AVSIM Vendor Survey

[mgh 89]

I added a post to the following thread pointing out a flaw in the calculation of percentages in this survey that doesn't seem to heve been picked-up:http://forums.avsim.net/dcboard.php?az=sho...ing_type=searchTo reiterate, the flaw is that the percentages are calculated by dividing the the number of replies to a particular aspect of a vendor by the total number of replies, which includes replies from those who haven't bought anything from that vendor.Taking Abacus as an example (simply because it's the first alphabetically) there were 12 replies to the statement "Product quality is typically excellent". The corresponding percentage is calculated by dividing this figure by 1343 (the total number of replies) to give 0.9% However, I suggest this should be based on the number of replies from those who have bought a product from that vendor, ie the total of 1343 less the 805 who replied "I have never bought one of their products so cannot rate them". Using the figure of 538 (1343-805) purchasers would give a percentage of 2.2% - which differs by a factor of 2.It also introduces a bias against against vendors with lower sales and hence lower response rates. If all 200 replies gave a particular vendor the highest rating, that's a 100% rating. However basing the calculation on the current total number of replies would reduce that figure to 14.9%.Can this be changed?