ABSTRACT
Chapter 7 reports findings concerning differences (across task types and proficiency levels) and changes (across test occasions) in sociolinguistic and strategic competencies as reflected in the students’ texts. Each competence was measured holistically using human ratings of register and metadiscourse use as well as several computer-based indices of register (incidence of contractions and passivizations) and metadiscourse use (i.e., incidence of textual, engagement and stance markers). Findings indicated a general trend of improvement in terms of human ratings of register and metadiscourse use after instruction, but findings concerning changes in specific computer-based indices varied across indices. There were also significant differences in terms of some of the register and metadiscourse indices across tasks and learner proficiency levels. Tables and graphs show the significant main and interaction effects for test occasion, task type, and proficiency level on indices used to measure sociolinguistic and strategic competencies.
