ORIGINAL: bluebird2
For those that still believe that the shift in sample size is responsible for the 5% statewide decrease in breeding rates ,here is an example that shows that is simply impossible. If you take 3 WMUs with a breeding rate of 96% and 200 doe sampled and 3 WMUs with an 86% breeding rate and 100 does sampled the average breeding rate for all 6 WMUs is 92%.
Now if you reduce the sample size of the first 3 WMUs to 100 does sampled and keep the same breeding rate,while keeping the sample size in the other 3 WMUs constant, but increasing the breeding rate by just 4%, the average breeding rate for all six WMUs increases to 93.9%.
Therefore, despite the shift in sample sizes it is impossible to get a 5% decrease in breeding rates unless the statewide breeding rates decreased by at least 5% in most WMUs.
Blueboy smoke and mirrors at it's finest!
Lets look at his example one more time without inserting any assumptions...
200 samples @ 96%
100 samples @ 86%
200x96=19200 100x86=8600 8600+19200=27600 27600/300=
93%
nowletsreducethe more productive samples by 100...
100 @ 96%
100 @ 86%
100x96=9600100x86=8600 8600+9600=18200 18200/200=
91%
But if the weight is shifted to favor the less productive area (what RSB tells us is what really happened)...
100 @ 96%
200 @ 86%
100x96= 9600 200x86=17200 9600+17200=26800 26800/300=
89%
so a shift in sampling sizes can have a significant effect. Add in seasonal differences like weather, localized disease, mortality etc etc and all of a sudden RSB's explanations make a lot more sense.
Given enough time with a calculator and a lack of scruples, anyone can put their own twist and spin on things.