We felt safe virtually the entire time we were there. We tried to avoid driving at night, but did do so on a few occasions. There were some crazy moments on the road: passing slow trucks, fast drivers, winding curves, etc, but nothing that would shock you if you’ve been road-tripping Mexico, Madagascar, and…
I hope people do read this responses, because they clearly show there’s no problem in the data.
As I showed, interpolation method makes no difference.
And as far as the data for the latest period… why is it invalid? It’s not at all unreasonable that Amish population would be about the amount described from surveys of Amish groups.
I feel like you’re being willfully thick.
There are *direct* counts of Amish population in, e.g. 2017 and 2012. As in, we actually know how many there were. We can count them.
The question is, how do we fill in the missing space? Let’s take 2001, 2012, and 2017 as an example. In 2012, there were…
There’s considerable debate about appropriate interpolation of population data. But in the end, it makes no difference, because there is agreement about the start/end points between which interpolations are made. I prefer linear interpolation as that yields population trendlines that more closely-match real-world…
Take the series of coordinates, 0,0;NA,1;NA,2;NA,3;NA,4;5,5
To linearly interpolate between 0,0 and 5,5, we would make the coordinates between them 2,2; 3,3; and 4,4.
Now, that means an increment of 1 in each period. So our values are 0,1,2,3,4,5. Following me?