Flutterby™! : Autism, immunity, inflammation, journalism 2012-08-27 18:14:07.572189+00

Autism, immunity, inflammation, journalism

2012-08-27 18:14:07.572189+00 by Dan Lyke 2 comments

Emily Willingham: Autism, immunity, inflammation, and the New York Times.

I just read through a portion of Emily's post and a thought crossed my mind in regards to your Echinacea Study post:

I recently experienced a very traumatic event. After six years working on a PhD Dissertation, my committee came to the conclusion that, with a little more work, it was a decent Masters Thesis. As you are probably aware, a PhD Dissertation is supposed to advance the science in the field of study awarding the PhD. In a nutshell, my committee decided that my work didn't appear to fulfill that requirement.

This sort of thing happens all the time in science. Goals are missed. Results are inconclusive. And yet, we hate to throw out all that good work. We try to extrapolate on the results and sometimes push the boundaries of causation. The Echinacea Study did not produce statistically significant results. My dissertation did not produce original science.

It may be possible to carefully scrub the data from the Echinacea study to improve the statistical significance (which I suspect the authors already did based on the statistical methodology described in the paper). But if you do that, you have to be very careful not to try to produce significance which reflects a particular outcome. The author's decision to draw conclusions form their meaningless results makes me concerned about their ability to remain detached from the results.

Perhaps the better approach is to carefully look at the causes of variability in the study and redo the experiment with better controls. That is, do more science! You could also include surveys to try to gauge other factors, like patient-provider relationships to try to capture the qualitative aspects of Echinacea treatment.

This is the approach I am taking with my dissertation. First, I am doing more science. The work which my committee said was almost Masters-quality will be packaged up into a single publication. Other work I have been involved in will produce a second related publication. A third project will be undertaken under a new advisor to produce a third publication. The three publications will be threaded together to form a new dissertation. That is, I am doing more science!

People are loathe to give up on what appears to be genuine hard work even when the work produced erroneous results. It's like working on a Sudoku puzzle and the last square has to be a 9 but you already have a 9 in that section of the grid. People are working very hard to understand autism. It's not easy and a lot of deadends will be encountered. But we need to see deadends for what they are.

And from a slightly more oblique angle: The Science News Cycle. Th only thing they leave out is when your research ends up as a plot point on Rizzouli and Isles.