This paper proposes a method of handling limited problems in dialect research. In approaching the problem, it was necessary to devise a system for coding phonetic transcription which would take into account the variance in the diacritics of different field workers so that none of the material would be lost while permitting computer analysis. The design of the program also allows the researcher to isolate the significant variables found in the dialects examined. The author presents the coding system, the program organization and deck assembly instructions, a listing of the program and all the subroutines, and the informant coding. An accompanying computer print-out is available for inspection at the ERIC Clearinghouse for Linguistics, 1717 Massachusetts Avenue, N.W., Washington, D.C. 20036. Copies of the print-out are also available from the author at the Illinois Institute of Technology, Chicago, Illinois 60616. (Author/DO)