Comments on: Blog Debating and Experimentation http://facibus.com/onblogging/2007/09/09/blog-debating-and-experimentation/ Andrew looks at Blogging and Blog Life Wed, 07 Jan 2009 12:13:32 +0000 http://wordpress.org/?v=2.3.2 By: Andrew Boyd http://facibus.com/onblogging/2007/09/09/blog-debating-and-experimentation/#comment-715 Andrew Boyd Mon, 10 Sep 2007 03:00:27 +0000 http://facibus.com/onblogging/2007/09/09/blog-debating-and-experimentation/#comment-715 John, just took a few minutes to flick through some of the articles - very good stuff indeed. I've known about this in the past and had reverted to 1950s engineering fix-all-bar-one-variable thinking - fascinating. Best regards, and thanks, Andrew John,

just took a few minutes to flick through some of the articles - very good stuff indeed. I’ve known about this in the past and had reverted to 1950s engineering fix-all-bar-one-variable thinking - fascinating.

Best regards, and thanks, Andrew

]]> By: AndrewBoyd http://facibus.com/onblogging/2007/09/09/blog-debating-and-experimentation/#comment-713 AndrewBoyd Sun, 09 Sep 2007 21:20:08 +0000 http://facibus.com/onblogging/2007/09/09/blog-debating-and-experimentation/#comment-713 Hi John, thank you for your comment and for the references. You are absolutely correct - we do this with UX testing all the time. I will follow up. Best regards, Andrew Hi John,

thank you for your comment and for the references.

You are absolutely correct - we do this with UX testing all the time. I will follow up.

Best regards, Andrew

]]> By: John Hunter http://facibus.com/onblogging/2007/09/09/blog-debating-and-experimentation/#comment-709 John Hunter Sun, 09 Sep 2007 15:52:12 +0000 http://facibus.com/onblogging/2007/09/09/blog-debating-and-experimentation/#comment-709 I just like to point out that fixing all variable but one is not the best way to experiment. In fact <a href="http://curiouscat.com/management/doe.cfm" rel="nofollow">multi-factorial designed experiments</a> are much more effective (you learn much more much more quickly). In addition to many other benefits multi-factorial designed experiments allow you to capture interaction effects. Here is a nice simple intro: <a href="www.symphonytech.com/articles/romeos.htm" rel="nofollow">Three Romeos and a Juliet</a> - Our early brush with Design of Experiments. <a href="http://curiouscat.net/library/designofexperiments.cfm" rel="nofollow">More articles on designed experiments</a> I just like to point out that fixing all variable but one is not the best way to experiment. In fact multi-factorial designed experiments are much more effective (you learn much more much more quickly). In addition to many other benefits multi-factorial designed experiments allow you to capture interaction effects. Here is a nice simple intro: Three Romeos and a Juliet - Our early brush with Design of Experiments. More articles on designed experiments

]]>