Maze and cloze probes for progress monitoring

By the IEP Atlas editorial teamLast updated November 1, 2026

Short answer

A maze probe is a short timed passage where every seventh word or so is replaced by three choices, and the student circles the word that fits. Count correct choices in the time limit, usually 2.5 to 3 minutes. Give one probe a week at a fixed level and graph it. Cloze leaves a blank instead, which is harder to score the same way twice.

A maze probe measures whether a student is making sense of text while reading silently. The student reads a passage at a fixed level and, at regular gaps, picks the one word out of three that fits the sentence. The score is the number of correct choices in a set time. Because it takes about 3 minutes and can be given to a whole group, maze is one of the few comprehension measures you can repeat every week.

This page follows one fictional student through a year of maze data, then lays out the rules in a table. For the wider question of how to measure comprehension goals, see how to measure reading comprehension for an IEP. The pillar on reading comprehension strategies covers what to teach once the data says a change is needed.

Maze vs cloze

Both formats delete words from a passage. They differ in what the student does at each gap.

Feature	Maze	Cloze
What the student sees at a gap	Three choices: the correct word and two distractors	A blank line
Response	Circle or click one word	Write or say a word
Scoring	Right or wrong, with a key	Exact word or an accepted synonym, which takes judgment
Timing	Timed, often 2.5 to 3 minutes	Often untimed
Group administration	Easy, on paper or computer	Possible, but slower to score
Best use	Weekly progress monitoring	Instruction, vocabulary in context, informal checks

Cloze is useful in teaching. It makes students think about meaning and syntax, and it pairs well with context clues activities. For a weekly graph, maze wins. Two adults scoring the same cloze sheet often disagree about synonyms, and that disagreement shows up as noise on the graph.

One student across a school year

Dev is a fictional 5th grader. His oral reading fluency is close to grade level, but he answers few questions correctly after reading. The team suspects he reads words without building meaning. His reading goal will be monitored with a grade 5 maze probe.

Baseline. The teacher gives three grade 5 maze probes, one on each of three days in early September. Each probe lasts 3 minutes.

Date	Correct choices	Errors
Sept 8	8	6
Sept 10	11	4
Sept 12	9	7

The median is 9 correct. That number goes into present levels with the date, the passage level and the time limit. The baseline data guide explains why the median beats the average for three probes.

The goal. The team uses the growth guidance that came with the district's maze tool to choose an ambitious weekly rate. For this illustration, it is 0.4 more correct choices per week. Across 32 weeks of instruction, that adds about 13, so the target is 22.

The error limit matters. A student who races and guesses can raise the correct count while the error count climbs too. Writing both into the goal stops that.

Weeks 1 to 8. Dev gets a 30-minute intervention 4 days a week that targets the gist of short sections and monitoring. His weekly scores are 9, 10, 9, 12, 11, 11, 12 and 13. The aimline from 9 to 22 expects about 12 by week 8. He is on track.

Weeks 9 to 14. Scores flatten: 12, 11, 12, 12, 11, 12. Four points in a row fall below the aimline, so the team applies its decision rule (see the four-point rule). It adds a short daily summary routine and checks that the intervention is delivered as planned.

Weeks 15 to 24. Scores climb again, from 13 to 18. The rate of improvement for this stretch is about 0.5 correct per week, slightly above the goal rate. The progress report says so in plain numbers.

How to give a maze probe

Pick the tool. Use passages from one published set at one level for the whole year. The National Center on Intensive Intervention rates progress monitoring tools, including maze measures, on its tools chart.^[1]
Read the directions word for word. Do one practice sentence together.
Time it exactly. Start the timer when the student starts reading. Stop at the publisher's limit.
Score with the key. Count correct choices. Count errors separately. Follow the publisher's rule for guessing, such as stopping the count after 3 errors in a row or subtracting part of the errors.
Record and graph the same day. Enter both numbers. A week-old stack of unscored probes is how teams miss a flat line.

Maze can be given to a small group in the same 3 minutes. Students who need a read-aloud accommodation on class tests usually take maze without it, because hearing the passage changes what is measured. Note any change to standard administration on the data sheet.

What maze tells you and what it does not

Maze shows whether the student is building meaning sentence by sentence. It moves with fluency and with sentence-level comprehension. It says little about retell, inference across paragraphs, main idea or theme.

For that reason, a maze goal should not be the only comprehension goal when the student's need is higher-level. Pair it with a skill goal, such as a retell goal scored on a rubric, or an inferencing goal. Maze is the weekly general outcome measure. The skill goal shows whether the taught skill is landing. The difference between the two kinds of measure is covered in curriculum-based measurement vs mastery measurement.

Rules and options at a glance

Some of these rules come from federal law. Others are set by your state or district, and they vary.

Question	What applies	Varies by state?
Must the IEP say how progress is measured and when it is reported?	Yes. Each annual goal needs a description of how progress will be measured and when reports go to parents.^[3]	No, but report schedules vary by district
Which maze tool is used?	Usually chosen by the district	Yes
How often to probe	Weekly is common for intensive intervention; repeated measures over time are the core of progress monitoring^[2]	Some states set minimums for intervention tiers
Can maze data count toward a learning disability evaluation?	Federal rules let states use a process based on the student's response to research-based intervention^[4]	Yes, states set their own criteria
Must parents get the progress data?	When intervention data is used in a learning disability evaluation, repeated assessments at reasonable intervals must be documented and given to parents^[5]	Timing and format vary
Is a universal comprehension screener required?	Not by IDEA	Yes, some state reading or dyslexia laws require screening

Check your state's eligibility guidance before using maze data in an evaluation. This page describes the rules in general and is not legal advice.

Writing maze into the IEP

Put three things in present levels: the median baseline, the passage level and the time limit. Then write the goal with the same three. A goal that says "will improve maze scores" has no level, no time and no target.

For full goal wording, the comprehension goals page lists 100 templates by grade, and the reading comprehension strategies hub groups the strategy goals. For other weekly reading measures, see reading progress monitoring tools. The glossary entries for probe and curriculum-based measurement define the terms for families.

Frequently asked questions

How long is a maze probe?

Most published maze probes last 2.5 or 3 minutes. Use the time your tool's directions give and keep it the same all year, because a longer probe gives a higher count and breaks the comparison with earlier scores.

Should I count errors on a maze probe?

Yes. Record correct choices and errors separately. A student who guesses fast can raise the correct count while errors rise too. Many tools adjust for guessing, and writing an error limit into the goal keeps the score honest.

Can a student with a read-aloud accommodation take a maze probe?

Maze is usually given without a read-aloud, because hearing the passage changes the task from reading to listening. If the team decides otherwise, give it the same way every time and note it on the data sheet and in the goal.

Is maze a good measure for high school students?

It can be, if your tool has passages at the right level. For older students whose need is analysis or inference across a text, pair maze with a skill goal scored on a rubric so the graph reflects the skill you teach.

Use it in IEP Atlas

Sources

Intervention & Assessment Tools Charts, National Center on Intensive Intervention
Progress Monitoring: Data & Goal Setting, National Center on Intensive Intervention
34 CFR §300.320: Definition of individualized education program, Legal Information Institute, Cornell Law School
34 CFR §300.307: Specific learning disabilities, U.S. Department of Education (IDEA website)
34 CFR §300.309: Determining the existence of a specific learning disability, U.S. Department of Education (IDEA website)

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.