assuming it's a 2x2 ANOVA).
It really works, Thank u sir
Hello, thank you for the informative video. Do you know how to bootstrap for ROC curves in SPSS? Thank you!
PS. I should also add that I have the bug version of SPSS 25... Thank you!
THANK YOU!
Ehrenmann
Thank you sir