Two-Minute Cultural Comparison: Score or Re-Teach?
Ms. Vidal teaches Spanish III to 24 tenth graders. For the first presentational-speaking benchmark, students have four minutes to prepare and two minutes to record a cultural comparison explaining how a celebration in a Spanish-speaking community compares with one in their own community. Marisol, a fluent heritage speaker, records 1 minute 45 seconds. Her transcript has 55 seconds describing Día de los Muertos in Mexico, 20 seconds describing her family's altar, 25 seconds listing foods, and 5 seconds saying 'and in the United States, Halloween.' She reads from notes, says 'el ofrenda' twice, and uses 'so' three times. The rubric has four equal dimensions: task completion/comparison, comprehensibility, vocabulary, and grammar. Ms. Vidal must assign a benchmark level and choose the most useful feedback for a 20-minute conference. Marisol's oral fluency is the strongest in the class.
1. What benchmark level is defensible, and what feedback should Ms. Vidal prioritize in the conference?
Read the worked answer
The defensible level is developing, not proficient, even though Marisol sounds fluent. On an equal-weight rubric, comprehensibility and vocabulary are strengths, and grammar has some control despite 'el ofrenda.' But task completion is the weakest dimension: 55 seconds describe one community, 20 seconds describe a personal altar, 25 seconds list foods, and only 5 seconds mention the United States. That is a mention, not a sustained comparison. Fluency does not compensate for missing content because the task is a cultural comparison. The conference should prioritize content and organization first: a Venn diagram with two concrete examples per community, a comparison frame such as 'A diferencia de...' and 'Mientras que...,' and a requirement to explain one similarity and one difference. Pronunciation and 'so' are lower priorities. She should recast 'la ofrenda' naturally and then have Marisol re-record a 60-second comparison, not the full two minutes.
2. Design a 20-minute re-teach that addresses the rubric weakness without embarrassing Marisol or lowering the standard.
Read the worked answer
Use a short, criterion-based workshop rather than a general 'try harder' conference. Minutes 1–3: show the rubric and identify the missing comparison, not the student's accent. Minutes 4–8: analyze a short anonymous exemplar that names two communities, gives one similarity, one difference, and evidence. Minutes 9–14: Marisol completes a Venn diagram with two specific examples from each community and drafts four sentence frames. Minutes 15–18: she records a 60-second practice segment comparing one similarity and one difference. Minutes 19–20: she self-scores against the task-completion dimension and names one revision. The standard stays the same: two communities, explicit comparison, supporting examples. The scaffold is temporary. This protects her confidence because the feedback targets the task, not her Spanish identity, and it gives her a concrete revision she can hear.
3. Why would giving Marisol an advanced score because she is a fluent heritage speaker be a rubric problem?
Read the worked answer
It would conflate fluency with presentational proficiency. The rubric separates task completion/comparison, comprehensibility, vocabulary, and grammar; a high score on one dimension cannot substitute for a missing comparison. Marisol's ease of speech may reflect home acquisition, but this task asks her to select, organize, and compare cultural information for an audience. Giving advanced for fluency would also send the message that heritage speakers are judged by accent or smoothness rather than by evidence and structure. That is a validity and equity problem: it lowers expectations for the very students who need explicit academic-language instruction. The fair decision is to score each dimension on evidence, praise her strong comprehensibility, and require the same comparison standard as everyone else. If the program wants to recognize oral fluency, it should use a separate fluency measure, not distort this benchmark.