Hi everyone,
I am running some linear regressions (3 way interactions) with bootstrapping (1000 samples with BCA correction) and all of the main analyses run fine. However, when I run a kind of robustness check with a slightly different variable as a predictor (and the same 3 way interactions) the bootstrapping does not complete the full 1000 samples (around half are actually completed) and I do not get any error messages (such as insufficient memory) in the output. I 've tried different things (such as skipping the BCA method and just doing the default percentile method, or tried setting seeds (different ranges)) but still haven't managed to get
the bootstrap to complete the full 1000 samples. Any ideas of how to find what might be going wrong there?
Thank you in advance
Nikolaos
Locked
