Student problem-solving ability appears to be tied to the representational format of the problem (math, pictorial, graphical, verbal). In a study of a 367-student algebra-based physics class, we examine student problem solving ability on homework problems given in four different representational formats, with problems as close to isomorphic as possible. In addition, we examine students' capacity for assessing their own representational competence by giving follow-up quizzes in which the students can choose between various problem formats. We report student performance and consider factors that may influence their ability or choices. As a control, part of the class was assigned a random-format follow-up quiz where students received quiz formats at random. We find that there are statistically significant performance differences between isomorphic problems. We also find that allowing students to choose which representational format they use improves student performance under some circumstances and degrades it in others.