Sunday, February 5, 2012

How measuring clouds the effect of multiple hidden parameters

Many things are affected by multiple parameters. (heart rate affected by mineral balance, mood, room temperature, sitting of lying down etc.)

At any given time most of the effects are hidden. But many of them act simultaneously,

Many more effects are potentials. These do not act commonly, but have the potential to affect strongly, (heart drugs etc.)

Scientific investigation will prefer the measurable parameters. And also the easily manipulable ones. Thus, a drug will take precedence is research over a frown (which clearly affect the heart rate). Any clear parameter will take precedence on the subtle ones (combinations, harder to measure, or those parameter that are not easily taken to the extreme, hence do not show strong effects in laboratory manipulations)

These biases will over time make a part of the effects scientifically known, while all other effects will not be seen. And the clouding of knowledge will be considered "scientific".
Instead of full appreciation of the variety of parameters that make things happen, we will have a focus on those measurable. Those manipulable in the lab etc.

when many things affect a single outcome the bias is much stronger.
To the exclusion of all other effects, the effects that enter easier into a scientific experiment get all visibility, and the illusion builds that these are central.

In fact, there are many causes. Many of whom are subtle and not easy to see and measure.

Also, many effects are usually minor. But become large when artificially enlarged. See my piece extreme and special cases and causes

To sum up. We have complex phenomenons that are affected by multiple small and big, clear and subtle effects. Scientific enquiries will concentrate to those that are: easy to measure and Easy to take to the extreme. Creating the illusion that those are the central causes to the exclusion of all else and of complexity.

See also the academic bias. This article is a special case of the academic bias

No comments: