Timeline for Is a T-Test a suitable model for this data?
Current License: CC BY-SA 3.0
13 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Mar 14, 2018 at 10:49 | history | tweeted | twitter.com/StackStats/status/973873715955097600 | ||
| Mar 14, 2018 at 10:18 | answer | added | Sal Mangiafico | timeline score: 0 | |
| Mar 13, 2018 at 23:23 | comment | added | badatstats | @SalMangiafico Thanks heaps for the help, I'll implement one of the methods you mentioned. Can you copy/paste your comment into a reply, I'll accept that as the answer :) | |
| Mar 13, 2018 at 22:05 | comment | added | Sal Mangiafico | There's also a concept of inventory as a percent of sales. With your data, I'm not sure how would get "sales" or if this makes any sense , but it might make sense to categorize the day by the sum of the amounts of change between time stamps. | |
| Mar 13, 2018 at 21:52 | comment | added | Sal Mangiafico | I think that would also make sense. Or you may want to use some other statistic to summarize the day's numbers. Like the median, or a trimmed mean. These statistics are less sensitive to outliers, like a single really high value during the day. | |
| Mar 13, 2018 at 21:01 | answer | added | rolando2 | timeline score: 0 | |
| Mar 13, 2018 at 20:52 | comment | added | badatstats | @SalMangiafico I was wondering, regarding the proportion method you suggested would it be better if I did: (Final Inventory)/(Max Day Inventory), or would it better if instead I take the absolute value of the Inventory, and then sum the values together, and then put that in the denominator such that I now write: (Final Inventory)/(sum(abs(Inventory))) ? Sorry if this is a dumb/trivial question, I'm mostly just not sure how best to approach this problem | |
| Mar 13, 2018 at 20:31 | comment | added | badatstats | @AdamO Not really a trend in inventory, but it's more that I want to see whether the Inventory for a particular ID reverts toward zero on a particular day. In my dataset I have a 4 month period and I have about 20 different IDs but I just put a short example above as to what I was after | |
| Mar 13, 2018 at 20:29 | comment | added | badatstats | @SalMangiafico Yes perhaps that might be a bit better since I'm checking up further methods now and can't find much :( what sort of method do you believe is appropriate? | |
| Mar 13, 2018 at 20:26 | comment | added | Sal Mangiafico | t-test is definitely not the test you want. It sounds like maybe you want to assess the final number relative to the maximum for the day, maybe as a proportion? | |
| Mar 13, 2018 at 19:50 | comment | added | AdamO | Are you saying that you are trying to detect a trend in Inventory as the time stamp increases? Do you have data on more than 1 store? | |
| Mar 13, 2018 at 19:36 | review | First posts | |||
| Mar 13, 2018 at 19:58 | |||||
| Mar 13, 2018 at 19:31 | history | asked | badatstats | CC BY-SA 3.0 |