The IBM developerWorks site has a new articleposted today from Paul Meagher - Part 2 of the Conduct Web experiments using PHP series.

In Part 1 of this two-part series, Paul Meagher showed you how to improve the quality of your Web offers by developing PHP-based tools to set up and run Web experiments. In Part 2, you focus on simulating and analyzing the contingency table data arising from these Web experiments.

My purpose in this article is to analyze the resulting data. (You will absorb the reasoning in this article more easily if you have read or are familiar with concepts discussed in the prerequisite articles suggested in Resources.)

Using the data from part 1's experimental designs to fuel the formulas, this second part can help you unserstand which layouts and items were the most effective with your users. There's plenty of heavy math here, so it's not for the faint of heart. He does, however, explain it all in very good detail, so following it isn't too much of a problem...