In this study the researcher investigated the feasibility of constructing a valid paper-and-pencil measure of problem solving ability. (Rationale and design of the study are discussed in Part 1.) The principal feasibility criterion, correlation of at least .71 with scores on taped and coded individual "thinking aloud" problem-solving sessions, was not met; however, the obtained correlation (.68) for one test suggested to the researcher that more reliable tests might achieve the criterion. Rank ordering of subjects on the "thinking aloud" procedure and written tests were highly correlated. The use of the "thinking aloud" procedure to establish concurrent validity was evaluated and questions about the validity of this procedure with seventh-grade students were raised. Investigations of the functional differences between audiotaped and videotaped interviews revealed no differences in subject performance, but supported the superiority of videotaping as a research tool. Instruments used in the study and data displays are presented in appendices to this report. (SD)