The Keyed Answer vs. Two Defensible Readings
An 11th-grade team has finished a unit on Dickinson and Frost and administers a 10-item multiple-choice quiz on "My Life had stood - a Loaded Gun." Item 7 reads: "The tone of the poem's final stanza is best described as (A) triumphant (B) resigned (C) playful (D) bitter," and the key is A. In the post-quiz discussion, 14 students defend A; 9 argue for B, several quoting "Though I than He - may longer live"; and two students say the ending is deliberately unresolved. The department chair wants the quiz placed in the common bank because a colleague teaching the same grade will use it next month, and asks the teacher to submit it as is so scores are comparable across sections.
1. What should the teacher do about Item 7 before it enters the common bank, and what is the reasoning?
Read the worked answer
The item should not be banked as written. The problem is not that students disagreed; it is that the format forces a single key onto a text where more than one reading is defensible, so the item measures agreement with the test-writer's interpretation rather than the skill of reading closely. A student who argues B and cites the shift into conditional, unresolved language is doing exactly the work the standard asks for and would be marked wrong.
Options, in order of preference:
- Rewrite as constructed response: "Choose the reading you find most defensible and support it with two pieces of textual evidence, including one image or figure of speech." Score with a short guide listing acceptable readings and the quality of evidence, not the position chosen.
- If it must stay multiple choice, tighten the stem so one option is clearly better warranted by a specific line, and add a required one-sentence justification that is also scored. A justification item makes the reasoning visible even when the stem is imperfect.
Either way, the teacher should tell the chair why comparability is better served by a valid item than by a shared invalid one; a keyed answer that punishes defensible reading produces consistent but meaningless data.
2. How should the teacher use the 14–9 split instructionally rather than treating it as a discipline or classroom-management issue?
Read the worked answer
The split is evidence that students are already interpreting; the teacher's job is to make the warrants visible. A workable sequence:
- Position and evidence chart. Each side lists the two strongest lines supporting its reading and the one line that is hardest to explain. The B side will likely cite the conditional close; the A side will cite the closing image of the master's pillow.
- Test the readings against the whole poem, not the last stanza alone. Ask each group what its reading cannot account for. The 14–9 majority often rests on the connotation of a single word; the discussion surfaces that.
- Exit paragraph: defend one reading in six to eight sentences with two cited lines and one acknowledgment of a counterreading.
Grading the paragraph rather than polling the room rewards reasoning over consensus, and it gives the teacher a record of who can handle ambiguity and who is pattern-matching on word connotation.
3. A colleague argues that ambiguous literature "can't be assessed with objective items anyway," so the team should switch to all multiple-choice on comprehension questions. How should the teacher respond?
Read the worked answer
That move solves the wrong problem. Objective items are perfectly usable for some literary knowledge - identifying a speaker, recognizing a shift in tense, matching a term to its function - but the skill of constructing and defending an interpretation of an ambiguous text is the point of literary study, not an obstacle to it. Dropping interpretation from the assessment would raise reliability while lowering validity.
The better response is to match item type to target: objective items for knowledge and comprehension, constructed response for interpretation, and a shared scoring guide for the constructed response. If the team is worried about inter-rater consistency, the fix is calibration on anchor responses, not retreat to recall items. The teacher can offer to bring three sample student paragraphs - a strong, a middling, and a weak - to the next team meeting as anchors.