Hi, Thank you for these exceptionally well done tutorials. Do you know if jamovi has a module which allows you to do multivariate analysis - basically if I want to predict if baseline clinical variables can predict for a given outcome, e.g. response, or survival - I first want to run a univariate analysis - and then include the significant variables in a mutivariable model to confirm the findings Any suggestions would be much appreciated. Thanks,
Can Jamovi be used to conduct a *mixed-effects* multinomial logistic regression? (i.e., including both fixed and random effects)? I can see options to do a mixed-effects binomial logistic regression but not multinomial.
Not necessarily. "Multinomial logistic regression" means that you have an outcome with more than two categories. On the other hand, "multiple logistic regression analysis" means that you you using several predictor variables to classify cases into categories. So, both terms are technically incomplete. It's important to specify whether your outcome has just two categories (binomial logistic regression, which is the most common kind) or more than two categories (multinomial, which is much more complicated). Separate from that, you could specify whether you have just a single predictor (bivariate or simple regression, which would be unusual for logistic regression of either kind) or more than one predictor (multiple regression, which is more common). So, things get wordy. Sorry for the confusion! Bart
This was a fantastic explanation and demonstration. Thank you!
Thank you
Can you provide some guidance on how to conduct a power analysis to determine the sample size for a multinomial logistic regression?
Hi,
Thank you for these exceptionally well done tutorials.
Do you know if jamovi has a module which allows you to do multivariate analysis
- basically if I want to predict if baseline clinical variables can predict for a given outcome, e.g. response, or survival
- I first want to run a univariate analysis
- and then include the significant variables in a mutivariable model to confirm the findings
Any suggestions would be much appreciated.
Thanks,
Can Jamovi be used to conduct a *mixed-effects* multinomial logistic regression? (i.e., including both fixed and random effects)? I can see options to do a mixed-effects binomial logistic regression but not multinomial.
Is a multinomial logistic regression the same as a multiple logistic regression analysis?
Not necessarily. "Multinomial logistic regression" means that you have an outcome with more than two categories. On the other hand, "multiple logistic regression analysis" means that you you using several predictor variables to classify cases into categories. So, both terms are technically incomplete. It's important to specify whether your outcome has just two categories (binomial logistic regression, which is the most common kind) or more than two categories (multinomial, which is much more complicated). Separate from that, you could specify whether you have just a single predictor (bivariate or simple regression, which would be unusual for logistic regression of either kind) or more than one predictor (multiple regression, which is more common). So, things get wordy. Sorry for the confusion!
Bart