Realtors are not Statisticians

Realtors are not statisticians, and they should’t play them in the media. Given how much data realtors have at their disposal, you would think their licensing requirements included some basic statistics and data analysis. Unfortunately, some of them misrepresent the data on a regular basis.

For example:

• They call a trend with only 2 data points. You can’t look at data in a vacuum. If this month’s sales are up over last month’s it dosen’t necessarily mean buyers are getting back into the market. Seasonality and pricing can also be affecting the data.

• If they happen to look at year-over-year, to account for seasonality, they don’t offer context. Like last year was a really bad year, so the fact that we are up 2% over last year, is actually pretty bad, not something to celebrate.

• They don’t seem to grasp the concept/importance of sample size, and weighted averages. You can’t combine two averages to get the overall regional average. For example, the average of a zip with 30 homes for sale, at an average of $225 a square foot, and a zip with 10 homes for sale at $275 a square foot is not $250 (the average of $225 and $275), as much as the realtors would like it to be its. The properly weighted average is $237.5, which is a much less exciting number when compared to $250.

I guess I am just frustrated because this misrepresentation adds fuel to the fire.
I do want to caveat that there are some savvy realtors who understand the data and offer context, like the great work done over at sacrealstats by AgentBubble and Max.