Assessment Data That Doesn't Match the Classroom
Mrs. Nakamura teaches 4th grade in a bilingual program. On the state English reading assessment, her English learners scored, on average, 18 points below the campus mean. But in class, these same students read grade-level English texts with comprehension, discuss them in small groups, and write organized English paragraphs. Mrs. Nakamura suspects the test is measuring something other than reading comprehension for her students. She has the item-level data: her ELs missed disproportionately on items with long, complex sentence stems and on items requiring cultural background knowledge (e.g., a passage about a baseball tryout, a passage about a school fundraiser). Her ELs are mostly from rural Guatemala and have strong Spanish literacy.
1. Interpret the gap between classroom performance and test scores. What is the test likely measuring, and how should Mrs. Nakamura use this in instruction?
Read the worked answer
The pattern points to construct-irrelevant variance: the test is partly measuring English academic language and cultural schemas, not just comprehension. Evidence:
- Students comprehend grade-level English texts in class-so decoding and basic comprehension are intact.
- They miss complex sentence stems-a linguistic-demand issue (syntax, academic vocabulary), not a reading-comprehension deficit.
- They miss culturally loaded passages (baseball, school fundraisers)-a schema/background-knowledge issue.
So the score gap is real but misattributed if read as 'these students can't read English.' Instructionally, Mrs. Nakamura should:
- Teach test-language explicitly: unpack sentence stems, question types, and academic vocabulary-without turning class into test prep.
- Build background knowledge broadly (and preview unfamiliar cultural content before similar passages).
- Continue strong comprehension instruction; do not lower text complexity.
She should also flag the pattern to administration for assessment review, since a test that systematically penalizes ELs for language and culture is a validity problem, not a student problem.
2. How should Mrs. Nakamura communicate these results to families so the data is accurate but not demoralizing?
Read the worked answer
Families deserve the real score and the real interpretation, in a language they understand. She should:
- Present the score plainly, then immediately contextualize: 'On this English test, your child scored below the campus average. In class, your child reads and discusses grade-level English texts well. The test includes language and topics your child hasn't had exposure to yet.'
- Avoid euphemism that hides the score (families find out anyway and lose trust) and avoid framing it as a fixed ability.
- Give one or two concrete things families can do (continue Spanish reading at home-it supports English; talk about school and community topics to build background knowledge).
- Use an interpreter or bilingual written summary; do not rely on the child to translate.
This respects the family, keeps the data honest, and frames the gap as addressable.