My theory will make psychological predictions, and it will be possible to test the theory using psychological experimentation. I will probably not run these experiments, as my focus is on computer science aspects, but I will at least describe possible experiments that could be run to test my theory. These descriptions will help demonstrate my theory is falsifiable and making testable, non-vacuous claims.
I will describe one possible experiment as an example. One prediction of my theory is that visual representations will be used, primarily, when causal representations fail. Typically these analogy experiments involve the participants reading about problems, solved and unsolved. My theory predicts that if symbolic mismatches in the text description (regarding objects and their causal properties) participants, when solving the problem analogically, will make more diagrams and experience more mental imagery.