More often, analytical reports rely on Mean values (aka average) in the analysis and to reach conclusions. But is it a good idea?
To answer this question, I compared the Mean and Median house prices per UK region for residential properties sold since October 2021. The data are shown in the plot [raw data obtained from Land Registry].
If we take London as example, the Mean price was £628K while the Median was £500K only. That’s a substantial 25% difference! Other regions show similar “skewness”, albeit to a lesser degree. This indicates a consistent regional trend that fewer (and pricier) properties make a rather large impact on house price data.
So which measure should you consider in your analysis?
I would argue that both should be used. Bigger difference is a good indicator that care must be taken in the analysis.
It’s also more helpful to use more localised statistics. For example, per district or even per small area.
If you use AccuVal to value a property, pay close attention to the Mean and Median prices of similar properties. A large gap indicates “troublesome” areas.