Your guide to NYC's public proceedings.
Q&A
Detailed explanation of research methodology and replication challenges
2:32:59
ยท
150 sec
The IBO analyst provides a comprehensive explanation of how the research was conducted, including challenges faced and potential improvements for future studies.
- Matching CDF grantee location data with public school addresses
- Challenges included address validation, co-located schools, and District 75 schools
- Suggested improvements: faster processing methods and inclusion of school IDs (DBNs) in CDF reporting data
Carlina Rivera
2:32:59
And if you could just tell us very briefly, you talked a little bit about it, but how the research was conducted and how easy would it be to replicate it for other years.
Arden Armbruster
2:33:10
Sure, so I think on its face it's pretty straightforward.
2:33:12
You schools had a CDF program.
2:33:15
We're matching the location address information for the CDF grantees that is produced as part of the end of year reporting to addresses of traditional public schools in this case.
2:33:27
So we used District 1 Through 32 and District 75.
2:33:31
But we did run into a few challenges that we were able to, I think, satisfactorily address.
2:33:38
One of them is the DCLA addresses aren't sort of validated, so you'll have spelling errors and things like that that we need to clean up to be able to match to the DOE addresses, which are maintained by our education team and quality checked.
2:33:53
And then you have the co located schools.
2:33:55
Schools are on a campus where you might have five schools at one address and then if you're matching to the CDF data then you might have you know sort of five programs at five different schools.
2:34:06
So we had to go through manually and sort of assign programs based on a description and the CDF data.
2:34:13
And then the last thing I'll mention on the sort of challenges front is that the district 75 schools, where you do have schools that have multiple sites, We did see a couple of instances where there was a D75 school, or it appeared to be a D75 school, and we had to sort of assign it back to the main location to make sure that we were sort of capturing that school.
2:34:36
But those, both the campus issue and the D75 issue, a really small number of programs that we weren't ultimately able to sort of assign to a school and include in the data.
2:34:46
Sort of saying all of that, having undertaken this once, I think that I have some ideas for how we might do it faster in the future.
2:34:54
And then one thing I wanted to mention, sort of from the perspective of an analyst that would reduce the amount of time needed to conduct an analysis like this and also improve the quality of sort of the matches that we're talking about, is to include a school ID in the CDF reporting data.
2:35:10
I know this is something the council does for some of its reporting.
2:35:13
These are called DBNs, district borough numbers, and it would allow us to match directly on that DBN instead of on an address.
2:35:20
You might have just an accidental flipping of numbers in address or avenue with two V's that can complicate the research process.
2:35:27
So I think that would improve the quality of
Carlina Rivera
2:35:29
the match.