Can you please explain on my there's no t value in the bootstrap result? I've been facing some difficulties in finding the answer related to this problem. Thank you
Great video. I'm wondering, is there a way to access the bootstrap samples themselves? I'd like to see the 1000 or so cases of estimates directly for various reasons.
Stat faculty here. if 0 is in the confidence interval try re-running the analyses and increase the number of samples. If 0 is still in the CI, then I would not trust that your result is significant.
I want to know can we compare statistical significance with bootstrap significance? Making my question more clear, for my regression coefficient, I have b1 = -0.26(p = 0.039) regular way and b1 = -0.026(0.098) with bootstrap. What should I conclude about significance? Which significance is more valued regular significance or bootstrap significance?
I am doing a multiple regression and not bothering to report any confidence intervals (my lecturer didn't say I need to), so do I just select 'percentile' as opposed to BCa if I am not reporting the confidence intervals anyway? - Thanks, great video!
Thanks. Nice video. Which version SPSS do you have? I have SPSS 26 base and I don't find the option. I checked online and it shows that Bootstrap is a feature of SPSS 26 Base. Can you help?
I was wondering the same thing. It looks like SPSS has a "bootstrapping option" that's an add-on. IBM offers a two week free trial, so I am hoping I can get that to work. www.ibm.com/products/spss-statistics/pricing www.ibm.com/support/knowledgecenter/SSLVMB_24.0.0/statistics_mainhelp_ddita-gentopic5.html
Hello, Thank you for the video. Recently, I completed several multiple regression analysis with bootstrap (n = 1000). Upon the write up, I have difficulties for reporting the effect size for my predictors. If I am going to report the effect size from the results, what values should I use? Am I reporting the 'bias' values as the effect size for the predictors? or should I use the unstandardized coefficient beta which are statistically significant in the bootstrap result?
This was very helpful- thanks! Reporting results in a table typically involves B, SE B, Standardised beta and BCa Confidence intervals - is there a way to calculate the beta for the bootstrapped results or are these perhaps the same as the beta values in the regular coefficients table? Thanks!
In a bootstrapped analysis, the point-estimates (e.g., beta-weights) are the same as those you would obtain from the normal theory estimation approach (i.e., the "regular" approach). What changes are the standard errors, the confidence intervals and the p-values.
