Thread:User talk:Yair rand/Definition editing gadget/reply (2)

I actually think that the key part of the test is whether bad contributions are encouraged. More total edits and a "small" increase in vandalism/bad edits would show success without requiring careful measurement.