Saturday, January 15, 2011

What's a Good Scientist and Why Should One Be "Good"?

http://www.alanaragonblog.com/2010/02/19/a-retrospective-of-the-fructose-alarmism-debate/#comment-1334
Jamie Hale, posting on Alan Aragon's blog, wrote the following:

What is a Real Scientist?
Although a person with a science degree might claim to be a scientist, the true test of the scientist is how one thinks. A good scientist:
· Accepts nothing in science absolutely.
· Is willing to change his opinions based on new data.
· Does not rely on Authority.
· Thinks critically.
· Knows that extraordinary claims require extraordinary proof.
· Has an open mind.
· Relies on logic and reason.
· Knows how to form hypotheses and test them.
· Respects the scientific method.
· Examines all the data, not just the data that support his or her view.
· Builds on the work of others, giving them appropriate credit.
· Documents his or her experiments so they can be duplicated by others.
· Knows that if a claim is made, the claimant must provide the proof. (It is not up to others to disprove it.)
· Is intellectually honest
. And drops the ego and admits when their belief system has been refuted by scientific data. It’s not enough to consider the weight of data, but the quality and strength must also be considered.


I like this quite a lot, and will enjoy chewing on it.

I assume he would agree that the point of being a good scientist is to learn what is true, and these techniques are proven, if not the only, approaches to accomplish that end.

Aragon comments in his blog that 99% of the science done on human health and weight loss is useless.  A few moments reflection on how hard it is to do good human science would confirm that assertion.

One has to get to work on themselves, or on helping others, knowing that there's a mountain of uncertainty, most of the science is not conclusive, and that something's eventually going to emerge that will render your best understanding obsolete.  In the mean time, you and I can do the science on ourselves, testing options and measuring results, to see what works for with the only test population for which we're responsible.

No comments:

Post a Comment