The Composite Score That Hides Two Students
Your fourth-grade team gives a computer-adaptive universal screener three times a year, and the fall window closed Friday. It reports a single composite percentile plus subscores. Your principal has read only the composite and has asked each teacher to submit 'the bottom 20 percent' for a new 30-minute Tier 2 block that begins in two weeks and will replace part of the science block. Your 26 students' composites run from the 9th to the 96th percentile. Two students complicate the list. J. scored at the 14th percentile overall but earned 85 percent on your place-value unit test, reads fluently, and explains her reasoning out loud. D. scored at the 71st percentile, yet she is failing your weekly application problems; her papers show correct computation followed by an answer to a question that was never asked. You have subscores for both.
1. What does each student's pattern suggest, and what would you check before either is placed?
Read the worked answer
The composite is an average, and averages hide profiles. J.'s 14th percentile is roughly one standard deviation below the mean, but a screener score carries a standard error, so treat it as a band, not a point. The key question is which subscore is dragging the composite down. If it is computation or fluency, the intervention target is different from comprehension, and her strong place-value performance and fluent reading suggest her conceptual understanding is intact - she may have a narrow fluency gap, or she may have had a bad testing session. Check her subscore pattern and her instructional history: did she miss the fluency-building unit?
D. is the false negative the rule will produce. A 71st percentile composite says 'probably fine,' but her weekly work shows a consistent failure to select the operation or answer the question asked, which a computation-heavy screener may barely sample. That is a comprehension and self-monitoring issue, not a calculation issue, and it will not be caught by a percentile cutoff. Neither student should be placed or excluded on one score.
2. How would you respond to the principal's 'bottom 20 percent' rule?
Read the worked answer
Respond to the rule rather than to the principal. A single-measure, single-time-point cutoff produces both false positives - a fluent student having an off morning, or one unfamiliar with the device - and false negatives like D. It also cannot tell you what to teach, only who scored low.
Propose a defensible decision rule: a student is placed if at least two of three criteria are met - composite at or below the 20th percentile, a relevant subscore at or below the 20th percentile, or classroom evidence (unit tests, running records, work samples) showing the same skill deficit. Where the measures disagree, teacher judgment with documented evidence resolves the case. Ask that the intervention target the identified skill and that progress be monitored every two to three weeks with a brief measure, with a defined exit criterion, so the block does not become permanent placement.
Raise the cost of the schedule honestly: 30 minutes pulled from science twice a week for eight weeks is real lost instruction, and the students most likely to be placed are also those most likely to benefit from rich content. Suggest the block run outside core content or that science time be protected. Finally, note that no screener can identify a disability; child find requires multiple sources and documented intervention response.
3. What would you say to J.'s parents, who have seen the percentile?
Read the worked answer
Be concrete and avoid both reassurance you cannot support and alarm you have not earned. Explain what the screener is: a brief check that flags students for a second look, with a margin of error, not a verdict or a grade. Show them the contrasting evidence - the place-value unit test, the running record, J.'s verbal explanations - and say plainly that the picture is mixed, which is exactly why the school is not making a placement from one number.
Tell them what happens next: a short diagnostic on the specific skill the subscore flagged, classroom progress monitoring, and a decision within four to six weeks with a defined criterion for entering or not entering the intervention group. If J. does enter, describe the target skill and the exit criterion, not just the block.
Invite them to share what they see at home, and give them one concrete thing to watch. Steer clear of framing the score as a prediction about J.'s ability; percentile ranks invite parents to hear a ceiling. Close by scheduling the follow-up so the conversation has a date attached.