![]() |
z test, t test, or chi squared test If I run an AB split test for a marketing campaign and get these results: A sent: 100000 A success: 50000 A conversion rate: 50.0% B sent: 101000 B Success: 50400 Conversion rate: 49.9% Which test would be better to run in order to find out if the results are statistically significant? Also, what results did you get if you ran both or either? |
Re: z test, t test, or chi squared test Additional info: I've used both of the following: Excel spreadsheet from The Analytical Group (http://www.analyticalgroup.com/download/sigtest.xls) using two independent samples AB Test Calculator: http://www.splittestcalculator.com/ Not sure if either is right/none are right/both are right (it appears that the analytical group uses z test while the calculator uses Yates' chi squared test) |
All times are GMT -8. The time now is 12:10 AM. |
Copyright © 2019 My Math Forum. All rights reserved.