The effect of process writing software on the quality and length of ESL students\u27 writing

The purpose of the study was to investigate the effect of process writing software on the quality (holistic score) and length (total number of words) of the writing produced by ESL students enrolled in an advanced reading and writing class. Four advanced-intermediate ESL students who had enrolled in an advanced reading and writing class volunteered to participate in the study. During the six weeks of the first summer session of 2004, participants received regular classroom instruction and utilized process writing software ( Essay Punch) to write two academic essays. In addition to the two essays written with process writing software, participants wrote a pretreatment essay, a take home essay---immediately after the second essay with Essay Punch---and a posttreatment essay. Three ESL experts assigned each essay a quality rating according to the scoring guidelines of the Test of Written English Guide (2004). The total number of words per essay was also calculated using Microsoft Word. The quantitative analysis did not produce conclusive results. While all participants obtained their highest quality rating in one of the two essays written with Essay Punch and also wrote their longest essay with this software, their quality ratings and word totals per essay tended to be lower in the posttreatment essay than in the pretreatment essay. To gain a better understanding of the participants\u27 experiences using the software, the researcher also gathered qualitative data through a demographics questionnaire, field notes, a posttreatment questionnaire, and a teacher\u27s survey. Qualitative data were analyzed using inductive analysis, which yielded a list of categories that was later organized into major themes. The qualitative analysis revealed that factors such as the poor integration of the software with the curriculum, the short duration of the experiment, the limited time spent writing with the software, and the individual goals of each participant may have negatively influenced participants\u27 writing performance