Automated writing evaluation program's effect on student writing achievement

Lester Donnie Holman, Tennessee State University

Abstract

In an ex post facto causal-comparative research design, this study investigated the effectiveness of Automated Writing Evaluation (AWE) programs on raising the student writing achievement. Tennessee Comprehensive Assessment Program (TCAP) writing achievement scores from the 2010 administration were utilized for this study. The independent variable for the study was the usage of an AWE program. The dependent variable for the study was TCAP writing achievement scores. In this study, the TCAP writing achievement scores of two groups of eighth grade students from each school were analyzed and compared. School A was the experimental group that received AWE as a treatment, and School B was the control group that did not utilize AWE as part of its writing instruction. Both schools were from similar demographic backgrounds. This study took place in rural, upper middle Tennessee. Four research questions with corresponding null hypotheses guided this study. The first three null hypotheses related to the following variables: comparison of control group and experimental group, economically disadvantaged students, and gender. For these three null hypotheses, an ANCOVA test was performed at the 0.05 level of significance. The covariate for these tests was pre-test data. The first null hypothesis dealt with the comparison of the experimental group and the control group and was rejected due to the statistically significant difference in TCAP writing scores. The second and third null hypotheses dealt with economically disadvantaged students and the gender of students, respectively, with the use of AWE. Both of these null hypotheses were retained due to no statistically significant difference in TCAP writing scores between the comparison groups. The last null hypothesis pertained to special education students and the use of AWE. Non-parametric tests were utilized for this subgroup with no statistically significant difference being found between these two groups. Therefore, the fourth null hypothesis was retained. From the study, one may conclude that students who participate in AWE programs do increase in writing achievement more than students who do not participate in AWE programs as part of their writing instruction. One recommendation for practice is that teachers should understand the value of AWE programs when working with students of different ability levels. A recommendation for further research is that a longitudinal study should be considered to evaluate the long-term results and transferability of writing skills.^

Subject Area

Education, Evaluation|Education, Leadership|Education, Middle School

Recommended Citation

Lester Donnie Holman, "Automated writing evaluation program's effect on student writing achievement" (2011). ETD Collection for Tennessee State University. Paper AAI3487560.
http://digitalscholarship.tnstate.edu/dissertations/AAI3487560

Share

COinS