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
Demographic stats were introduced in commit 132397b for budgets and in commit be68c8c for polls; this bug has always been present in these statistics
Current Behavior
A few years after a poll/budget has finished, the displayed stats might change. It happens because the stats reflect people's current age, gender and geozone, instead of their data at the time they voted.
Expected Behavior
Demographic stats should refer to the time when the process took place.
Notes
For poll voters, we actually store their demographic information at the time they voted in the poll_voters table, so maybe tweaking the participants method could be enough 🤔.
For budget investments, however, we don't currently store this information. We could calculate the age based on when the process took place, but there's no way to track whether someone has changed their gender or their geozone since then.
Pull request #5533 improves the calculation of age stats (not as precise as using the information in poll_voters would be, in the case of polls, though), which solves the most important problem in this issue; after all, not everybody changes their gender or their geozone, but most people certainly change their age as time goes by.
The text was updated successfully, but these errors were encountered:
References
Current Behavior
A few years after a poll/budget has finished, the displayed stats might change. It happens because the stats reflect people's current age, gender and geozone, instead of their data at the time they voted.
Expected Behavior
Demographic stats should refer to the time when the process took place.
Notes
For poll voters, we actually store their demographic information at the time they voted in the
poll_voters
table, so maybe tweaking theparticipants
method could be enough 🤔.For budget investments, however, we don't currently store this information. We could calculate the age based on when the process took place, but there's no way to track whether someone has changed their gender or their geozone since then.
Pull request #5533 improves the calculation of age stats (not as precise as using the information in
poll_voters
would be, in the case of polls, though), which solves the most important problem in this issue; after all, not everybody changes their gender or their geozone, but most people certainly change their age as time goes by.The text was updated successfully, but these errors were encountered: