You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Many thanks for creating this app, the package is great to visualize the terms.
However, I would like to know which contrasts are possible with a given model.
I'm not sure how easy would it be to expand it but I think it would help to understand designs if one could say, something like: "Ok, with this design I cannot compare the two conditions I wanted.", or "Great, this design takes into account the covariates when comparing A vs B".
I understand that this could be quite a long list in complex designs but for two variables there wouldn't be too much combinations (Although, all depends on the variables and the number of factors in them). Perhaps in the shiny app a new menu with a dropdown option to select which fitted values could be used to build a contrast would help reduce the burden.
The text was updated successfully, but these errors were encountered:
Thanks for the suggestion. I have to think about what we can do here (if I remember correctly, the question has also come up once before) - it may not be so easy, for an arbitrary design, to extract "all" meaningful contrasts and describe them in an interpretable way. Our idea with the current implementation was rather that the user decides which groups to compare, and the visualizations/tables directly provide the contrast that should be used to achieve the comparison. I'll keep this issue open while we discuss it in the development team.
Many thanks for creating this app, the package is great to visualize the terms.
However, I would like to know which contrasts are possible with a given model.
I'm not sure how easy would it be to expand it but I think it would help to understand designs if one could say, something like: "Ok, with this design I cannot compare the two conditions I wanted.", or "Great, this design takes into account the covariates when comparing A vs B".
I understand that this could be quite a long list in complex designs but for two variables there wouldn't be too much combinations (Although, all depends on the variables and the number of factors in them). Perhaps in the shiny app a new menu with a dropdown option to select which fitted values could be used to build a contrast would help reduce the burden.
The text was updated successfully, but these errors were encountered: