Assessment
Assessment
This domain tests how you select, administer, and interpret assessments for ELLs, distinguish language proficiency from content achievement, avoid bias, and use accommodations appropriately.
Core concepts
- Formative assessment is ongoing and informs instruction; summative assessment measures outcomes. Diagnostic assessment identifies needs; progress monitoring tracks growth.
- Norm-referenced tests compare students to a norm group; criterion-referenced tests compare performance to standards. Validity means the test measures what it claims; reliability means consistency. A test can be reliable but not valid for ELLs.
- Language proficiency tests (e.g., WIDA ACCESS, WIDA Screener) measure English development; achievement tests measure content learning. A low content score may reflect language demand, not lack of knowledge.
- Accommodations change how a test is delivered without changing what is measured (extended time, small group, directions read aloud, heritage-language glossary). Modifications change the content or expectations and affect score comparability.
- Test bias occurs when items assume cultural or linguistic knowledge unrelated to the construct. Review items for cultural load, idioms, and unnecessary complex syntax.
- Alternative assessments-portfolios, performance tasks, rubrics, oral interviews-can capture what ELLs know when standardized tests underrepresent their ability.
Worked example
An ELL scores below grade level on a science unit test. The teacher notices the test uses long, conditional sentences and idioms (“if the plant were deprived of light...”). Before concluding the student lacks science knowledge, the teacher should analyze the language demand, offer a glossary and simplified syntax as an accommodation, and use an oral interview or labeled diagram to check understanding. Reasoning: the construct is science, not complex English syntax; reducing linguistic load while keeping the science construct intact improves validity. A bilingual dictionary might help, but it is not automatically appropriate if it gives away answers or slows the student excessively.
Traps that cost marks
- Treating accommodations as modifications, or assuming all ELLs need the same accommodation.
- Confusing reliability with validity; a consistently biased test is still invalid for its purpose.
- Equating a language proficiency score with academic achievement.
- Relying on a single test score; multiple measures are required for high-stakes decisions such as exit from ESOL services.