Oral Reading Data That Don't Add Up
Ms. Okafor's third-grade class takes winter screening. Jordan's oral reading fluency is 48 words correct per minute against a benchmark of 70. Accuracy on the passage is 91%. On a graded word list, Jordan reads at grade level with 95% accuracy. Spelling samples show phonetically plausible forms such as 'becuz' and 'frend,' but few standard spellings of vowel teams. During the passage reading, most miscues are meaning-preserving substitutions ('pony' for 'horse,' 'house' for 'home'). On the MAZE comprehension screener, Jordan scores 8 of 20; the class median is 14. In a small-group discussion after reading, Jordan explains why a character lied and supports it with two details, but cannot retell the story's opening and middle in order. The school's intervention menu offers: (a) move Jordan to Level J texts, (b) 20 minutes daily of systematic phonics, (c) repeated reading with comprehension retell.
1. Which of the three intervention options is best supported by Jordan's data, and which is least supported? Explain what each data point tells you.
Read the worked answer
Start with what the numbers actually measure. Accuracy at 91% is below the typical 95% instructional-level threshold, so the passage is hard. But the miscues are meaning-preserving substitutions, and Jordan reads grade-level words in isolation at 95%. That pattern points away from a primary decoding problem: Jordan is using meaning and syntax to compensate, not failing to map letters to sounds. The spelling sample supports partial alphabetic knowledge, not a student who needs a full phonics program. The low MAZE score is a timed cloze measure; it can be depressed by slow reading rate rather than by weak comprehension. The discussion data are the most diagnostic: Jordan makes an inference and cites two details, but cannot retell sequence. That combination suggests fluency and story-structure/retell are the constraints.
Option (c) is best matched: repeated reading builds automaticity and prosody, and structured retell builds narrative comprehension. Option (a) is not supported as a first move; lowering text level reduces demand but does not build capacity and risks limiting access to grade-level content. Option (b) is least supported unless further assessment (nonsense-word reading, multisyllabic decoding) shows a specific decoding gap.
2. What caution should Ms. Okafor apply to the MAZE score before placing Jordan in a comprehension group?
Read the worked answer
MAZE is a screening indicator, not a diagnosis. An 8/20 is below the class median of 14, but it comes from one brief timed sample. Performance on a timed cloze task depends on reading rate, attention, vocabulary, and background knowledge as well as comprehension. Jordan's oral discussion is counter-evidence: when questions are oral and untimed, Jordan makes an inference and supports it with two details. That means the low MAZE score may reflect fluency or test format rather than comprehension capacity.
Before placing Jordan, Ms. Okafor should triangulate: a timed ORF plus a retell rubric, untimed oral comprehension questions, and a vocabulary check. If retell is weak but oral inference is strong, the target is retell and fluency, not comprehension strategy alone. The decision rule should be multiple data points converging, not a single cut score.
3. Describe a three-week plan and the evidence you would use to decide whether it is working.
Read the worked answer
Keep Jordan in grade-level text with scaffolds rather than dropping levels permanently. Daily 12-15 minute repeated reading of a short poem or fluency phrase set, with a cold read, two practice reads, and a hot read; graph words correct per minute. Three times weekly, a 10-minute guided retell using a story-grammar frame (someone-wanted-but-so-then), first with a model, then with Jordan. Preview vocabulary before new texts and allow partner reading or audio support so comprehension work is not blocked by rate.
Progress evidence: ORF gain of roughly 8-10 words correct per minute over three weeks, accuracy at or above 95% on practiced and new text, and retell including at least three story elements in order. Also track whether Jordan's oral inference continues to be strong. If ORF stalls and retell stays weak despite consistent attendance, add diagnostic decoding measures and bring the data to the student study team rather than continuing the same plan.