RE: Does Prayer Really Work? Does God Even Care?
August 17, 2014 at 8:53 am
(This post was last modified: August 17, 2014 at 9:42 am by Michael.)
Thump. I think if you look at what the experiment was testing then the design was good. It asked and answered one small specific question and was appropriately controlled for that question (I very much doubt it would have passed peer-review otherwise). Anyways, I wasn't going to add any more, but I just wanted to address that specific point on controls.
Just as a complete aside, you may be a little surprised to know that some more modern scientific methods have no specific control group. The statistics behind these experiments are a little more complicated, but these experiments often have the most powerful designs, and avoid a statistical 'flaw' of more traditional designs where one control group can have too much weight in an experiment. Though people are often introduced to scientific experiments with the idea that you change one thing at a time, and have a control group, in practice the more sophisticated designs actually break both those rules. One relatively straight forward example is 'response surface design' ...
http://en.m.wikipedia.org/wiki/Response_...ethodology
Just as a complete aside, you may be a little surprised to know that some more modern scientific methods have no specific control group. The statistics behind these experiments are a little more complicated, but these experiments often have the most powerful designs, and avoid a statistical 'flaw' of more traditional designs where one control group can have too much weight in an experiment. Though people are often introduced to scientific experiments with the idea that you change one thing at a time, and have a control group, in practice the more sophisticated designs actually break both those rules. One relatively straight forward example is 'response surface design' ...
http://en.m.wikipedia.org/wiki/Response_...ethodology
