'Significance test for AB testing of email marketing campaign - data in MySQL

I have 50,000 email ids divided into Group A and Group B equally. An email is sent to both groups with different subject. The Open Rate is as follows:

Group | Emails_Opened | Total Emails | Open Rate
A     |  24332        |  34471       | 70.5869% 
B     |  24020        |  33761       | 71.1427%

Which significance test should I run to make sure the results are statistically significant and how? The data is in SQL database.

sqlexcelab-testing


Solution 1:[1]

So you fill in the column totals and row totals, then each expected value is

<corresponding column total>*<corresponding row total>/<grand total>

e.g.

=B$4*$D2/$D$4

enter image description here

Then use

=CHISQ.TEST(B2:C3,F2:G3)

Sources

This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.

Source: Stack Overflow

Solution Source
Solution 1

Related Questions