Twelve Points Below, and a Monday Deadline

Say yes to intervention, no to the decision rule as written. She should not refuse-the students below the 25th percentile need support-but she should ask for the identification to be validated before it hardens into a placement.

What is wrong with the current comparison:

• The median comparison is not fair. Her class contains all six students with reading IEPs. Comparing class medians without disaggregating compares rosters, not instruction.
• The testing conditions are suspect. Post-evacuation, right after lunch, is a real threat to validity on a reading benchmark.
• The 11 are not necessarily 11 new needs. Some overlap with the six students already receiving IEP-specified services. Those services cannot be replaced or duplicated by a Tier 2 block; the IEP governs.

Before Monday:

1. Cross-check the 11 names against IEP and 504 rosters.
2. Look at item-level data-is the shortfall spread across comprehension items or concentrated in one area?
3. Triangulate with a second measure: running records, a writing sample, last spring's state results.

Then bring the principal a short memo: here is the overlap, here is the alternative explanation, here is the group that genuinely needs Tier 2, and here is what I will start Monday. That is compliance with the spirit of the request and protection against misidentification.

Do not use it for grading, and do not simply decline it. The test's recall-only items cannot measure the standard, which requires citing and explaining evidence. If she uses it, she gets scores that look like data and cannot inform instruction-worse than no data, because they invite instructional decisions.

Handling the colleague:

• Assume good faith. He wrote what he had time to write, and a recall test is a reasonable first draft.
• Bring the standard and the test side by side and ask a shared question: 'Which items would show a kid can cite evidence?' That is a collaboration, not a critique.
• Offer to co-revise: turn three recall items into evidence items, keep the rest as a quick comprehension check.
• If he declines, use the common assessment for grading and his test only as a low-stakes recall check, and say so plainly rather than quietly sidelining it.

What she should not do: escalate to the department chair before talking to him, or use the test and later cite it as evidence about students. Both damage the relationship she needs for the rest of the year, and the second also corrupts her data.

The tempting mistake is waiting for the winter benchmark. Six to ten weeks with no intermediate check means a semester can pass on an intervention that is not working, and the students lose the time.

Build the monitoring now:

• Baseline on the specific skill the intervention targets-not a general reading score.
• Brief weekly probes (one to three minutes) on that skill, graphed.
• A decision rule set in advance: for example, if four consecutive data points fall below the goal line, change the intervention rather than waiting out the cycle.
• A separate check for the six students with IEPs, coordinated with the special education teacher, so the Tier 2 data and IEP progress data tell one story rather than two.

On the benchmark: when it is re-administered, she should check the testing conditions and disaggregate by subgroup again. If the class median rises but the IEP subgroup does not, the class average is hiding the students the intervention was meant to serve-and that is the precise error that started this case.