Learning styles comparison based on a classification methodology

Learning styles have been used to explain students' differences in approaching their learning, but there are still deficiencies in interpreting the results of their application, and there are authors that indicate that there are no elements that support their credibility in achievement-based education. For this reason, we proposed a methodology to analyze the results obtained after the application of the test of learning styles. Our methodology is designed in four steps: identify the population and its characteristics, establish the time for the samples, determine the existing conglomerates in the population and to extract the characteristics of the constructed conglomerates. To evaluate the this aproach, the methodology was applied and validated by the Felder and Silverman's dichotomous test vs its diffuse version in an adult population in continuous formation and we found evidence that there are external aspects to those evaluated by the traditional learning styles instruments that could influence the changes of preferences over time.