Star Assessments: A Teacher's Guide to Scores and Reports

Star Assessments are short computer-adaptive tests from Renaissance that tell you where each learner is in reading or maths, and how fast they are moving. There are three: Star Reading, Star Maths and Star Early Literacy. Over 7,000 schools in the UK and Ireland use them, according to Renaissance's UK Star Assessments page.

This guide is for teachers holding a Star report and trying to decide what it means. It explains every score on the report, the default benchmark groups, how accurate the tests are, and what to do next. Star is one part of a school's reading picture, so it links to our guide to reading comprehension strategies. Every product fact comes from Renaissance's own pages, checked on 24 September 2026.

‍

What are Star Assessments?

Star Assessments are online tests a learner takes on a computer or tablet. Nobody marks them, because the report is generated automatically as soon as the learner finishes. Renaissance says a class can be screened in as little as 20 minutes. Each test stays short because it adapts to the individual child, so nobody spends time on questions that are far too easy or far too difficult.

• Star Early Literacy: 27 questions across 41 skill areas in ten domains, for children with a sight vocabulary of fewer than 100 words, from age 3, as set out on the Star Early Literacy page.
• Star Reading: 34 questions across 43 skill areas in four domains, for ages 6 to 16, according to the Star Reading page.
• Star Maths: 24 questions across 50 skill areas in ten domains, for ages 6 to 15, covering number, algebra, geometry and measure, and data, as the Star Maths page lists.

Schools use Star for three purposes: screening a whole class to identify who needs help first, checking progress at set points in the year, and planning teaching from the skills report, which Renaissance calls the Instructional Planning report. In the UK, Star often sits beside Accelerated Reader, the reading practice programme, which uses the Star Reading score to suggest book levels.

Renaissance also publishes its own study of how well Star predicts Key Stage 2 SATs. With results from over 12,000 children in over 500 schools, it reports correct predictions for 79% in reading, 89% in maths and 82% in spelling, punctuation and grammar. That is the company's own research, so the independent evidence further down matters.

‍

How does a Star test work?

A Star test is computer-adaptive. When a learner answers a question correctly, the next one is harder. When they get one wrong, the next one is easier. The test homes in on the level where the questions stop being easy and start being hard, and that level becomes the score.

This explains why two children in the same class see different questions, and why Renaissance can keep the assessment comparatively short. A child who answers hard algebra questions correctly is not then asked basic addition, because the system already knows they can do it.

Adaptive testing often confuses families and learners at first. One former learner described the design on X as progressively easier as you get more wrong, which is exactly the point. For a teacher, the practical lesson is simple: a Star score is only as good as the child's effort on the day. A child who clicks at random drags the test down to easy questions and scores low.

A successful testing session needs a quiet room, working devices, and a teacher who explains beforehand that the questions will become harder and easier. Tell the class that nobody is expected to get everything right, because the test is designed so that everyone eventually meets questions they cannot answer.

‍

What do Star scores mean?

A Star report carries several numbers that are easily confused, and the diagram shows how they relate to each other. Renaissance's help page on the Scaled Score says it is the most fundamental score and that all the other norm-referenced scores are derived from it.

‍

Scaled Score (SS)

The Scaled Score runs from 0 to 1400 and spans Year 1 to Year 13. It depends on how hard the questions were and how many the learner got right. Because every year group sits on the same continuous scale, it is the appropriate score for tracking one child's progress over time.

On Renaissance's Unified Scale, Star Reading and Star Maths scores run from 600 to 1400 and Star Early Literacy from 200 to 1100. The same number does not mean the same thing on different tests. A score of 900 on Star Early Literacy is not the same as 900 on Star Reading.

One threshold is worth knowing. On Star Reading, a Scaled Score below 615 triggers a Probable Non-Reader alert for teachers whose email address is verified with Renaissance. Treat that alert as a prompt to check word reading and phonics immediately, before interpreting anything else on the report.

‍

Standardised score (NRSS)

The Normed Referenced Standardised Score compares a learner with others of exactly the same age, in years and months, in a national sample. The average is 100, and the range runs from 50 to 150. Above 100 is above average for age; below 100 is below it.

This is the score to use when age matters, for example in a class where the oldest child is almost a year older than the youngest.

‍

Percentile Rank (PR)

The Percentile Rank runs from 1 to 99. A rank of 60 means the learner scored higher than 60% of the national sample in the same year group. Star Reading uses separate norms for the UK, Wales, Scotland and Northern Ireland, and international norms elsewhere.

The report also gives a Percentile Rank Range. That is the band the child would probably occupy if retested soon, so a child close to a cut-off could fall on either side of it on a different day.

‍

Reading Age (RA)

The Reading Age gives the typical age at which readers obtain a Star Reading score. It is written as years and months: 7:10 means 7 years and 10 months. The range runs from 6:00 to 16:06, so a strong reader at the top of secondary can hit the ceiling.

Reading ages are easy to explain to families, but they stretch unevenly across childhood: twelve months behind means considerably more at seven than at fourteen. The standardised score allows for that difference; the reading age does not.

‍

Zone of Proximal Development (ZPD)

In Star Reading, the ZPD is a range of book levels for independent reading practice, not too hard and not too easy. The name borrows from Vygotsky, whose idea is explained in our guide to the zone of proximal development. Renaissance's ZPD help page calls the range approximate and asks teachers to adjust it for the child's interests and knowledge.

‍

Growth percentile (SGP)

The growth percentile, or SGP, compares a learner's growth with children in the same year who had a similar Star history. It runs from 1 to 99. Renaissance's report for each learner labels growth low (0 to 33), typical (34 to 66) and high (66 to 99). It needs at least two tests in different testing windows.

‍

US scores: Grade Equivalent and Lexile

American reports can also show a Grade Equivalent and a Lexile measure. The Grade Equivalent is the highest grade level at which a learner can read short passages, not the level they can read comfortably for long. A Lexile measure places reader and book on one scale, and the Lexile range for a reader runs from 100L below to 50L above their measure. Schools can choose Lexile ranges in place of ZPDs.

If you have a number and are not sure which kind it is, the mini app below reads it for you. Choose the score, type the number, and it tells you what the score says, what it does not say, and what to do next.

‍

What is a good Star score?

There is no single good score, because each measure answers a different question about the learner. The same Scaled Score can be strong for a younger child and weak for an older one. For a quick answer, look at the standardised score or the Percentile Rank, because both compare the child with others of the same age or year group.

Schools usually mean the benchmark group. The Star Screening report sorts a class into groups by Percentile Rank. Renaissance's UK default with four categories is:

• At or Above Benchmark: Percentile Rank 40 or above.
• On Watch: 16 to 39.
• Intervention: 6 to 15.
• Urgent Intervention: below 6.

These categories are configurable settings, not fixed rules. A school can choose two, three, four or five categories and move the lines. The US defaults are different: Renaissance's US benchmark guidance uses below 25 for Intervention and below 10 for Urgent Intervention. So before a meeting about who is behind, check which settings your school's report uses.

We have made a set of slides for talking this through with colleagues. They cover the scores, the benchmark groups and the questions to ask before acting on a result.

Use the slides to agree three things as a team: which score you report to governors, which benchmark settings you use, and what extra check happens before a child joins an intervention group.

‍

Reading a Star report: two worked examples

The two children below are invented, and so are their numbers. They illustrate the reasoning rather than a conversion table, because real reports take their numbers from Renaissance's technical manuals.

‍

Year 3 (US grade 2): a reader below benchmark

Maya is 7 years 10 months old. Her autumn Star Reading report shows a Percentile Rank of 12, a Reading Age of 6:08 and, from last year's tests, a growth percentile of 58. On the UK default, a rank of 12 puts her in Intervention. Her growth is typical, so she is moving at the usual rate from a low start.

Her teacher does not place her straight into a comprehension group. First she listens to Maya read a short book aloud, and notices that Maya reads slowly and sounds out common words. That evidence points to word reading and fluency rather than comprehension, so the next step is phonics catch-up and daily fluency practice. Our guide to reading fluency strategies covers what that looks like.

‍

Year 9 (US grade 8): a reader the score does not explain

Callum is 13 years 11 months old. His Percentile Rank is 45, At or Above Benchmark, but his growth percentile has dropped to 18 since last spring. His English teacher has also noticed that his written answers on the class novel have become noticeably thinner.

The Star score suggests Callum reads competently, while the growth figure and the written work suggest something has changed. His teacher asks him about the test, and he says he rushed it to get to lunch. She checks his understanding with two inference questions on a passage from the class text, as in our guide to inference and deduction, and he answers well. The next step is a conversation about effort, and a retest at the next window.

‍

How accurate are Star Assessments?

The fullest independent summary for UK schools is the EEF's entry for Star Reading in its measures database. It reports reliability of 0.94 to 0.97 from the technical manual, and correlations of 0.74 to 0.89 with the Suffolk Reading Scale. It also reports that Star Reading predicted which of 12,212 children would reach the expected standard in Key Stage 2 reading with 79% accuracy.

Two numbers in that entry matter more than the headline. In screening terms, sensitivity was 0.84, so Star caught 84% of the children who went on to miss the standard. Specificity was 0.70, so 30% of children who went on to meet it were flagged anyway. The EEF also notes the UK norming sample over-represented disadvantaged schools and England, compared with Scotland and Wales.

Independent studies, mostly American, tell a similar story. A 2026 meta-analysis of 30 studies found computer-adaptive reading screeners predicted later tests well, but sensitivity of 0.77 fell below the 0.80 usually wanted for screening (Shen et al., 2026). In a study of 13,009 learners in grades 3 to 8, a single Star Reading screen was accurate enough for reading, and cut-offs set on local data did better still (Forcht and Van Norman, 2024).

For young children the picture is weaker. Star Early Literacy predicted year-end word reading only moderately, and its cut scores classified the lowest scorers poorly. Short paper checks of letter sounds and word reading improved the prediction (Clemens et al., 2015). A grade 3 study found an oral reading fluency check gave the best accuracy for its cost, and Star plus that check picked out useful teaching groups (Barrett et al., 2024).

So Star is a reliable first sort of a class, but it is not accurate enough on its own to decide which child receives which intervention.

‍

How often should you test with Star?

Renaissance's growth percentile is built on three testing windows a year, and many schools test once in each: autumn, spring and summer, or fall, winter and spring in US schools. Testing more often does not make any individual result more accurate; it simply produces more results to interpret.

Some schools use Star between windows to check whether an intervention is working. The research says be patient. An analysis of Star Reading results from over six million learners found growth estimates needed at least five tests over about 20 weeks before they were stable (Bulut and Cormier, 2018). A later simulation found seven to nine monthly tests were needed for reasonably accurate decisions (Van Norman and Forcht, 2024).

For week-to-week checks, a short oral reading probe usually tells you more, sooner. Our guide to progress monitoring with curriculum-based measures explains how. Where a school runs tiered support, our guide to MTSS and response to intervention shows where screening fits.

‍

Using Star for SEND and intervention planning

Star can flag a child who needs a closer look, but it does not identify a special educational need. A low score can come from weak word reading, weak language, low effort, anxiety, a bad morning or unfamiliarity with the screen. The test cannot distinguish between these explanations, but the adult who watches the child usually can.

When Star flags a child, the useful sequence is short. Listen to them read aloud. Look at their recent written work, and ask them what the test felt like. Then decide whether to monitor, to adjust your teaching, or to consult the SENCO. Our overview of special educational needs sets out the wider process, and our guide to supporting learners with dyslexia covers what to look for in reading.

When Star informs support planning, keep it in its proper role. Use it to track broad attainment over terms, and use closer tools, such as precision teaching probes, to track the specific skills you are teaching.

‍

Talking to parents about Star results

Families see Star through the Star Parent Report or the Star Family Report, and often through a single number. Families posting on X mostly mention the percentile. One parent posted that reading and comprehension had been a big focus with their daughter ahead of the test.

Three sentences cover most conversations: Star is a short test that adapts to your child and shows where they are now; we interpret it alongside what we observe in class every day; and here is what we are doing next, with how you can help at home. Avoid drilling for the test: practice questions do not build reading, and reading widely does.

‍

What teachers and parents say about Star

We searched posts on X from late July to late September 2026 for how teachers and families talk about Star. The results were sparse, and that absence is informative in itself. No UK teacher or parent discussed a Star score in public. American posts were about the testing week, the devices and the percentile.

The strongest objection was practical, not about the scores. A US maths teacher said the district's Star maths test only takes 35ish minutes to complete, then described two days lost to broken laptops and chargers. Plan the testing session as carefully as you interpret the results, with charged devices, working logins and a quiet room.

‍

Limitations of Star Assessments

Star is a carefully constructed assessment with strong reliability, but it has limitations every teacher should understand before interpreting a report.

• It is a screen. Independent studies find it misses some children who need help, and flags some who do not.
• It is silent reading on a screen. Star Reading cannot hear fluency, expression or decoding errors. Listening to a child read fills that gap.
• One sitting is one day. Effort, anxiety and device problems all move a score. Growth needs several tests before it means much.
• The norms have a shape. The EEF notes the UK sample over-represented disadvantaged schools and England.
• Much of the evidence is the publisher's own. The independent studies are mostly from the US, with US tests as the outcome.

None of this makes Star a weak assessment; it makes it a starting point. Its best use is to identify quickly which learners to examine first, then to rely on your own assessment to decide what happens next.

Your next step: take your class's latest Star Screening report and pick the three learners nearest a benchmark line. Listen to each read aloud for two minutes this week, and write one sentence on what you heard. Then compare it with the report. For the reading strategies that follow, see our guide to teaching reading comprehension.

‍

Frequently asked questions

‍

How long does a Star Reading test take?

Renaissance says a class can be screened in as little as 20 minutes, and the EEF lists 15 to 30 minutes for Star Reading. Star Reading has 34 questions, Star Maths 24 and Star Early Literacy 27. Allow extra time for logging in.

‍

How do learners log in to Star?

Go to your school's Renaissance web address, select I'm a Student, enter the user name and password, and select Log In; staff select I'm a Teacher/Administrator instead, as the Renaissance logging-in guide explains. If you do not know the address, schools.renaissance.com opens a page headed Find your School or District.

‍

What is the highest Star Reading score?

The Scaled Score runs to 1400, the Reading Age reaches its maximum at 16:06, and the Percentile Rank stops at 99.

‍

Can learners practise for Star?

They can become familiar with the screen and the question types, but practice tests will not improve their reading. Because the test adapts to every answer, the most reliable route to a higher score is genuinely better reading.

‍

Is Star Reading the same as Accelerated Reader?

No. Star Reading is the assessment, while Accelerated Reader is a separate reading practice programme that uses the Star Reading score to recommend a range of book levels. A school can use either programme independently.

‍

References

• Barrett, C. A., Johnson, E., Truckenmiller, A. J. and VanDerHeyden, A. (2024). Comparing the cost–accuracy ratios of multiple approaches to reading screening in elementary schools. Remedial and Special Education, 45(2), 71 to 84.
• Bulut, O. and Cormier, D. C. (2018). Validity evidence for progress monitoring with Star Reading: slope estimates, administration frequency, and number of data points. Frontiers in Education, 3, 68.
• Clemens, N. H., Hagan-Burke, S., Luo, W., Cerda, C., Blakely, A., Frosch, J., Gamez-Patience, B. and Jones, M. (2015). The predictive validity of a computer-adaptive assessment of kindergarten and first-grade reading skills. School Psychology Review, 44(1), 76 to 97.
• Education Endowment Foundation (2026). Renaissance Star Reading. EEF measures database. Checked 24 September 2026.
• Forcht, E. R. and Van Norman, E. R. (2024). Comparison of screening methods for computer adaptive tests to predict reading and math performance. Psychology in the Schools, 61(4), 1590 to 1610.
• Ochs, S., Keller-Margulis, M. A., Santi, K. L. and Jones, J. H. (2020). Long-term validity and diagnostic accuracy of a reading computer-adaptive test. Assessment for Effective Intervention, 45(3), 210 to 225.
• Renaissance Learning (2026). Scaled Score (SS), Normed Referenced Standardised Score (NRSS), Percentile Rank (PR) and Percentile Rank Range and Reading Age (RA). Star Assessments knowledge base, UK. Checked 24 September 2026.
• Renaissance Learning (2026). Default Percentile Rank (PR) values for school benchmark categories. Star Assessments knowledge base, UK. Checked 24 September 2026.
• Renaissance Learning (2026). Star Assessments. Renaissance UK. Checked 24 September 2026.
• Shen, L.-L., Huang, Y., Ma, Z., Chen, M., Ma, C. and Clemens, N. (2026). How accurate are computer-adaptive reading screeners? A systematic review and meta-analysis of validity and classification accuracy. School Psychology Review, 1 to 21.
• Turner, M. I., Van Norman, E. R. and Hojnoski, R. L. (2022). An independent evaluation of the diagnostic accuracy of a computer adaptive test to predict proficiency on an end of year high-stakes assessment. Journal of Psychoeducational Assessment, 40(7), 911 to 916.
• Van Norman, E. R. and Forcht, E. R. (2024). Accuracy of progress monitoring decision rules to evaluate response to instruction with two computer adaptive tests. Journal of School Psychology, 105, 101319.