C3X: Correlation, Causation and Controlled Experimentation for C2
Abstract
This paper examines the key role of controlled experimentation in testing causal hypotheses on the warfighting effectiveness of C2 technologies and procedures. Through the years many hypotheses have been advanced regarding the factors making for effective warfighting. More troops, more firepower, higher speed of maneuver, superior doctrine, better training and superior C41 (Command, Control, Communications and Intelligence) are all factors hypothesized to make for more effective warfighting. Warfighting, itself is adjudged more effective when enemy combat losses appreciably exceed own force losses. How does one go about testing the many hypotheses on the causes of warfighting effectiveness against observational evidence? For example, what difference does it make to warfighting effectiveness if we put in a particular new C2 technology? We don't need to know simply what's the difference between military units with and without the new C2 technology; we need to know what actually happens with the new C2 technology compared with what would have happened without it. We shall prove that controlled experiments provide the only unequivocal tests of such causal hypotheses; otherwise the observed results are open to rival explanation in terms of causation by some of the uncontrolled factors. We introduce causal hypothesis testing with observations on a single group and then move to the method of using simple correlational data for two groups. This forces us to confront the open ended issue of control groups and control variables in testing causal hypotheses which in turn leads us to consider the most conclusive method, controlled experimentation.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 2004
- Accession Number
- ADA466544
Entities
People
- Paul J. Hiniker
Organizations
- Defense Information Systems Agency