Skip Navigation

MEDIAN vs AVERAGE PRICE TREND

Started by joedavis
about 18 years ago
Posts: 703
Member since: Aug 2007
Discussion about
MEDIAN IS NOT AFFECTED BY A FEW PROPERTIES -- SORRY TO REITERATE THIS point but I keep seeing posts here that say that when xxx super high priced properties are not on the market the median price will come down. The AVERAGE will certainly change because of these few outliers. The MEDIAN is pretty much unaffected by these properties here is an example 11 properties whose prices are listed as below... [more]
Response by nyc10022
about 18 years ago
Posts: 9868
Member since: Aug 2008
Ignored comment. Unhide
Response by stevejhx
about 18 years ago
Posts: 12656
Member since: Feb 2008

That is simply not true in the case of a small sample, where many apartments might close in a single building in a short time frame. That is, while it is true that the actual median does not change in the sequence 1, 2, 5 and 5000, and 1, 2, 5 and 10, it does change if the sequence is 1, 2, 5, 500, 5000 and 1, 2, 500, 500, 5000. In the first case the median is 5, in the second it is 500. All it took was one price change.

Dream on.

Ignored comment. Unhide
Response by joedavis
about 18 years ago
Posts: 703
Member since: Aug 2007

sorry for reposting -- i see this often enough that I decided it may deserve its own thread

Ignored comment. Unhide
Response by joedavis
about 18 years ago
Posts: 703
Member since: Aug 2007

steve
I usually just ignore you. However, your blissful ignorance does need to be rocked
As the sample size increases the median is much more robust -- what you did was to change 50% of the numbers and then you saw a change -- please get yourself a calculator or excel or go take STAT 101 and try it out

Ignored comment. Unhide
Response by nyc10022
about 18 years ago
Posts: 9868
Member since: Aug 2008
Ignored comment. Unhide
Response by kylewest
about 18 years ago
Posts: 4455
Member since: Aug 2007

I said this on the other thread and it's worth repeating I think since the heading is misstated.

ust for clarification: it isn't really correct to call something and "average" when distinguishing it from the "median". A median IS an average. Generally there are two types of averages people refer to: the more typical MEAN average and then the MEDIAN average referred to above. While different, both mean and median are types of averages. Together they provide a rounder picture of a situation.

For example, a mean average of the life expectancy in a country is meaningful in comparing to year over year figures to look at the health of the population overall. If the number goes up, all it means is that the population as a whole is living longer. What is not revealed by the mean average, however, is whether the higher number means every single person on average is living longer, whether just the old people are living longer or if some vulnerable population is living longer. If there is a high infant mortality rate in this country, because of some disease for which a vaccination is suddenly developed, and the 30% infant mortality rate is dropped to 1%, the mean average will shoot up making it appear that people who don't understand averages that adults are living further into old age. Yet here, not a thing has changed for older people. The median average, however, would reveal this because year over year the mean average in such a country would likely hardly shift. To look at just the mean average would make you think nothing had changed. It is taken together that the averages reveal something going on.

Ignored comment. Unhide
Response by nyc10022
about 18 years ago
Posts: 9868
Member since: Aug 2008

no reason to repeat things over here. just let the thread die, and follow the conversation where it is actually happening...

http://www.streeteasy.com/nyc/talk/discussion/4857-raise-your-hand-if-you-are-a-foreigner-or-a-cash-rich-sideline-sitter-waiting-for-the-right-moment-to-buy

Ignored comment. Unhide
Response by stevejhx
about 18 years ago
Posts: 12656
Member since: Feb 2008

joey: "what you did was to change 50% of the numbers and then you saw a change"

No, actually, there were 5 numbers and I changed 1 of them. That is 20%, not 50%. So, to regurgitate: "please get yourself a calculator or excel or go take STAT 101 and try it out."

If all of 15CPW or the Plaza close in a particular quarter, that will affect the median as Jonathan Miller said it did. If a building full of studios all closes in a quarter, that will likewise skew the median.

Sorry.

Ignored comment. Unhide

Add Your Comment