CLRP, and the data generated by it, represented an important point in the evolution of dispute processing research. Three basic approaches could be used to collect data on dispute processing behavior. Each one would select a different fundamental unit for sampling -- the case, the institution, or the participant. Because each of these approaches has limits, a mixed strategy is more useful. CLRP chose the case as the common denominator and designed a sample for a case-focused data set, using a combination of the three approaches. The project developed a sample stratification matrix to permit both institutional comparisons and comparisons among the various types of participants. The project then devised a mixed sampling procedure to ensure a sufficient number of observations in each of the cells of the matrix. Several practical problems of implementation arose, and appropriate steps were taken to deal with them. The project succeeded in completing 3,873 interviews with dispute participants, covering 2,011 disputes. Two serious problems will face any future surveys of this type: how to locate the actual litigants and how to identify appropriate organizational representatives. This study shows that survey methodology, particularly telephone surveys, can be used to study complex phenomena. A major legacy of CLRP will be in the data collected, data will ultimately be available to the scholarly community to explore and exploit. About 800 references, 25 notes, and 1 figure are included.