Re(1): Dania Median Tri Prices:This Year Compared To Last Year Posted on January 27, 2024 at 02:39:33 PM by Craig G
Or the average of the middle 2 values if the count is even.
Also, you need to decide if you care about differentiating tri's from Spec 7 and Spec 9.
I would like to see your thoughts on the following. Let's say we had a huge number of prices, eg 50,000, and we selected a sample of 500 values randomly, and then calculated the median. Repeatedly.
Do you (or Tiger, or anyone) have a formal statistical way of predicting the amount of variance from one sample to the next? Maybe I should say average difference.
I already wrote some code to test this empirically, which is quite revealing, especially as you vary sample size, but would like to know how it matches up with theory.