Publications & Resources

Impact of Selected Background Variables on Students’ NAEP Math Performance

Jul 1998

Jamal Abedi, Carol Lord and Carolyn Hofstetter

The effects of students’ background characteristics on their NAEP math performance was examined in this study. Secured NAEP math items were administered to 1394 eighth-grade students from schools with large Spanish-speaking student enrollments, sizable LEP student populations, and varying socioeconomic, language and ethnic backgrounds. Three test booklets were developed (original English, linguistically modified English, original Spanish) using the 1996 NAEP Grade 8 Bilingual Mathematics booklet. The three booklets were randomly assigned to the students within a given class. All booklets contained the same math items, differing only in their linguistic demands. During the linguistic modification process, only linguistic structures and non-technical vocabulary were modified; mathematics vocabulary and math content were retained. The results of our analyses suggested that students performed highest on the modified English version, lower on the original English version, and lowest on the Spanish version of the math assessment. Additionally, non-LEP (fluent English proficient, initially fluent in English) students performed better on the math test than LEP students, both in general and across test forms. These results were maintained even after controlling for students’ reading proficiency. Finally, students may have performed lower on the Spanish version because, in most cases, the language of instruction was English only or sheltered English. Additional analyses suggested that students tend to perform best on math tests that are in the same language as their math instruction. The results of this study also indicated that clarifying the language of the math test items helped all students improve their performance. Certain types of linguistic modifications may have contributed more than others to the significant math score differences. Multiple regression analyses, predicting math and reading scores from students’ background questions, indicated that language-related background variables, such as length of time of stay in the United States, students’ grade point average, and the number of times the student changed schools, are good predictors of students’ performance in math and reading.

Abedi, J., Lord, C., & Hofstetter, C. (1998). Impact of selected background variables on students’ NAEP math performance (CSE Report 478). Los Angeles: University of California, Los Angeles, National Center for Research on Evaluation, Standards, and Student Testing (CRESST).