Facilitating student growth and development in diagnosing and solving technical problems remains a challenge for technology and engineering educators. With funding from the National Science Foundation, this team of researchers developed a self-guided, computer-based instructional program to experiment with conceptual mapping as a treatment to improve undergraduate technology and engineering students' abilities to develop a plan of action when presented with real technical problems developed in collaboration with industrial partners. The pilot-testing and experimental portions of the project confirmed that the subjects found the training interesting and useful. However, the experimental data did not support the conclusion that the treatment caused an increase in diagnostic performance. Analysis of individual data demonstrated that for a portion of the sample, significant gains occurred. Qualitative analysis demonstrated that the majority of the students treated the experience as a trivial academic exercise, which seriously limited their efforts. A major step forward occurred in the ability to automatically compare student maps with the expert's reference map using a modified version of the similarity flooding algorithm. Student feedback was based on this automatic comparison. In all, the experience has encouraged the team to continue and expand the use of conceptual mapping as a tool for improving diagnostic and problem-solving skills.