We conducted a series of experiments to investigate the extent to which prompting the construction of a force diagram affects student solutions to simple mechanics problems. A total of 891 university introductory physics students were given typical force and motion problems under one of the two conditions: when a force diagram was or was not prompted as part of the solution. Results indicated that students who were prompted to draw the force diagram were less likely to obtain a correct solution than those who were not prompted to solve the problem in any particular way. Analysis of the solution methods revealed that those students prompted to use a diagram tended to use the formally taught problem-solving method, and those students not prompted to draw a force diagram tended to use more intuitive methods. Students who were prompted to draw diagrams were also more likely to depict incorrect forces. These results may be explained by two factors. First, novice students may simply be more effective using intuitive, situational reasoning than using new formal methods. Second, prompting the construction of a force diagram may be misinterpreted by the student as a separate task, unrelated to solving the problem. For instruction, the results of this study imply that ignoring students' prior abilities to solve problems and their necessary developmental stages in learning formal problem-solving techniques may lead to serious mismatches in what is taught and what is intended to be learned. (Contains 4 tables and 8 figures.)