Methodological problems in measuring teaching effectiveness

Abstract

This paper does not argue for more reliability in using traditional evaluation methods, but for abandoning some of the basic starting points of empiricist methods. Objectivity of measurement, in the manner of the natural sciences, cannot be increased by greater effort and precision, because this principle is antagonistic to the laws of social, symbolically mediated interaction. It is argued that comparative evaluation, or teacher assessment, is not important. Our aim should be to develop evaluational methods for innovational settings. Most traditional methods of evaluation assess the individual, not the system, and tend to gather the subjective opinion of the individuals. As a result, conclusions tend to be based on present views in ignorance of alternatives. In consequence it is a rather conservative approach which fixes the status quo of the system and immobilises it against more radical innovations. Much more fruitful would be the abandonment of the attempt at quasi-scientific objectivity and its replacement by on-going evaluation of the complex interactions in the innovation process via ‘action research’.