ABSTRACT
When I began my doctoral studies in 1983, research conduct was relatively clear. The modal investigator in reading, writing, and language studies conducted research in order to identify best teaching practices, typically employing experimental designs to contrast two or more “treatments” to determine their relative effects. Hillocks (1986) describes a typical study of this sort:
Studies of this sort, designed to identify factors that contributed to higher writing scores as determined by such factors as the presence and detail of primary traits (e.g., in an argument, the presence of claims, evidence, and warrants), were de rigueur through the early 1980s. The researcher’s task was to identify treatments for contrast, determine the variables to contrast across treatments, set up the experiment to control for other variables, construct appropriate pre-test and post-test tasks and counterbalance them in the design to avoid task and order effects, develop valid scoring rubrics and train raters to evaluate the essays reliably, persuade teachers to run the study, observe the instruction to corroborate the teachers’ use of the different treatments, and run appropriate statistical tests to calculate the signifi cance of differences in the students’ change scores from pre-test to post-test and contrast the effectiveness of the experimental and control groups according to these change scores.
