Hacker Read top | best | new | newcomments | leaders | about | bookmarklet login

I was reacting to a CNN article earlier which did not make it clear if they were using the metropolitan or city statistics. Apologies.


sort by: page size:

Title is misleading--the article only discusses metro areas. I'm interested in seeing the stats within city limits.

I'm assuming they're asking for city-wide averages (they say "statistics" in the article), not individual results.

Did you read the article? It gave the sources for the data for both cities.

It wouldn't explain the wide statistical variation between cities.

It's unfortunate that these are (or at least appear to me to be) raw counts, rather than percentages, normalized by each metro area's population or job totals.

They use country-level data for many factors, including for "Inclusivity & Tolerance". Having also lived in both cities, it's clear why this list is flawed.

The methodology is highly suspect. They chose 50 of the most populated cities that had data available. Not the 50 most populated cities that had data available.

Where's Indianapolis? Cleveland? Bakersfield?


You need to compare on a per-capita basis, not per square mile.

Sorry, what I meant was I was looking at the data for my specific city. For some indicies statistics Canada has a breakdown to specific cities.

I believe that they're basing this data on stats collected by the census, so you have take the city as defined by the census bounds, which aren't necessarily similar to common conceptions of what area is 'in' a city.

I'm pretty sure its a city and suburbs issue (just from news articles), but I am at a loss as to how to get the statistics to prove either case.

This assumes the data is granular to the city level. ...it's probably just state level data - and they are likely just using the state center point.

They're comparing zip-code level Case-Shiller index based on distance from city center. Metro level (which is what is widely quoted as far as I know) does not get at what they're trying to characterize, which is relative demand for housing in more urban environments.

Seemed like he stopped at a visualization of the data, without any real interpretation.

I'd like to see some information on the density of the areas. For example, Chicago is listed as the metro region (8 million people) when the city itself only has about 3 million, and public transport hardly exists outside the city.

If the total rides had been divided by the city population, it would equal much closer to 200 rides per capita.


That is interesting. I haven't looked at that data set. At the moment, this only looks at cities with a population > 1000 obtained from GeoNames.

Also, the cities on the lists are the only cities they even have data for, so it's possible for other cities to have higher counts.

So do you believe this article should be disregarded, because it uses data that differs from California's and Chicago's?

Or would you instead prefer to only use the parts of this article that support your position, and replace those that don't with data from California and Chicago that does?


Your commment implies they were cherry-picking the data, which isn't true. They only included data that could definitively be marked as within Seattle vs outside Seattle.

Not helping: I think they've done an exceptionally poor job of clarifying how they defined the metropolitan areas. For example, the label for the data in the Boston area (lifted off the map, which was the only place I could find the slightest mention) is "Boston-Cambridge-Quincy, MA-NH". Grouping MA and NH?! Conversely, Riverside and Los Angeles are separated despite covering a similar geographic range. I'd love to see the geographic regions in this study shown as polygons on a map, along with some sense of the total population per region.
next

Legal | privacy