This is a report from the Project on Individually Guided Mathematics, Phase 2 Analysis of Mathematics Instruction. The report outlines some of the characteristics of probability measurement procedures for scoring objective tests, discusses hypothesized advantages and disadvantages of the methods, and reports the results of three experiments designed to learn more about the technique and compare it with standard procedures of scoring objective tests. The procedure used required the students to specify a degree of belief probability for each of the given alternatives to a question. The students were given a multiple-choice item and asked to specify what they believed to be the probability of correctness of each choice. The initial intent of these experiments was to see if a non-standard test-taking and scoring procedure would provide useful, reliable information for such tests. The studies indicated that the problem of getting useful, reliable information on difficult tests has not been solved. (Author/FL)