The article is devoted to the influence of subjective representations about one's capabilities on sensorimotor task solution efficiency. The similarities and differences of theoretical concepts of A. Bandura and V.M. Allakhverdov in relation with this issue in question are discussed. The problem of contradictory predictions of the empirical results in a situation where the task is performed well above the expected level is raised. As a part of experimental design, two tasks are set: how to influence the real efficacy of subjects and how to measure their representations about their capabilities. The first problem might be solved with the help of illusory stimulation and distorted feedback; the limitations of these approaches are also discussed. To tackle the second problem, various first-order and metacognitive measurement methods, commonly used in experimental studies of consciousness, are discussed; their advantages and disadvantages are described in the context of the problem. The conclusion states the possible combination of the methods described in the experimental design, but it also states the necessity of a mandatory series of preliminary studies to select and calibrate the stimuli.