In this paper we: 1) discuss the need for quantitatively reproducible experiments in the study of top-down design; 2) propose the design and writing of tutorial papers as a suitably general and inexpensive vehicle; 3) suggest the software science parameters as appropriate metrics; 4) report two experiments validating the use of these metrics on outlines and prose; and 5) demonstrate that the experiments tended toward the same optimal modularity.

During the past few years, several investigators have noted definite patterns in the distribution of operators in computer programs. Their proposed models have provided explanations for other observed software phenomena and have suggested possible relationships between programming languages and natural languages. However, these models contain notable deficiencies.

Nearly a generation ago, I was involved in sizing and scheduling a Navy programming project. Because it was estimated to be "half-again" as complicated as the one we had just done, it was predicted that it would require 75 man-years. When the system went to sea on time and on budget, those of us who had participated in the scheduling thought that we had demonstrated scheduling ability. None of us realized that we might instead have only demonstrated that a tossed coin can come up heads five times in a row.