jueves, 2 de enero de 2014

Cómo evaluar, según Eric Mazur

Para aquellos entre cuyos propósitos para el año nuevo esté el mejorar su docencia, aquí hay una conferencia fantástica de Eric Mazur (el que hizo famosos los clickers) sobre la evaluación: su papel central en el aprendizaje, y cómo mejorarla. Me la pasó Antonio Muñoz y hasta hoy no había podido verla, pero es más que recomendable. Abajo os pongo los momentos que más me han gustado (el resumen está en inglés porque me resultaba más fácil transcribirlo así directamente):

25min: We are using inappropriate assessment tools, we are not testing for real problems. Good tests should cover all the skills in the pyramid, in particular: making assumptions, developing a model, applying a model
And then he goes on to show a good problem, and how this compares to actual textbook problems.
30min: Unless you are able to transfer what you've learned in one context to another context, you have not really learnt.
33min: We should focus on the solution, not on procedures (which will be addressed by computers)
34min: Our grading system has become incompatible with real problem solving, because it penalizes students for making mistakes, they become risk averse instead of encouraging them to take risks.
35min: In real life we work together, we can look up things whenever we want.
36min: No matter how innovative the approach to teaching, it is assessment that drives how students learn.
38min: only lowest order thinking skills can be judged objectively
39min: four ways to improve assessment:
- mimic real life: it's about how to use information, not about retrieving it. If it can be googled, then it is not an assessment question. Aquí hay un par de videos apasionantes
- focus on feedback, not ranking. Ranking is a myth, it can be unfair on what they know
- focus on skill, not content: and plan accordingly
- resolve the coach/judge conflict: use external evaluators, or calibrated peer review

No hay comentarios: