Visitor stats


Stats

Tuesday, July 6, 2010

How “Sound” Science Can Be Unreliable

There are three fundamental ways scientific research can produce unsound results:
1. the design of the experiment fails to replicate the conditions on the farm to the extent that it skews the results;
2. the data sets selected for analysis are not a statistically accurate representation of the phenomena being studied; and
3. the reporting of results ignores some aspects and makes prominent others in an unbalanced manner.

The causes of these flaws in scientific method can be fourfold:

1. ignorance of farming conditions
2. lack of farming skills
3. predisposition to anticipate outcome
4. predisposition to explain away findings (paradigm defence)

Following are three case studies where ‘sound science’ led to unsound results.

No comments:

Post a Comment