Social "Sciences"

All the social "sciences" use elements of science. But the subjects the social sciences study do not lend themselves to results as consistent as the "hard" sciences, whether or not u apply the scientific method. The environment does not support it.

I use this analogy. If you r executing the Butterfly stroke to Phelpsian perfection, but u r executing it while falling from 20,000 feet in the air. Are you swimming?

Your Butterfly technique may be useful 20,000 feet in the air. But it is not swimming. The environment does not support it.

Elements of the scientific method may be useful for a social research project. The entirety of the scientific method may be useful for a social research project. But the nature of the area of study does not allow the same degree of certainty that a hard science does.

So what? That does not mean that the work is not valuable. As far as I can tell G. B. Shaw did not use the scientific method at all. And I consider his work valuable.

Literature is 1 extreme. Physics is the other. Literature stands on its own. Any academic area that aims to be a discipline should do as physics should do. Physics papers also often fall short. <Delete?> Science works better and faster when rules r stated and followed. </Delete?> But sloppiness in physics papers usually involves insufficient detail to allow replication of the experiments. Or specious patterns derived from firm data.

In social research the data is a lot less firm. U do not get to run experiments on a large scale. And small-scale experiments do not extrapolate well. U will get much less certainty out of social research. But u can maximize the reduced certainty using the same principles that hard sciences should use. Declare your methods, why you used them, and the certainty of each factor in ur research.

Nate Silver does a good job of this on 538. Some factors are close to precise (like votes in the last presidential election). Some factors (like the polling samples matching the segments of the population the samples are supposed to represent) are known to a probability. Some factors (like the proportion of the segments of the population that will make up the eventual electorate) are an educated guess.

Silver gives the factors, their certainties, and his reasoning. Silver also usually does a postmortem analysis of his projection. Especially after hls projection is wrong. I wish all areas of research worked the same.