I have often wondered if it would be worthwhile for ballots to verify that voters understand the issues they are voting on, and whether this could be achieved objectively.
Case in point: I had a conversation a few years ago in which a person told me "I am voting 'Yes' on ballot measure X because I want Y to be legal". I told them that ballot measure X makes Y illegal, so they should vote No, and asked if they had read their voter guide. They seemed not to comprehend the issue after my attempts to explain it several different ways, and at the end I told them they should probably read their voter guide or refrain from voting.
I have read a couple similar questions on this site asking about voters' general "merits" and voters' general intelligence, and obviously those suffer from the need for everyone to agree upon strict definitions for overly broad and dubious concepts. They are also based on the voter as a person, not the voter's knowledge of a specific issue.
For the purposes of this question, assume that there is one true/false or a/b/c question per ballot measure, verifying a basic understanding of the measure's primary purpose. Assume that the questions are composed by the same parties we already trust to author the official voter guides which are distributed with our sample ballots. Also assume that this verification question is presented through all of the same assistive technologies (audio, braille, etc.) that the rest of the ballot is presented through, so that no group is excluded based on literacy alone, or some other handicap. Could such a step make our voting process better, without devolving into some new method of discrimination?