Do you ever read something and cringe knowing that you were pegged. Peter Coy at Business Weeks Hot Property has a great post about statistical misuse by those of us who write about real estate. And you know what? He is right and
I am guilty. Very guilty.
I am a stats whore. I love market trends reports. I love Top 10 lists. Got a prognostication on the market or a report that cites a number, I am all over it. I can not help myself. Reel me in cause I am already on the hook.
As a child I would spend hours pouring over the statistics in the sports section of Newsday and the New York Times to see if Micky Rivers was still batting over .300. Today I check the NHL stats on a near daily basis to see how my Thrashers players are stacking up. And then do not get me started as I obsess over my stats package for this site.
So he has a reader Dave that had the audacity to point out that Peter, and of course myself since I am projecting, never talk about the margin of error in the statistics I use. And of course, once those variances are added to the numbers, the wonderful story they allude to tends to fall apart. Of course I am speaking tongue and cheek about Dave’s comments because he is exactly right.
Housing Starts Rise in December (We think…)
Resales Bottom Out in 2006 (Maybe…)
The dirty truth is that almost all the reports we comment upon are estimates. They are taken from a sample of the marketplace and then have a margin of error. Happens everyday on the front of the newspaper when a poll is announced saving politician A has a 5 point lead over politician B. The margin of error may be 4 percent either way and only 1,000 people are surveyed.
But it makes a good headline.
So, am I going to change? Most likely not. But will there be a nagging thought in the back of my mind that this wonderful story may be build on a less than sturdy foundation? Most likely. But I will do my best to see if there is any thing underneath the screaming headline that gets me going enough to write about it.
I wonder if there is a 12 Step program for a stats junkie? If so, sign me up and send over the statistical analysis of success cures.
No related posts.
{ 2 comments… read them below or add one }
I feel your pain.
Even accurate statistics can be misleading. A wise person once told me, "Statistics are like a bikini. What they reveal is suggestive, what they conceal is vital."
The trouble is that everyone wants an answer as to what the market is going to do, and it's easy to be sucked into trying to give them one.
What I can predict is this. The market will go both up and down, and over the long term go up. Everything else is just educated guessing.