Many believe that spatial reasoning and visualization contribute to success in engineering. To investigate this view, we a) studied how students in engineering and engineers in professional practice solved spatial reasoning problems, b) designed and implemented spatial strategy instruction, and c) characterized the impact of spatial instruction on engineering course performance. In the span of 4 years, over 500 students have used our spatial strategy instruction that includes hands-on activities, innovative computer courseware, and problem-solving assessments. We studied 153 students in an introductory engineering course. Overall, students made significant progress in spatial reasoning. In addition, gender differences in the ability to generate orthographic projections on the pre-test disappeared on the post-test. Spatial reasoning ability was a significant predictor of overall course grade, and strong spatial skills were necessary for success on the course exams. Spatial strategy instruction helps students build a repertoire of approaches for engineering problem solving and contributes to confidence in engineering, especially for women. We recommend starting instruction on spatial strategies used by practicing engineers in introductory engineering courses and building on these skills throughout the curriculum.

To help introductory programming students gain an integrated, generative understanding of LISP, we designed, implemented, and evaluated the LISP Knowledge Integration Environment (LISP‐KIE). The LISP‐KIE reflected a conceptual framework which featured (a) scaffolding of students as they control their own learning rather than telling students what they should know and (b) activities engaging students in expert problem‐solving practices rather than exercises emphasizing syntax and small problems. We conducted two in‐depth studies and one comparison study to show that the LISP‐KIE fostered knowledge integration. By knowledge integration, we mean linked, organized, and connected information about such aspects of programming as design, testing, specific problem solutions, and self‐monitoring.

This paper describes the experiences encountered by new users of a natural language interface for ad hoc query of a relational database. During the study, subjects with wide ranging computer experience performed queries of varying complexity using a commercial natural language system. Their experiences are compared with those of similar subjects working on artificial language and graphical user interfaces doing the same queries. The study revealed strengths and weaknesses of the natural language interface studied as well as natural language interfaces for database query in general, and it showed that interaction with the natural language interface was qualitatively different than interaction with either of the other systems while the overall performance was quantitatively very similar.

This paper describes a human factors experiment performed to compare three different interface styles for database query. Over sixty subjects with wide ranging computer experience performed queries of varying difficulty using either an artificial, graphical, or natural language interface. All three interfaces were commercial products. The experiment showed that none of the interfaces was best for all queries or users.

The authors describe an exploratory study performed to compare three different interface styles for ad hoc query to a database. Subjects with wide-ranging computer experience performed queries of varying difficulty using either an artificial, a graphical, or a natural language interface. All three interfaces were commercial products. The study revealed strengths and weaknesses of each interface and showed that interaction with the natural language interface was qualitatively different than interaction with either the graphical or artificial language systems