$$Best$$ is not well$-$defined in this experiment; we would ideally like to pick the subject line that earns us the most votes, but with the election one month away, we don$$t yet have that data. All we have is the data described above. Which one$($s$)$ of available metric$($s$)$ is$($are$)$ appropriate to be used to define the $$best$$ variant? Select all that apply.

Question $3$ options:

For each variant, percentage of people who donated $.$ Because anyone who donates is likely to vote, so donation percentage is a good proxy for number of votes.

Average donation amount of people who received the same variant. Because in politics, sadly, money buys votes through advertising and marketing.

For all who see the same variant, percentage of people that shared on social media. Because people are very cognizant of how their peers perceive them.

For each variant, average donation amount of those who donated. Because in politics, sadly, money buys votes and has the influential power.