This study evaluates the effects of an intervention aimed at improving representational flexibility in linear-function problems. Forty-nine students aged 13-16 participated in the study. A pretest-intervention-posttest design with an experimental and control group was used. At pretest, both groups solved a choice test, where they could freely select a table, a graph, or a formula, and three no-choice tests, where they used predetermined representations. Twenty-five students in the experimental group were exposed to the intervention, where they learnt to fine-tune their choices both to the task at hand and to their own characteristics as representational users. The control group was not exposed to any intervention. The posttest was similar to the choice pretest. A flexibility score was calculated per student based on Siegler and Lemaire's score. At posttest, the experimental group became more flexible than the control group. The impact of increased flexibility on problem-solving accuracy and speed is discussed.