There’s a lot of debate around the Alternative Vote (AV) voting system ahead of the 5th of May referendum in the UK. AV may or may not become the way we elect members of the UK parliament, but I think it could definitely be put to good use foreliciting user feedback.
We’ve just completed a usability testing project where (among other methods) we asked participants to rank 6 potential designs in order of preference. As expected, not one of the designs had more than 50% of the votes as a 1st choice.
We could at this stage conclude that the design that got the highest percentage of 1st choice votes is the best, but what if this design was also the worst design for the rest of the participants? We’d end up with a design that say 32% of people love and the rest of the sample hate, which is hardly a satisfactory outcome. What we’re looking for is a design that can satisfy most participants.
That’s where using the AV system comes in. By eliminating the least popular choices and counting 2nd and 3rd preference votes, we can make sure that the “winning” designs are the ones that have broad support across most of our sample. For this reason, AV is used in the Oscars to decide who gets the Best Picture award so that ‘a film hated by a clear majority of people [won't] go on to win with a minority vote’.
What methods do you prefer for analysing user preferences towards a set of designs? Let us know in the comments below!
And just to be clear, we never advocate using just the results of a vote to decide which is the best design. Always ask people why they’ve ranked a design as their top or bottom choice, and make sure you checkhow people behave in addition to what they say.
Photo: Liz West via Flickr/Creative Commons