On-the-Fly Translation During a Content Test: Validity and Equity
In a seventh-grade social studies class, students take a 25-item unit test in English on the Silk Road and map skills. A Cantonese-dominant English learner, who has been in the U.S. for two years, receives a bilingual word-to-word glossary as an accommodation in his English learner plan. During the test, he asks the teacher to explain 'latitude,' 'longitude,' and 'primary source.' The teacher orally translates the terms into Cantonese and, for one item about the Bill of Rights, adds a sentence explaining that the Bill of Rights protects individual freedoms. He scores 18/25, up from 11/25 on the previous test. The teacher realizes that another Cantonese-speaking English learner did not receive the same clarification, and that the district's assessment policy allows oral clarification of directions but not of content. The teacher must decide how to handle the score, whether to adjust the test, and how to prevent this in the future.
1. How should the teacher interpret and use the 18/25 score?
Read the worked answer
The 18/25 cannot be treated as a clean measure of his social studies knowledge. The accommodation plan allowed a bilingual glossary and oral clarification of directions, not explanation of content. When the teacher explained that the Bill of Rights protects individual freedoms, she supplied content that the item was designed to assess. That item, and any item affected by the explanation of latitude, longitude, or primary source, should be invalidated for this administration. The glossary and translation of directions are legitimate because they reduce language barriers without giving away content. The teacher should record exactly what was clarified, mark the affected items, and use the remaining items formatively. She should not report 18/25 as mastery or compare it directly to the other English learner's score. The increase from 11 to 18 is partly a language effect and partly a content effect, and we cannot separate them without item-level analysis. The score should trigger re-teaching and a better-designed assessment, not a grade that misrepresents what he knows.
2. What assessment design would be appropriate next time?
Read the worked answer
Next time, accommodations should be planned before the test and applied consistently. I would simplify the English stems, add visuals and a word bank for key terms, and translate directions and a glossary-not the content. For a student at his English level, I would allow an oral response in Cantonese to a prompt, scored with a content rubric, so that reading and writing demands do not mask social studies knowledge. I would separate map-skills items from reading-heavy items so that I can see which construct is being measured. I would preview vocabulary such as latitude and longitude in both languages during instruction, not during the test. I would also decide in advance which accommodations are available to all English learners at a given proficiency level and put them in the English learner plan. The goal is to reduce construct-irrelevant language load while keeping the content demand intact.
3. What equity and policy issues arise, and how should the teacher address them?
Read the worked answer
The equity problem is that one student received ad hoc content help and another did not, so their scores are not comparable. The district policy exists to prevent exactly this. The teacher should disclose what happened to the English learner coordinator, document the accommodation, and avoid using the score for placement or accountability without an item analysis. She should also speak with the student and family in a way that does not blame him: explain that the test conditions were not fair and that he will have another opportunity. On a classroom level, she should standardize a menu of accommodations for English learners, train herself to distinguish clarifying directions from teaching content, and use a bilingual glossary or translated test version prepared in advance. If she cannot prepare a translation, she should not improvise content explanations. The judgment call is between responding helpfully in the moment and protecting the validity and fairness of the assessment.