[glynjackson]

There are many, many ways to test! A control group, quasi-experiment evaluation etc. "Correlation does not imply causation". All she has is a hypothesis, a theory based on observational research, while this is ok, the article tries to pass this off as 'fact'.

I'm not staying I disagree with her 'hypothesis' but, people should require evidence before accepting it as fact. Maybe if she have quoted studies, instead she used well know experiments to justify the hypothesis. When the one experiments she did quote by Sergey Metalnikov even concludes...

"The difficulty for the investigator lies not so much in inducing such responses, but in employing the proper controls, both immunological and psychological, in order to demonstrate that these responses exist and to explore the underlying mechanisms." - http://www.ncbi.nlm.nih.gov/books/NBK3908/

I guess a book entitled "What I think" would not sell as well.