Sunday, February 28, 2010

Say it ain't Pittsburgh ever again

Another short note.   Gary R. brings up the old story of Pittsburgh's #1 ranking as the most livable place in the US in 1985??  I should have realized it was in fact the 25th anniversary of that story.  The theme of the article today is about how inconceivable Pittsburgh's ranking was at the time and the disbelief folks had in the publication that produced it.

How big a deal was it here? A video highlighting the whole thing: Pittsburgh's No. 1

I have posted this before, but the disbelief in that result goes a lot deeper than even what Gary gets into. It really is a case study in what we call aggregation theory.  The cognitive dissonance of Pittsburgh's top ranking prompted Professor of Psychology Geoff Loftus of Washington University to write an article in Psychology Today in 1985 about how screwed up the ranking system must have been. You can read it yourself: Say it ain't Pittsburgh.  

Basically he explained how survey based ordinal rankings of preferences really added diminishing information beyond the top picks. Pittsburgh didn't really come out on top in 1985 because it excelled in any one category, it was sort of moderately ok across the board. Thus, according to Professor Loftus, the Pittsburgh #1 ranking was really an artifact of over-interpretation of the data. Just one contrarian voice? Maybe, but it turns out that the Places Rated publisher David Savageau was so taken by Professor Loftus' critique of the system that he brought him onboard and became co-author of the almanac in 1996.


Blogger Jim Russell said...

The Loftus piece is a hoot. Mountain-fresh Denver? I love this stuff.

Sunday, February 28, 2010 9:37:00 PM  
Blogger Unknown said...

Worth noting that even in Loftus's revised methodology Pittsburgh would have finished #11 in 1985.

Of course, 25 years later, the multifarious Forbes rankings still just add all their numbers together.

Monday, March 01, 2010 1:51:00 AM  
Blogger Mark Arsenal said...

I've been going out on a limb and declaring that in fact PGH does have a higher quality of life.

Ever since I started visiting PGH I've decided it has the nicest people this side of the 49th parallel.

I've deduced that when people are living so cheaply and upwardly-mobile corporate jobs are so scarce, the competitive human instinct present in expensive cities diminishes.

Businesses can provide better service because they know that "wasting" time talking to you won't hurt profits; there's no profit anyway. Waitresses make a career out of their occupation - it's not reserved strictly for high school students who don't care whether you live or die. Entrepreneurs can experiment with a business idea in a $900/mo storefront without having to sign a personal guarantee in the blood of their first-born.

Basically, I would argue that the cheap housing really does make life qualitatively better, because it makes you less concerned with social mobility, money, or growth. Score one for mediocrity!

Thursday, March 04, 2010 5:05:00 PM  
Blogger Dennis Galletta said...

In 1985 I accepted a faculty position in Pittsburgh and saw the Loftus article while preparing to move there. I don't like multiplying ranks when you have raw scores available. So I normalized the raw scores by calculating a z-score for each rating (raw score minus mean score for that criterion divided by standard deviation).

The z-scores have a mean of zero and standard deviation of 1. They are negative when below average and positive when above average. For instance, NY had a whopping 11.5 for culture, and -4.3 for crime. Add all together and NY acculumated a total z-score of 18.2.

Then I added the z-scores together. Pittsburgh's ranking? #12 just like Loftus! However, NY was #1 and Chicago was #2 in my calculations, whereas Loftus had San Francisco as #1 and NY as #2.

Tuesday, February 03, 2015 3:29:00 PM  

Post a Comment

<< Home