Statistical Weighting - HELLLLPPP!
I've currently been doing some work which is based around all house fires that have occurred between April 2009 - 31st July 2012. Each individual incident has been allocated a risk score based on a range of factors (casualties, damage to building e.t.c.). However, these scores now have to be historically weighted and I'm stuck!!!
Those incidents that occurred in April 2009 (the least recent) are considered to have less of a weighting and all scores for incidents for this month have been split by two (or 0.5). All incidents that occurred during July 2012 are considered to have the greatest weighting and so remain at 1 with no change made to their score. I now have all incidents for the 38 months falling between 0.5 and 1 that need to be proportionally weighted!!! How on earth do I do this? I tried 0.5/38 and then cumulatively subtracting the answer from each month but I got some very strange figures. I am quite frankly flumoxed and need to sort it asap. Any assistance would be greatly appreciated.
Re: Statistical Weighting - HELLLLPPP!
post stat questions in the appropriate statistics forum