The Data Wall vs. the Role-Play: Defending a Skills-Based Summative

Frame the problem as validity, not as stubbornness. The unit's target is a performance skill: refusing pressure under real conditions. A 30-item multiple-choice test can sample knowledge about refusal (what a refusal sounds like, which strategies exist) but cannot observe whether a student produces a refusal while being pressured. If the multiple-choice score becomes the unit's reported outcome, the number on the data wall will answer a different question than the one the curriculum asks. That is a construct-validity problem, and it is exactly the kind of thing a board can be embarrassed by later.

At the same time, treat the principal's need as legitimate. Administrators are accountable for showing growth. Opposing the request without offering an alternative will get the role-play quietly overruled.

Proposal:

1. Keep the performance task as the summative and report it quantitatively: percent of students proficient (scoring 3 or 4 on at least three of four criteria), plus criterion-by-criterion breakdowns. An analytic rubric produces numbers.
2. Answer the 'subjective' objection with procedure, not assertion. Build an anchor set of three scored exemplar videos or live demonstrations (below, approaching, proficient). Run a 75-minute calibration with the department. Double-score a 20 percent random sample and report inter-rater agreement. Subjectivity is reduced by calibration and moderation, not by changing the item type.
3. Add a short knowledge component (10-12 items on refusal strategies and pressure tactics) so the data wall can display a paired 'knowledge plus skill' profile. This satisfies the appetite for a test without letting the test stand in for the skill.
4. Anchor the SLO to pre/post performance, with the baseline collected in week one of the unit and the post in week five, leaving a week for scoring and reporting.

If the principal insists, propose a two-section pilot comparing both measures and a written one-paragraph rationale citing the district curriculum and the standards. That converts a values conflict into an evidence question you can win with data instead of volume.

Design the task so students generate the evidence while also practicing. Run the role-play in triads: one student applies pressure, one refuses, one scores with the same rubric. Before any scoring, spend 20 minutes training the class on the rubric using two teacher-modeled exemplars so students can recognize a 2 versus a 3. This does three things at once: it gives every student repeated practice, it makes the criteria transparent, and it distributes the first pass of scoring.

Teacher scores only the summative. Student-scored rounds are formative - feedback, not grades. During the summative round, the teacher circulates and scores six to eight triads per period, so a full section is scored across two class meetings. With 12 sections, that is a real cost, so protect it by scheduling the unit so the summative days fall on days when the Chromebook cart is not needed for other work.

Guard against drift. Score one criterion at a time across several students rather than one student across all criteria; halo effects drop sharply. Keep a simple scoring sheet with the four criteria as columns and brief evidence notes, which also protects you if a grade is challenged.

Build in a second rater for high-stakes cases. Any student scoring a 1 or a 4 gets a second look, either by you re-scoring later without the first score visible or by a colleague. This is where most grade disputes originate.

Report honestly. Publish the percentage proficient by criterion alongside the class median. If 'proposing an alternative' is the weakest criterion schoolwide, that is the most useful finding the data wall could show, and it justifies the next unit's revision. None of this requires an aide; it requires structuring student time so practice and evidence collection are the same activity.

Separate what is yours to decide from what is not. Their classroom assessment choices are theirs unless the department has agreed on common summatives; if it has, that agreement is the right lever, not a personal appeal.

Lead with artifacts, not argument. Invite them to a 20-minute scoring session using three anonymized role-play recordings or live demonstrations and the rubric. Most skepticism about performance assessment dissolves when a teacher discovers the rubric produces consistent judgments and reveals learning that a quiz hides. Bring the calibration data and the criterion-level results once you have them.

Name the equity issue directly but without accusation. Students with strong recall and weak self-advocacy score well on the quiz; students who can actually handle pressure but test poorly do not. If the department reports only quiz data, the students who most need the skill will look fine on paper.

Escalate only if the pattern matters institutionally. If the building's health data are used for placement, intervention decisions, or the school improvement plan, raise the comparability question with the department chair or curriculum coordinator and ask for a common assessment agreement. If the data are purely local and informal, note your concern once, share your materials, and move on - spending political capital here costs you the ability to spend it on the role-play itself.