How to score a writing sample against a rubric for IEP progress

A rubric turns a piece of writing into a number. For an IEP, that number has to mean the same thing in October as it does in May, whoever scores it. Most rubric data fails that test because the rubric changes, the prompt changes, or two adults read the same paragraph differently.

This page walks through scoring a writing sample against a rubric so the scores can carry a progress report. For the goals themselves, see the pillar guide to IEP goals for writing. For counting words and word sequences instead, see how to measure writing with TWW and CWS.

What you need before you start

• The goal, with the rubric named in it, such as "earning at least 10 of 12 points on a 12-point body paragraph rubric."
• An analytic rubric of 3 to 5 rows, each scored 0 to 3 or 1 to 4, with a written descriptor for every score.
• A set of equivalent prompts, enough for one every one or two weeks for the year.
• Fixed conditions: time limit, planning time, mode (handwritten, typed or dictated) and any allowed supports.
• The rubric scoring data sheet, set up with one column per row of your rubric.
• Two or three anchor papers: real samples (names removed) that show a 1, a 2 and a 3 on each row.

Step 1: Build rows that match the goal

The rubric should score what the goal targets and nothing else. A goal about organizing a body paragraph needs rows for the point, the evidence, the explanation and the link back to the claim. It does not need a row for handwriting or spelling.

Use an analytic rubric, with separate rows, rather than one overall score. A single 1 to 6 score can stay flat for months while the student gains a whole row, and the team never sees it. Row scores show which part moved.

Keep it short. Four rows of 0 to 3 points give 12 points, which is enough range to show growth over a year and few enough decisions to score in 2 or 3 minutes.

Step 2: Write descriptors a second scorer could use

Each score on each row needs a sentence that says what the writing looks like at that level. Vague levels ("developing," "proficient") produce scores that drift.

Test the descriptors before baseline. Two adults score the same three samples alone, then compare. If they disagree by more than 1 point on any row, rewrite that row's descriptors.

Step 3: Fix the prompt and the conditions

Rubric scores only compare if the task stays the same. Write the conditions into the goal and the data sheet header: prompt type, text level if there is one, planning time, writing time and mode.

Use prompts of the same type and difficulty. For body paragraphs, pair each claim with a short text at the student's reading level. For narratives, use personal-experience prompts. A prompt about an unfamiliar topic measures background knowledge and will drop the score for reasons that have nothing to do with the goal.

Probe every two weeks for most paragraph and essay goals, or monthly for long pieces such as research papers. Weekly rubric probes are possible for short paragraphs, but they take class time and give students little room to apply feedback between samples.

Step 4: Score each row against the descriptors

Score the sample with the descriptors and anchor papers in front of you. Read once for the whole piece, then score row by row. Write the row scores on the paper and on the data sheet.

Do not give credit for parts the student would have written with more time, or for what you know the student meant. Score what is on the page under the stated conditions. If you gave a prompt or a reminder during the probe, note it, because it changes what the score means.

Step 5: Check agreement with a second scorer

Have a second adult score every fourth or fifth sample alone, without seeing your scores. Count the rows where you agree exactly or within 1 point. Many teams aim for agreement within 1 point on at least 80% of rows. That is a common working standard, not a rule.

If agreement drops, sit down with the anchor papers and talk through each disagreement. Most drift comes from one row, usually explanation or organization.

Step 6: Graph the totals and make decisions

Plot the total for each sample on a graph with a goal line from the baseline median to the target. Progress monitoring depends on repeated measures taken the same way over time, compared with that goal line.^[2] Review after 6 to 8 data points. The four-point rule is one way to decide when to change instruction.

The row scores tell you what to change. If explanation sits at 1 while the other rows climb, the next teaching move is explanation, not another round of the whole paragraph.

Worked example: Jonah's body paragraphs

Jonah is a fictional sixth grader. His goal comes from the body paragraph template: given a claim, a short text at his reading level and 20 minutes, he will write a body paragraph scoring at least 10 of 12 on the body paragraph rubric above, on 3 of 4 consecutive scored paragraphs.

His three September baseline paragraphs score 3, 4 and 3, a median of 3. He states a point and copies a sentence from the text, then stops. Here is his data sheet through January, one probe every two weeks, with a co-scored sample marked:

The totals tell the team Jonah went from 3 to 11 in about 19 weeks. The rows tell them why the October scores stalled: explanation stayed at 1 for three probes. The teacher added sentence frames for explaining evidence ("This shows that... because...") in November, and the explanation row moved within two probes. Jonah has met the criterion on 3 of the last 4 paragraphs (10, 10, 11), so the team is ready to raise the target or move to multi-paragraph essays.

Both co-scored samples agreed within 1 point on the total and on every row. The progress report can state the scores with confidence.

Common mistakes and better versions

Weak: Using the district's 6-trait rubric for every goal.

Better: Use a short rubric whose rows match the goal. A trait the goal does not target adds noise to the total.

Weak: Levels labeled "beginning, developing, proficient, advanced" with no descriptors.

Better: One sentence per score per row that says what the writing contains.

Weak: Scoring classwork that had teacher conferencing, peer editing and unlimited time, then comparing it with a timed baseline.

Better: Score probes written under the goal's conditions. Use classwork as extra evidence, not as the data point.

Weak: Recording only the total.

Better: Record every row. Row scores show which part to teach next.

Weak: Changing the rubric in January because a better one turned up.

Better: Keep the rubric for the whole goal period. If it must change, re-score the earlier samples with the new rubric so the graph stays comparable.

Weak: One adult scores every sample, forever.

Better: Co-score every fourth or fifth sample and talk through disagreements.

Rubric goals for self-assessment and conventions

Two template goals put the rubric in the student's hands or use it for mechanics:

• Score own draft on the class rubric: the student scores each row, writes one sentence of evidence per score, and names one row to raise in revision. This builds the habit of checking writing against criteria. The What Works Clearinghouse secondary writing guide recommends using assessments of student writing to inform instruction and feedback, and student self-assessment is one form of that.^[3]
• Score conventions on monthly writing samples: a 16-point conventions rubric for capitals, end marks, commas and grammar in a monthly on-demand sample. More mechanics goals are in IEP goals for capitalization and punctuation.

Templates, data sheets and tools

• Rubric scoring explains the measure and when to choose it over counts.
• The rubric scoring data sheet has columns for row scores, totals and notes.
• Every rubric-scored writing goal in the written expression goal bank names its rubric size, so you can copy the structure.
• The IEP must say how progress on each goal will be measured.^[1] Name the rubric and its point total in the goal, and attach the rubric. Then run the goal through the goal checker.

Frequently asked questions

Use it in IEP Atlas

• Rubric scoring measurement guide
• Rubric scoring data sheet
• Rubric in the IEP glossary
• Written expression goals (writing)
• Written expression in the IEP glossary
• IEP goal checker

Related guides

• IEP goals for writing: written expression, spelling and mechanics
• How to measure writing: total words written and correct word sequences
• Paragraph writing IEP goals: 20 examples by sub-skill
• IEP goals for narrative writing: 20 examples from labels to dialogue
• Decision rules: the four-point rule and trend analysis
• IEP goals for writing simple sentences

Sources

1. 34 CFR §300.320: Definition of individualized education program, Legal Information Institute, Cornell Law School
2. Progress Monitoring: Data & Goal Setting, National Center on Intensive Intervention
3. Teaching Secondary Students to Write Effectively, What Works Clearinghouse, Institute of Education Sciences

Written and checked by the IEP Atlas editorial team under our editorial policy. This guide describes general practice and federal rules. It is not legal advice; your state may add requirements, and the IEP team makes the decisions for each student.