Early intervention has worked so well in reading, many educators believe it should be put to work in other key areas of education such as writing. But how do you assess children’s progress in writing if they are still at the stage when writing a letter and a word or two is an impressive achievement? A recent study in Exceptional Children, describes a curriculum-based measurement /(CBM) of writing for kindergarten and first grade that acknowledges that limited production is typical of writing for this age group.
With many CBMs of writing, children are given a prompt and asked to write as much as they can in 3 minutes. In this modification, called the Sentence Writing measure, children are given 3 minutes to write one sentence in response to a prompt.
Once the 3 minutes are up, students are given a second related prompt and have 2 minutes to write another sentence. “Scoring procedures for CBM measures of writing have predominately focused on production indices, such as the number of words written, number of correctly spelled words, and number of correct word sequences,” the authors write. “As assessment that elicits single or multiple sentences might be appropriate for students at this developmental level because the expectations for length and thematic complexity are modest.”
The Sentence Writing measure was developed by two University of Delaware researchers and tested with 233 kindergarten and first grade students from a single district in the northeastern United States. As well as measuring students’ word production (total number of words including correctly spelled words, incorrectly spelled words, nonsense words separated by a space and word sequences), the Sentence Writing measure includes a qualitative score.
The researchers note that teachers often resist scoring based solely on production. The qualitative score used in the Sentence Writing measure takes into account appropriateness of response, whether the response was in words, a sentence or multiple sentences, and whether correct spelling, punctuation and grammatical structure was used. Students are given a qualitative score of 0-3 for each sentence.
A guiding principle in qualitative scoring is that the most serious syntactical errors are those that made the response difficult to understand. Errors involving verb agreement (e.g. use of plurals) are not seen as significant because they do not obscure the meaning of the response. In this study, both production and qualitative scores were found to have “acceptable validity, interrater agreement, alternate-form reliability, and sensitivity to growth,” the authors write.
The scores were sensitive to students’ bimonthly growth, and there were also grade-level differences for all production scores and for the qualitative score. Sentence Writing was administered 3 times during a 6-month period beginning in January. The measure was administered individually in kindergarten and in small groups in first grade. At the end of the school year, first-grade children wrote 11.67 more words than kindergarten students.
For total correctly spelled words, first-grade students averaged 10.21 more correctly spelled words than kindergarten students. There was a similar pattern for total correct word sequences. At the end of the school year, first-grade students averaged 8.55 more correct words sequences than did kindergarten students. Their sentences were rated 9.91 points higher on the qualitative measure than were kindergarten students’ sentences.
To establish validity of the measure, children’s performance on the Sentence Writing measure was compared with their performance on other norm-referenced measures of written language. Kindergartners were administered the Test of Early Written Language-2nd Edition (TEWL-2; Hresko et al., 1996) and first graders were administered 2 writing sub-tests of the Woodcock Johnson Achievement Assessment 3rd Edition/Normative Update (WJ=3; Woodcock, McGrew, Schrank & Mather, 2001). These were the Spelling sub-test, a dictated spelling test and the Writing Samples sub-test, which assesses students’ ability to write a word, phrase or sentence in response to a picture and/or verbal prompt.
Prompts and scoring were developed in a pilot test with approximately 45 kindergarten and first-grade students in a different school. The authors codified the scoring for each category by double-scoring approximately 150 sentences from students in kindergarten and first grade.
Selection of prompts requires care
An example of the directions given to the children for the Sentence Writing measure are as follows: “I want you to write two sentences for me. I will tell you when to start writing the first sentence and when to start writing the second sentence. If you come to a word that is tricky, spell it the best that you can.
Write about your favorite food for lunch.” Although teachers may feel that they could develop their own prompts, the researchers caution that prompt selection requires care and consideration. For example, one sentence prompt that was tested but discarded asked students to describe a favorite movie or television show.
The prompt was not suitable because of the wide range of potential responses that might be unfamiliar to adult raters. “Most developing writers would have difficulty responding to traditional CBM prompts,” the authors write. “Sentence Writing appears to be an appropriate assessment to capture the developing writing skills of young children.
Kindergarten and first-grade children were able to respond to the task, and the majority of children were able to write at least a word or phrase.
“Curriculum-Based Measurement of Writing in Kindergarten and First Grade: An Investigation of Production and Qualitative Scores,” by David Coker, Jr. and Kristen Ritchey, Exceptional Children, Volume 76, Number 2, pp. 175-193.