I have finished analyzing the results from last week's poll. Less than two dozen people bothered to vote but those who did varied fairly wildly in how they ranked the five projects. To figure out how people rated the projects, I used a fairly simple technique of averaging the scores. The scores were simply the ranks assigned to the projects. This means that the project with the lowest average would be the most popular. When these numbers were rounded off, I ended up with a five way tie.
This is actually quite significant, in that it means that the difference between the most popular to least popular project is not very high. That being said, I wanted to have a rank of projects so I changed the rounding to 2 decimal places and ended up with the following ranking. 2.62 Ultimate Retro Project. 2.95 Coffee Quest and One of those Weeks. 3.05 Dozen Days Pentalogy 3.43 Books and Articles.
Due to the small sample size, I can't say with certainty how accurately the poll reflects the wishes of my visitors. Likewise, when looking at individual votes, I am fairly certain that a few of the people voting never read my instructions and ranked the projects in the wrong order. While this would not have a huge impact on the results if hundreds of people voted, with only 21 votes this probably has a big impact. Still, after looking over things with the knowledge that I will soon be between paid projects I have some interesting decisions to make.
Friday, August 24, 2007
Subscribe to: Post Comments (Atom)
Post a Comment