Timeline for How to determine if there is a drop in revenue after a change in the system?
Current License: CC BY-SA 3.0
4 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 30, 2012 at 12:42 | vote | accept | jermeyz | ||
| Apr 28, 2012 at 0:06 | comment | added | pgericson | You are right about the autocorrelation, there is probably some kind of random walk with drift. This is a quick and practical test by looking at e.g. 1 month before the change of daily revenue data and 1 month after the change to see the levels of whether the drop was statistically significant. It was not meant as a perfect academic solution. | |
| Apr 27, 2012 at 22:45 | comment | added | Peter Ellis | Won't there be a problem because the t test assumes the observations are iid, whereas the data in the OP's post will probably be autocorrelated? | |
| Apr 27, 2012 at 18:53 | history | answered | pgericson | CC BY-SA 3.0 |