Valid research into foreign language teaching-two recent projects

Abstract

To be useful, foreign language teaching research must be controlled and valid. Although large-scale experiments have proved to be ill-controlled and hence invalid, if ‘small-scale’ experiments are used, valid comparative research into language teaching methods is possible. Since the disappointing Pennsylvania Project (Smith, 1970) two major ‘small-scale’ experiments, both aiming at improved internal validity, have been reported: the GUME Project (1972) and the York Study (1975). Although investigating different areas, both set out to compare teaching methods in a controlled and systematic fashion. Following a detailed discussion of their research methods and findings, it is argued that, although not perfect, both studies show that it is possible to improve on the internal validity of previous experiments. In both projects, the greatest threat to internal validity came from lack of control of the teacher variable. The writer has attempted (unpublished PhD thesis, 1977) to control this variable completely, by using the language laboratory to present grammatical material. However the teacher's influence was still demonstrated repeatedly in the analysis of the results. It is suggested that future research should employ some kind of index to monitor teacher attitudes and the existing school situation, to begin to control them as one would other background variables.