ABSTRACT
As explained in Chapter 2 , most experimental research focuses on some aspect of writers’ texts. Features of the text are usually the dependent variables that must be measured to assess the outcomes of some treatment. The reverse, however, is not true: Many studies analyzing learners’ texts are not experimental studies and instead have goals other than to assess the outcomes of a treatment. In addition, analyses of learners’ texts may be combined with other methods, particularly in case study research that seeks to describe students’ development over time with regard to text features. The various goals for analyzing learners’ texts are addressed in the first section of this chapter. In the second section, we provide a taxonomy of ways to analyze learners’ texts with a discussion of the issues related to choosing and using the techniques. We have included here mostly studies of L2 writers’ texts but include a few that also examine native speakers’ texts as a point of comparison to validate a measure or to show how learners’ texts differ from those of native speakers. The third part of this chapter focuses on issues related to using the various measures and text analysis techniques. We end with a description of how to design a study using text data, followed by a discussion of a focal study. We note that many experimental and correlational studies use holistic or analytic rubrics to assess features of students’ writing with regard to the quality of their writing. There is a large volume of research on assessing writing and validating holistic and analytic rubrics, but the development of such rubrics is beyond the scope of this book. We focus here, instead, on the more narrow and arguably objective measures to distance ourselves from the notion of quality per se.
