ABSTRACT
Statistics can be viewed as the general and abstract theory of this set of procedures. It is concerned with planning the acquisition of information, with processing it when acquired, and with applying it to the solution of the problems which motivated the inquiry. The median is a statistic which takes account of all the values, without being much influenced by the extremes. The median salary of a college professor, or the median cost of psychoanalysis, is likely to be a more informative datum than the mean or the mode for these values. The objections have come from two directions. The first is expressed in the familiar doctrine that human behavior is essentially unpredictable, because it is free. The second line of resistance to the role of statistics in behavioral science is not that it tries to do too much, but that it does too little.
