Confidence Intervals for Correlations - SPSS
HTML-код
- Опубликовано: 11 сен 2024
- I demonstrate a way to calculate confidence intervals for a correlation in SPSS. The method I demonstrate is a bit of 'trick' and it should be noted that it only provides approximate lower and upper bound confidence intervals. If you want a more accurate way to calculate correlation confidence intervals, check out this webpage calculator:
how2stats.blogs...
Very helpful! And your webpage with the calculator is a life saver! Thank you for putting all of this together!
Extremely helpful. The online links is indeed more precise!!!
You saved my research. Thank you!
Again, another great video. Thanks for this. On a separate note, SPSS 20 has added new built-in bootstrapping functionality, which seems capable of replicating the same process that you just described in your video. Would be nice if you could create a new series dedicated to bootstrapping describing when it can and cannot be used.
Dear, Prof. Does the link, the spreadsheet, support the correlation from nonparametric data? Thank you so much.
nice method to calculate rho confidence interval. Thank you.
Thank you
I'm not sure which version you are using, but in v.22 there's an option to estimate CI's for correlations by performing bootstrapping. Now the standard amount for bootstrapping is 1000, but it can be adjusted. So, if I adjust the amount of bootstraps performed to be my actual sample size, does this give me the accurate CI?
I tested this method, your regression trick, and your calculator on my data set. For the former it gave me 95% CI's of -.029 to .284. For the regression trick it gave me .001 to .294. Using your calculator, the 95% CI's were .002 to .288.
Omhril The technique demonstrated in this video was used before SPSS introduced bootstrapping as an option. To see how to bootstrap in SPSS, check out ruclips.net/video/9VjzPnoUBJQ/видео.html.
Performing bootstrap on SPSS wouldn't give narrows CIs and maybe as accurate as AMOS?
Well, someone can explain to me, how and why this SPSS method gives different results than on-line calculators?
Thank you for your wonderful insight. If possible can you help me to with the citation of the source for the approximate calculation using SPSS. Is the general approach to construct the approximate confidence intervals for correlation based on Zhou (2007) paper "Towards using confidence intervals to compare correlations" in Psychological Methods (Volume 12, No. 4, 399-413) or it is from another source. I would appreciate if you can help me to identify and acknowledge the source of the calculation algorithm. Thank you very much in advance.
Hi - what about using bootstrapping? I see that I get two different CI results if I use this method vs. bootstrapping).
Yes, bootstrapping will give you different results, as it is a different estimation technique. bootstrapping is increasingly popular.
can this method be also used to obtain the Spearman correlation intervals or only for Pearson correlation?
www-01.ibm.com/support/docview.wss?uid=swg21478368 Here You had info about it: Altman and Gardner (2000, p. 90-91) argue that the Fisher Z methods for computing confidence intervals for Pearson correlations can also be applied to Spearman Rank correlations as the distributions of the two correlations are similar. www.tbrieder.org/epidata/course_reading/b_altman.pdf
can we change the confidence interval if we use pearson correlation?
Technically, yes, but if you use alpha .05, it only makes sense to use 95% confidence intervals. It's up to you. Btw, I show a more accurate way to obtain confidence intervals in SPSS for Pearson's r in my textbook: ruclips.net/video/zWCgR3qsWy0/видео.html
you could also just bootstrap bruh
True, but not everyone has access to the bootstrap utility (plus, SPSS Version 25 has bug in the bootstrapping utility).