Press "Enter" to skip to content

Reviewing Experimental Results in the Process

John Cook talks philosophy of statistics:

Suppose you’re running an A/B test to determine whether a web page produces more sales with one graphic versus another. You plan to randomly assign image A or B to 1,000 visitors to the page, but after only randomizing 500 visitors you want to look at the data. Is this OK or not?

John also has a follow-up article:

Suppose you design an experiment, an A/B test of two page designs, randomizing visitors to Design A or Design B. You planned to run the test for 800 visitors and you calculated some confidence level α for your experiment.

You decide to take a peek at the data after only 300 randomizations, even though your statistician warned you in no uncertain terms not to do that. Something about alpha spending.

You can’t unsee what you’ve seen. Now what?

Read on for a very interesting discussion of the topic. I’m definitely in the Bayesian camp: learn quickly update frequently, particularly early on when you have little information on the topic and the marginal value of learning one additional piece of information is so high.