The joint_tests() function can return very sparse -- or even empty -- results for rank-deficient models (e.g., empty cells, leading to estimability issues). This can happen for a subtle reason -- an error in the estimability package, version 1.4.
You can check what version of estimability you have installed in various ways, including
packageVersion("estimability")
If you have version 1.4, you should update your installation to a different version. You may install the latest version from GitHub via
remotes::install_github("rvlenth/estimability")
Due to a CRAN "vacation", there will not be a new update on CRAN until after August 5, 2022. Another possibility is to downgrade to version 1.3, which will also solve the issue:
User advisory
The
joint_tests()
function can return very sparse -- or even empty -- results for rank-deficient models (e.g., empty cells, leading to estimability issues). This can happen for a subtle reason -- an error in the estimability package, version 1.4.You can check what version of estimability you have installed in various ways, including
If you have version 1.4, you should update your installation to a different version. You may install the latest version from GitHub via
Due to a CRAN "vacation", there will not be a new update on CRAN until after August 5, 2022. Another possibility is to downgrade to version 1.3, which will also solve the issue: