Computers Improve Students' Writing Performance on Tests
Testing Accommodations Summary Computer-Based Read-Aloud Tools: An Effective Way to Accommodate Students with Disabilities Computers Improve Students' Writing Performance on Tests Using Computers for Reading Tests Does Not Affect Scores Computer Accommodations Must Begin as Classroom Accommodations
Most state tests require students to write essays using paper and pencil, even though these students are accustomed to using computers for writing in class. Several small studies conducted over the past decade indicate that this practice may lead to inaccurate test scores for students. While studies initially found that students with disabilities could benefit from a shift to computers, it soon became clear that as computer use becomes more ubiquitous, all students can benefit.
Study Design
To examine how the use of computers for writing tests affects student performance,
the Enhanced Assessment Project conducted a large-scale study involving
approximately 1,000 students from three states. Students were randomly assigned
to perform the same writing test using paper-and-pencil, a basic text editor
that did not include spell or grammar checker, or Microsoft Word, which
does include a spell and grammar checker. Students' essays received two
scores; topic development (detail, supporting evidence, organization) and
use of standard English conventions (spelling, punctuation, capitalization,
grammar). All essays were transcribed to computer text so that the readers
could not tell whether the essay was originally produced with a computer
or on paper.
Students Do Better when They Write Essays on Computers
This study found that, on average, students performed better when they wrote
their essay using a computer. Students who had stronger keyboarding and
computer skills performed much better when writing their essays with computers.
Students with less developed keyboarding and computer skills performed about
the same using a computer or paper-and-pencil. (In past research, students
who had weak computer skills tended to perform worse when writing with a
computer.)
Further analyses looked at differences between boys or girls, minority or non-minority students, students who received special education services, or students for whom English was their second language. While the effect of using a computer did not differ for most groups, some differences appeared among minority students. Specifically, minority girls tended to perform about the same whether using a computer or pencil-and-paper, while all other groups tended to perform better when using a computer.
Recommendation
Because students are increasingly using computers for writing, a computer
format for writing tests is recommended. If the sample of students included
in this study is representative of students in New England, then the majority
of students have obtained sufficient computer skills to benefit-or at least
not be harmed-by composing an essay on computer.