Definitions. Interpretation of the categories listed in the data tables (see questionnaire) have been left to the perception of the program officer providing the data. Correspondingly, the data suffers from the problem that program directors who provided data may use the defined categories somewhat differently leading to unconstrained overlap in definitions. Since a uniformly unambiguous acceptance of definitions is an intractable task, it was felt, that in the compromise between rigor and the efficacy of collecting the data, that a more flexible approach was acceptable. However, it should be noted that the program staff providing data are professionals, who have a comprehensive grasp of their fields and the diversity of federal support. In addition, a fair amount of information exchange and shared programs among the different program managers assures that each manager has a reasonable grasp of the scope and style of other federal programs leading to comparable definitions.
Research Category under Agency Functional Goals. The category, ''Research," listed as one of the functional goals of the agencies has not been expanded to identify the ultimate purpose of the research. Correspondingly, it is not possible to identify how the research budget relates to the agencies' missions as shown in their functional categories.
Classification of "Solid-Earth Sciences. " By design the classification of "solid-earth sciences" that has been used in very broad. For example the areas of soil science, cartography, and bathymetry have been included. The data base includes all applied areas that may benefit from the application of skills derived from training in the basic disciplines of the solid-earth sciences, or use basic solid-earth science information as essential components to accomplishing mission goals. But, the data base can be used more selectively, because separate categories can be individually identified.
Programs Not Included in the Data Base. There are a few programs from which data were not collected and there is always the possibility that programs have not been identified. In terms of the total expenditures it is estimated that essentially 85 to 90 percent of the solid-earth science activities are reported. Unidentified programs are probably small because the major efforts are well known.
These points caution judicious use of the data base. One may not expect accounting accuracy and the exact figures in each pigeon hole should be assigned some error. An estimate of error is difficult but is probably reasonable to expect that values are within ten percent of the true numbers. The best use of the data is to get a qualitative to semiquantitative estimate of the scope of the federal agencies in the solid-earth sciences. In the latter sense the data are the only comprehensive accumulation of solid-earth science information that is currently available. Moreover, the information has been accumulated at the working level where there is a high degree of knowledge for the technical and scientific contributions and programmatic content of the federal solid-earth science effort.