The second level covariates are used to construct the path where CONN stores the data for each experiment. The problem is that paths are limited to a certain length, and so if you have too many covariates (with reasonably descriptive names), you quickly reach this limit and everytime you try to show the 2nd level result of an analysis, CONN will ask you where to store the data.
The problem is twofolds: 1. CONN will ask everytime where to store the data if you have too many covariates with long names, 2. CONN cannot detect nor reload results that are stored in another path than the default one.
To workaround this issue, I propose to add a description field for 2nd-level covariates: the description field won't be useful for the analyses, it will only serve as a note taking space so that we can use shorter names for the covariates (like "ads"), and still have a description field in case we forget what this covariate was about ("accumulated_days_in_space_standardized").
Also the same problem and solution might apply for sessions.
The problem is twofolds: 1. CONN will ask everytime where to store the data if you have too many covariates with long names, 2. CONN cannot detect nor reload results that are stored in another path than the default one.
To workaround this issue, I propose to add a description field for 2nd-level covariates: the description field won't be useful for the analyses, it will only serve as a note taking space so that we can use shorter names for the covariates (like "ads"), and still have a description field in case we forget what this covariate was about ("accumulated_days_in_space_standardized").
Also the same problem and solution might apply for sessions.