Reading assessment types: screening, diagnostic, progress monitoring and outcome

What are the four types of reading assessment?

Each type answers a different question. Trouble starts when a test built for one question gets used for another. A screener can't tell you which vowel team to teach on Monday. A state test can't tell you in October whether your small group is on track.

Read the table left to right and the logic shows. Screening casts a wide net, and diagnostic work narrows it down to the skills a flagged student is missing. Progress monitoring then watches how the student responds to the plan you built from those skills. Outcome tests judge the whole year, for the student and for the program.

What does a reading screener do?

A screener is a short test given to every student to find the ones who may struggle later. It takes a few minutes per student. Good screeners predict later reading well, so a low score is a reason to look closer.

In kindergarten, screeners usually check letter sounds and phonemic awareness. From the middle of first grade on, most add oral reading of a short passage, scored as words correct per minute (WCPM). Older grades often add a silent comprehension task.

A screener gives you risk, not reasons. Two students can earn the same low score for very different causes: one can't decode the words at all, while the other decodes accurately but so slowly that the minute runs out. The next step, a diagnostic check, sorts them apart and tells you what each one needs.

When do you need a diagnostic reading assessment?

Give a diagnostic check when a screener flags a student or when progress stalls. The goal is a teaching target. You want to finish the check knowing which sounds, patterns or skills to teach next.

Most diagnostic work in reading is informal and quick. A phonics survey lists real and made-up words by pattern, from short vowels up to multisyllable words. A spelling inventory shows the same knowledge from the writing side. A phonemic awareness check asks the student to blend, segment and change sounds with no letters in sight.

• Phonics survey: which spelling patterns the student can decode, and where accuracy breaks down.
• Spelling inventory: which patterns the student can spell, scored by feature.
• Phonemic awareness tasks: blending, segmenting, deleting and swapping sounds.
• High-frequency word check: which common irregular words are automatic.
• Oral reading error analysis: whether errors are guesses from the first letter or real attempts.

Formal diagnostic testing for a disability is a different thing. That is a full evaluation done by a qualified team, with parent consent, under federal special education rules.

How is progress monitoring different from screening?

Progress monitoring repeats one short measure often, so you can see a trend. Screening happens three times a year and compares a student with a benchmark. Monitoring happens every week or two and compares the student with their own goal line.

The measures are usually curriculum-based measurement (CBM) probes. Each probe is brief, timed and the same kind of task at the same level. A grade 2 student might read a new one-minute passage every Tuesday. The scores go on a graph with a goal line drawn from baseline to the target.

The 2009 IES practice guide on response to intervention recommends monitoring Tier 2 students at least once a month. Many schools go further and monitor Tier 2 every week or two and Tier 3 every week. More data points give you a trend you can trust sooner, which means fewer weeks lost on a plan that isn't working.

What is an outcome assessment for?

Outcome tests look back. They tell you whether a student reached grade-level standards and whether a program worked for a group. State reading tests and end-of-year district tests are the common examples.

They are too late and too broad to guide daily teaching. Results often arrive in summer. Use them for planning the next year, checking whether intervention groups closed gaps, and spotting patterns across classrooms.

How do the four types fit into one school year?

Screen at least two weeks after school starts so students have shaken off the summer. Avoid giving a screener the day after a long break. Scores dip, and you'll flag students who don't need it.

Which reading assessment should you give first?

Start broad, then narrow. For a new student, or a student you are worried about, this order wastes the least time.

1. Give a one-minute oral reading passage at grade level. Record WCPM and accuracy.
2. If accuracy is below about 95%, the problem is likely decoding. Give a phonics survey.
3. If the phonics survey shows gaps in short vowels or blends, check phonemic awareness too.
4. If accuracy is high but WCPM is low, decoding is accurate but slow. Plan fluency work and keep an eye on phonics patterns in longer words.
5. If both are fine but comprehension is weak, look at vocabulary and background knowledge.

Compare the WCPM score with the published Hasbrouck and Tindal oral reading fluency norms for the grade and season. A score more than about 10 words below the 50th percentile is worth a closer look.

Why do teachers mix up assessment types?

Because many tests can serve more than one purpose, and vendors bundle them. The same one-minute passage can be a screener in September and a progress monitoring probe in October. That's fine. What changes is who takes it, how often, and what decision follows.

A good habit is to write the decision next to each test on your calendar. "Fall screener: decide who gets a diagnostic check." "Weekly probe: decide whether to change the plan." If you can't name the decision, question whether the test is worth the minutes.

What can parents and tutors use outside school?

A tutor or homeschool parent can follow the same order on a small scale. Check letter sounds first, then give a short phonics survey to find the first pattern that breaks down. Time a grade-level passage and compare the score with published norms. Then pick one probe to repeat every week or two, and keep the scores in one place.

Keep results as teaching information, not grades. If you are worried about a possible learning disability, ask the school for an evaluation in writing. Only a qualified evaluator can make that call.

If you want the graphing done for you, Literacy Trail's progress monitoring tools plot weekly probes against a goal line. A paper graph works just as well if you keep it up.

How accurate are reading screeners?

No screener is perfect, and every one makes two kinds of mistakes. Some students are flagged who would have been fine without help, and some students pass who later struggle. Test makers usually set cut scores to catch more of the second group, because missing a struggling reader costs far more than giving a few extra students a closer look.

That trade-off is why a screening result should always lead to a second step rather than straight to placement. A quick diagnostic check, a look at classroom work and a word with the teacher will sort out most false alarms within a week. Students who pass but still worry their teacher deserve the same second look.

What do the numbers on a score report mean?

Score reports mix several kinds of numbers, and families often confuse them. A raw score is what the student did, such as 64 words correct per minute. A percentile compares that score with a national sample of students in the same grade and season. A benchmark is a target score tied to later success on an outcome test.

A percentile is not a percent correct. A student at the 30th percentile may have read 94 percent of the words correctly. When you explain results to families, give the raw score, say what it means in plain words, and then say what you plan to do about it.

What should you do when two tests disagree?

It happens often, and it usually means the tests measure different things. A student can pass a silent comprehension screener by using strong vocabulary and background knowledge while reading words slowly and inaccurately aloud. Trust the oral reading score for decoding and the comprehension score for understanding.

When results conflict, look at the task each test used and give a short diagnostic check to settle the question. Don't average the two scores. Averages hide exactly the information that tells you what to teach.

How much time should assessment take?

Less than most schools spend. A screener takes a few minutes per student three times a year, and a weekly probe takes one to three minutes. Diagnostic checks take ten to twenty minutes, but only for the students who need them. If assessment is crowding out teaching, the cause is usually extra tests that nobody uses to make a decision.