Exploring the Use of ESL Composition Profile for College Writing in the Indonesian Context

Abstract

Assessing writing is a demanding task. If a lecturer of writing is not prepared with a reliable scoring rubric, the students’ real performance might not be known. One of the well-known English as a second language (ESL) writing rubric is the Jacobs ESL Composition Profile which was developed by Jacobs, Zingraf, Wormuth, Hartfiel, & Hughey in 1981, known as Jacobs ESL Composition Profile. This scoring rubric is popular among writing teachers and researchers to score students’ composition. The present study is intended to 1) find out the internal consistency between raters who use the scoring rubric to assess the students’ essay, and 2) describe the level of the students’ writing performance when assessed by using Jacobs ESL scoring rubric, and 3) describe the raters’ opinion when using the profile. The study uses descriptive quantitative design. The instruments to collect the data are documentation and interview. The data were collected in three months, from February to April 2020. The subjects of the study were two writing lecturers who taught English as a foreign language (EFL) and became raters for a research grant. The raters were asked to score 37 essays of the fourth-semester students. The result of the study shows that there is internal consistency between rater 1 and rater 2 when scoring the students’ essay by using Jacobs ESL Composition Profile is high (r. = 0.674, α = 0.00 < 0.05). The Cronbach alpha analysis also shows 0.722 which indicates a strong and high level of consistency. The students’ writing performance fall in the average – good (moderate) level (49%), very good (high) level (40%), and only five students (11%) fall in the below-average – poor (low) level category. The result also reveals that Jacobs ESL Composition Profile is considered reliable to score essays even though it requires skills and practice because of its detailed description

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 15 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.