What makes a good theory?

My worldview requires explanations to be consistent, efficient, useful and verifiable. They should also build on previous theories and should not claim to be absolute proof.  

The checklist below can be used to assess the quality of a hypothesis or theory. It is taken from an essay by Robert Root-Bernstein... 


A. Logic

1. A simple, unifying idea that doesn't include anything unnecessary (Occam's Razor)
2. Logically consistent (no contradictions)
3. Logically falsifiable (there are situations in which the theory would be invalid)
4. Limited (doesn't presume to explain everything).

B. Empirical 

1. Empirically testable or can lead to testable predictions or retrodictions
2. Make verified predictions and/or retrodictions
3. Leads to reproducible results which others can check
4. Includes criteria to determine whether data is factual, artifactual, anomalous or irrelevant

C. Sociological 

1. Resolve known problems, paradoxes, and/or anomalies in previous theories
2. Creates new problems and questions to work on
3. Creates a new paradigm or model to use when working on problems
4. Provides concepts which to solve problems

D. Historical

1. Meets or surpass the criteria of earlier theories
2. Explains any and all data produced with earlier theories
3. Is consistent with any and all related theories