You’re welcome. Yes, I know how to weight. You will need to know the number, or have the means to calculate the number, in each of the groups you want to average together.
Let’s say a poll of 500 LV has 20% in the 19-44 age group:
number = .20 * 500 = 100 LV
And has 40% in the 45-60 age group:
number = .40 * 500 = 200 LV
Let’s say that Charlie Brown is the first choice of 70% of 19-44 and 50% of 45-60, and we want to know what the weighted average of the combined groups is:
((.70 * 100) + (.50 * 200)) / (100 + 200) = .5667
= 57% rounded
I hope that this makes sense.
BTW, I enjoy reading your analyses. :)
Thanks. It’s funny that the stat hounds on Twitter grew eerily silent on such things a Millennials and educated voters in polls. Earlier in the campaign they were constantly harping on these demos, which are overwhelmingly moved in Trump’s direction. But the categories would be subdivided and I would cite an average and they’d just say “well, you didn’t weight them”-— but wouldn’t produce the real number because it would undermine their side.