A Study Of Mixed Methods Research Used In Language Teaching Research

Language, as an important tool for human beings to communicate and exchange information, serves as a bridge between human being to understand each other. For many years, language teaching has been a hot point in both China and foreign countries. With gradually development, language teaching has now been regarded as the narrow sensed applied linguistics.Mixed methods research is commonly known as the combined use of both qualitative and quantitative components in a single study. In social sciences research, the mixed methods research is called "the third methodological movement"(Tashakkori & Teddlie,2010:ix), "the third research paradigm"(Johnson & Onwuegbuzie,2004:15). As a new research approach, mixed methods research (MMR) is developing fast, especially after the publishing of the first edition of Handbook of Mixed Methods in Social & Behavioral Research in 2003, which helps to legitimize MMR as an alternative to qualitative and quantitative methods. Now it has been widely applied in many social sciences including language teaching.In this paper, a thorough study on MMR in language teaching is conducted, based on empirical articles published between 2003 and 2013 on Language Teaching Research-a major western journal of foreign language teaching. This study employs content analysis as its main method. It is found that 24 articles employ mixed methods, taking a percentage of 15.7 among the overall 153 empirical articles. Based on priority or dominance given to quantitative and quantitative components, these 24 MMR are categorized into three types-equal status, (equal priority), quantitative priority (QUAN+qual), and qualitative priority (QUAL+quan). Based on sequence of data collection, they also can be classified into three types-concurrent design, quanâ†'qual sequential design, and qualâ†'quan sequential design. Based on stage of integration, they can be grouped into three types-integrated at research question stage, integrated at data collection and data analysis stage, and integrated at interpretation stage. It is found that unequal priority design takes more than equal priority design; sequential design takes more than concurrently design; most integration occurs at data collection and data analysis stage.