Abstract
This paper reviews empirically based research articles published in the LLJ between 1990 and 2002 in terms of readability and relevance for the readership. A set of evaluative criteria for conducting and reporting research in applied linguistics was used for the review. It was found that especially in the first five years of the LLJ there was a predominance of survey-based quantitative research, and that many articles did not fulfil the criteria for conducting and reporting research. Very few studies of this period were concerned with SLA phenomena. In the period from 1996 onwards, there was a clear improvement in the reporting style and therefore the readability of the articles. The research methodology, however, remained restricted to quantitative research, and in many studies only one data source was used. Some articles in the later period still had shortcomings in the reliability and relevance of the research. The need for more research that is based on a greater variety of methods and data sources for the LLJ readership is discussed. There is a particular need for more qualitative studies, leading to a deeper understanding of language learning and teaching processes, while at the same time providing a platform for teachers' research in their own classrooms.