You take a running record, mark the miscues, calculate the accuracy rate, and write the number in your gradebook. Next assessment window, you do it again. The number moves a little, or it doesn't, and either way the record itself rarely turns into anything a student experiences differently on a Tuesday.
That's the gap. A running record is diagnostic — it tells you *what kind* of reader is in front of you, not just *how good*. The miscue pattern (self-corrects vs. doesn't, meaning-based guesses vs. visual-only guesses, drops endings, ignores punctuation) is the actual instructional information. Most of it gets lost the moment it's reduced to a percentage.
What the miscue pattern is actually telling you
Two students can land at the same 92% accuracy rate and need completely different instruction:
- Student A substitutes words that keep the sentence meaningful ("house" for "home") but never self-corrects, and drops most word endings. That's a fluency/monitoring gap — they're reading for gist, not precision.
- Student B substitutes words that are visually similar but nonsensical in context ("form" for "from") and self-corrects often. That's a decoding-first, comprehension-monitoring-second profile — the phonics is shaky but the self-checking instinct is intact.
Same score, opposite next steps. Student A needs practice noticing when meaning breaks down. Student B needs more decoding reps before comprehension work will stick.
Turning the pattern into a plan
Here's the actual workflow:
1. After the running record, write down the miscue pattern in plain language — not just the tally, the pattern (e.g., "substitutes visually similar words, rarely self-corrects, strong on high-frequency words").
2. Feed that pattern, the accuracy rate, and the student's current level to an AI tool along with the standard/skill you're targeting. Ask for 2-3 concrete, dated intervention activities that match the specific pattern, not a generic "reading intervention" list.
3. Set a date to re-check. A running record without a re-check date is a data point. One with a re-check date three weeks out is a plan.
Our [assessment blueprint tool](/assessment-blueprint) helps you decide what to actually check for at that re-check point, so the follow-up assessment is built around the same skill you targeted, not a generic level check. And once you know the specific gap, our [differentiation plan tool](/differentiation-plan) can turn the miscue pattern into the actual tiered tasks and scaffolds a student works through between now and the re-check, so the intervention has a shape instead of being "read more."
A caution on over-fitting to one record
One running record is one data point on one day, with one text. A student having an off day, or a text that happens to be full of names and technical vocabulary, can produce a miscue pattern that isn't really their pattern. Before building a three-week intervention plan around a single record, check it against the last one or two — if the pattern is consistent, it's real. If it's a one-off, hold the intervention and take another record on a different kind of text first.
What still needs your judgment
- Reading level and reading confidence are not the same thing. A student who's technically ready to move up a level but is anxious about it needs a different conversation than a straightforward level bump.
- The re-check date is a commitment, not a suggestion. It's easy to let it slide when the week gets busy — put it on your actual calendar, not just the intervention document.
- Not every miscue needs fixing. Some substitutions (dialect-consistent pronunciations, for instance) aren't errors at all — know your students well enough to tell the difference before you build a plan around "fixing" something that isn't broken.
