Various types of calculators and computers are available today to assist in the manage- ment and analysis of research data. Much of this machinery allows persons untrained in computer programming and statistical methods to manage and analyze their own data. These machines have tremendous computational ability but cannot make any decision as to the appropriate analysis to use. Thus, a problem arises if the researcher chooses the wrong analysis, because the calculator will compute this analysis and cannot determine whether it is right or wrong. Even if someone else performs the statistical analysis, the researcher is the one who must justify the chosen analysis when presenting the results, and he must accept the responsibility for it.
Ronald G. Marks (Mon,) studied this question.