Non Overlap Drivetime tool: Understanding "grid" and need of use

[RickPack]

It appeared that I needed to use the Non Overlap Drivetime tool to avoid a sum of counts from the Allocate Append tool (like number of female residents) yielding the excessive counts seen when I used just the Trade Area tool with a Drivetime option. Is this because where trade areas overlap between my store locations, the Trade Area tool creates a unique polygon per store and then the Allocate Append tool calculates per polygon the count by its "block centroid retrieval technique"?

Then with the Non Overlap Drivetime tool, how do I quantify to business users the error introduced by the Grid Size value used? For example, if the Grid Size is 0.25 miles, is it correct to say, "Grid blocks of 0.25 miles were used to identify the block centroids in which residents lived, which meant a maximum buffer of 0.25 miles was allowed between stores' drive time ranges within which patients could be counted multiple times if sufficiently near more than one store."? 

Might someone offer better language?

[CharlieS · Accepted solution]

Hello again. I just saw this thread. 

---

@RickPack wrote:

It appeared that I needed to use the Non Overlap Drivetime tool to avoid a sum of counts from the Allocate Append tool (like number of female residents) yielding the excessive counts seen when I used just the Trade Area tool with a Drivetime option. Is this because where trade areas overlap between my store locations, the Trade Area tool creates a unique polygon per store and then the Allocate Append tool calculates per polygon the count by its "block centroid retrieval technique"?

---

You nailed it when you said "per polygon". It's important to remember that when you send a field of trade areas to the Allocate Append tool, it will append (just like the Append tool) on a per record basis. The block centroid methodology speaks only to the geographic accuracy of the values appended to each polygon. Also, I might suggest this documentation link.(for others reading this):

---

@RickPack wrote:

Then with the Non Overlap Drivetime tool, how do I quantify to business users the error introduced by the Grid Size value used? For example, if the Grid Size is 0.25 miles, is it correct to say, "Grid blocks of 0.25 miles were used to identify the block centroids in which residents lived, which meant a maximum buffer of 0.25 miles was allowed between stores' drive time ranges within which patients could be counted multiple times if sufficiently near more than one store."? 

---

I would start out by separating these two processes: 1) the spatial analysis and then 2) the demographic analysis. This focuses the effort to the spatial development before introducing any Allocate measures. So hopefully with that separation, and our discussion on our other post about grids, this all makes more sense.