[SOLVED] Compute new variables
Posted: Wed Jul 26, 2017 1:27 pm
First of all, thanks for your work on this project!
One thing that I hope can be added to jamovi in the near future is the possibility to compute new variables and have those added to the dataset. In part, this should go under the "Data" menu. Also, an analysis should be able to add new variables to the dataset (e.g., residuals from a regression model).
In part, I bring this up in connection with the meta-analysis module that Kyle is working on (viewtopic.php?f=11&t=10#p59). A common workflow in meta-analysis is to compute some kind of outcome measure (and corresponding sampling variances) based on some reported descriptive statistics (e.g., means, SDs, and sample sizes -> raw or standardized mean differences; counts for 2x2 tables -> (log) risk/odds ratios and risk differences) and then, in a second step, analyze those values with standard meta-analytic models. Without some kind of 'compute new variables' capability, this type of workflow isn't currently possible inside of jamovi.
Best,
Wolfgang
One thing that I hope can be added to jamovi in the near future is the possibility to compute new variables and have those added to the dataset. In part, this should go under the "Data" menu. Also, an analysis should be able to add new variables to the dataset (e.g., residuals from a regression model).
In part, I bring this up in connection with the meta-analysis module that Kyle is working on (viewtopic.php?f=11&t=10#p59). A common workflow in meta-analysis is to compute some kind of outcome measure (and corresponding sampling variances) based on some reported descriptive statistics (e.g., means, SDs, and sample sizes -> raw or standardized mean differences; counts for 2x2 tables -> (log) risk/odds ratios and risk differences) and then, in a second step, analyze those values with standard meta-analytic models. Without some kind of 'compute new variables' capability, this type of workflow isn't currently possible inside of jamovi.
Best,
Wolfgang