The QUALOG User’s Manual (Preliminary Version)

Anne Shelly, Syracuse University
Ernest Sibert, Syracuse University

SU-CIS-83-11

Description/Abstract

In qualitative research, data analysis is the stage of organizing, interpreting, synthesizing, conceptualizing, and reporting information collected over a period of time and by various means. The researcher organizes the information into “manageable units”, searches for recurring topics and patterns of words or phrases, decides on a focus, and refines the thesis or theme so that it accounts for supporting and negative evidence (Bogdan and Biklen, 1982). The major tasks of data analysis include organizing the data, developing categories around patterns or recurring pieces of information, discovering and testing the strength of relationships among categories, and developing a theme or thesis. Using a logic programming system, we are developing and refining a collection of computer programs to facilitate the mechanical tasks of formatting, coding, and hypothesis testing which are part of the analysis stage. Logic programming is a relatively new method of writing computer programs (Kowalski, 1974) based on earlier work in mechanized theorem-proving (Robinson, 1965). Unlike conventional programs, which specify the sequence of steps the machine should follow to obtain the desired result, logic programs specify the conditions the result should satisfy, and leave the details to the computer. The notation of logic programs is that of mathematical logic, more specifically a subset of the of the first-order predicate calculus, adapted somewhat for use in the computer. Logic programs are “executed” (one also says “interpreted”) by a theorem-proving program which establishes the existence of a result satisfying the specified conditions, and yields the result itself as a by-product of the proof.