What is the CORE Phonics Survey? The informal decoding inventory explained

The CORE Phonics Survey is an informal, criterion-referenced inventory of phonics and decoding skills that a teacher gives to one student at a time, in roughly 10 to 15 minutes, to find out which letter names, letter sounds, and spelling patterns that child can actually read. It is published by CORE, the Consortium on Reaching Excellence in Education, inside the book Assessing Reading: Multiple Measures, printed by Arena Press, a division of Academic Therapy Publications. Classroom teachers, reading specialists, and instructional coaches administer and score it themselves. The survey is not an IQ test, and it is also not a standardized norm-referenced test of any kind: it produces no standard scores, no percentile ranks, and no national norms, and a result is a count of items read correctly judged against an expectation of mastery. In a school's assessment system it works as a diagnostic follow-up to a screener. This article covers what the survey contains, who publishes it, how teachers use it, and what research has found.

The survey has two sections, one for alphabet skills and letter sounds and one for reading and decoding. CORE's teacher resource says it "assesses the phonics and phonics-related skills that have a high rate of application in beginning reading." On the current record form, which CORE distributes free as an interactive PDF, the parts run as follows.

• Parts A and B, letter names: The student reads a matrix of all 26 uppercase letters, then all 26 lowercase letters, scored out of 26 each.

• Parts C and D, letter sounds: Twenty-one consonant sounds, then the five long vowel sounds and the five short vowel sounds, scored separately.

• Parts E through K, single-syllable decoding: Seven parts of 15 words each, running from short vowels in consonant-vowel-consonant words through consonant blends, digraphs, r-controlled vowels, long vowel spellings, variant vowels, and low-frequency spellings.

• Part L, multisyllabic words: Twenty-four longer words such as consent, admire, and railways, testing whether a student can apply syllable patterns to unfamiliar words.

The design feature that matters most is the use of "pseudowords," made-up words such as nop, dilt, and gorf. Each of Parts E through K holds two lines of real words and one line of pseudowords, because a pseudoword must be sounded out; CORE's guidance notes that a student "cannot have memorized" it. A child who handles real words yet stumbles on pseudowords is probably leaning on sight memory rather than on the code. A parallel CORE Spanish Phonics Survey covers early Spanish literacy instruction.

CORE was founded in 1995 by the author Linda Diamond and Bill Honig, a former California Superintendent of Public Instruction, in response to what the organization calls California's crisis in reading achievement. It became a subsidiary of the nonprofit Pivot Learning in 2017 and merged with UnboundEd in 2022. The name causes confusion: the organization was originally the Consortium on Reading Excellence, which is why the research literature calls the instrument the "CORE-PS." Its current legal name, on the copyright page of the book, is Consortium on Reaching Excellence in Education, Inc.

The survey is not sold separately. It sits inside Assessing Reading: Multiple Measures, edited by Linda Diamond and B.J. Thorsnes and now in its Revised Second Edition, which CORE sells for $59. The book collects formal and informal reading assessments for kindergarten through grade 12 and groups the phonics surveys under decoding and word recognition, alongside phonological awareness tests, a high-frequency word survey, and an oral reading fluency measure.

The instrument has changed across editions. According to a study in The Elementary School Journal, the 1999 first edition included a subtest that assessed phonics through spelling, and the 2008 second edition dropped it. Teachers working from older photocopies online may be using ten-item subtests and a spelling section the current edition no longer contains.

Administration is deliberately low-tech. The directions sit on the record form itself, the student reads from material printed on the pages that follow, and the teacher covers the parts not being given with a sheet of paper. The form reproduces the same letters and words in reduced size, so errors are marked as the student reads aloud. CORE's teacher resource gives the grade range as kindergarten through grade 8 and the grouping as individual. If a student cannot read two or more real words in a line, the examiner skips that part's pseudoword line, and testing stops early if the child is tiring.

Scores are transferred to a Skills Summary on the front page, which lists every part with its point total and leaves space for two notes, "skills to review" and "skills to teach." The survey exists to generate a teaching list rather than a label. CORE's guidance calls it a mastery test on which students are expected to get every item correct eventually, advises retesting every four to six weeks on parts not yet mastered, and treats two or more errors in a five-word line as a signal that the element needs direct instruction.

Student profile forms in the same book schedule the parts by grade and season, so kindergarten begins with letter names in the fall and adds consonant and vowel sounds by spring. Teachers often reach for the survey after a fluency screener flags a student. When a child's oral reading rate on a measure such as DIBELS falls below benchmark, the survey shows whether the underlying code knowledge is missing.

This distinction is the one most worth getting right. A "norm-referenced" test compares a student with a representative national sample of same-age peers and reports that comparison as a standard score or percentile rank. A "criterion-referenced" test compares performance against a defined standard of skill and reports whether the skill is mastered. CORE states that interpretation in the book "is based on skill mastery and comparison of performance among students in the classroom," and researchers who examined the instrument describe it as a criterion-referenced mastery measure with no basal or ceiling rules.

There is no norm sample and no way to convert a raw count into a percentile, so 12 out of 15 on r-controlled vowels means the child read 12 of those words correctly and nothing more. There is no publisher technical manual either. When the research team behind the 2014 Elementary School Journal study telephoned CORE and the publisher, both confirmed they had no reliability or validity data for the instrument, so the evidence available today came from independent researchers.

None of that makes the survey a poor tool. Low cost, short administration, and close alignment with what teachers actually teach explain why it spread through Reading First projects, district training, and teacher preparation programs. It should not drive special education eligibility decisions on its own, and it does not screen for or detect dyslexia. Identifying a reading disability calls for a full evaluation by qualified professionals, usually including norm-referenced measures such as the CTOPP-2.

Two independent studies give the survey more of an evidence base than most informal inventories have. The larger, published in The Elementary School Journal in 2014 by Reutzel, Brandt, Fawson, and Jones, drew on 592 students in kindergarten through grade 3. They reported a two-week test-retest correlation of .98 for total scores, a Cronbach's alpha of .94 for the reading and decoding section, and a Cohen's kappa of .85 between trained raters. The content matched the Common Core reading foundational skills for phoneme-grapheme correspondence 97% of the time, and total scores correlated .84 with students' words correct per minute on DIBELS measures.

The same study is candid about weak spots. The alphabet section, Parts A through D, had an alpha of only .64 and correlated just .42 with fluency, largely because most students past mid-kindergarten already know their letters, which compresses the range of scores. The authors advise interpreting that section cautiously and note that their schools were a convenience sample.

A second study by Park, Benedict, and Brownell, published in Exceptionality in 2014, examined 165 upper elementary students with specific learning disabilities. A two-factor model of alphabetic knowledge and decoding skills fit best, and both factors significantly predicted the students' decoding and fluency a year later on norm-referenced tests.

CORE, the Consortium on Reaching Excellence in Education, based in Oakland, California. It appears in CORE's book Assessing Reading: Multiple Measures, printed by Arena Press, a division of Academic Therapy Publications.

No. It is criterion-referenced and informal. It reports raw counts of correct items against mastery expectations, with no national norm sample, no standard scores, and no percentile ranks.

CORE's guidance gives 10 to 15 minutes per student, administered individually. Classroom teachers, reading specialists, and trained coaches give and score it themselves.

Pseudowords such as nop and gorf cannot have been memorized, so reading them requires applying letter-sound knowledge. Comparing real words with pseudowords shows whether a student is decoding or relying on sight recognition.

No. It is a skills inventory showing which phonics elements a student has mastered. Diagnosing a reading disability requires a comprehensive evaluation by qualified professionals, and survey results are one input such an evaluation can use.

A phonics survey and a cognitive ability test sit at opposite ends of the assessment spectrum. The CORE Phonics Survey samples a narrow, directly taught skill and reports mastery of specific letter-sound patterns. An intelligence test samples reasoning with unfamiliar material, so the score reflects general cognitive ability rather than whether a scope and sequence has been covered in class. The two are related, since children who reason well tend to acquire the code faster. Parents who want an overview of how children's reasoning is measured can read our guide to IQ tests for kids.

For adults who want a measured estimate of their own reasoning ability, the Reasoning and Intelligence Online Test, the RIOT IQ test, was developed by RIOT IQ with psychometrician Dr. Russell T. Warne for people aged 18 and over. It includes 15 subtests across six cognitive indices (verbal reasoning, fluid reasoning, spatial ability, working memory, processing speed, and reaction time), takes about 52 minutes, and reports scores on the familiar mean-100, standard-deviation-15 scale. It does not replace an individually administered diagnostic evaluation. What it offers is a normed estimate of general reasoning for adults who want one. You can learn more or take the test at RIOT IQ.

1. Consortium on Reaching Excellence in Education. (n.d.). CORE's Assessing Reading: Multiple Measures, Revised 2nd Ed.. corelearn.com

2. Diamond, L., & Thorsnes, B. J. (Eds.). (n.d.). Assessing reading: Multiple measures, revised second edition [Sample pages]. Arena Press. corelearn.com

3. Consortium on Reaching Excellence in Education. (n.d.). Data entry forms: Assessing Reading: Multiple Measures, 2nd edition. corelearn.com

4. Consortium on Reading Excellence. (2002). CORE Phonics Survey [Teacher resource adapted from the CORE Teaching Reading Sourcebook]. Scholastic Red. broward.k12.fl.us

5. Consortium on Reaching Excellence in Education. (n.d.). About CORE. corelearn.com

6. Reutzel, D. R., Brandt, L., Fawson, P. C., & Jones, C. D. (2014). Exploration of the Consortium on Reading Excellence Phonics Survey: An instrument for assessing primary-grade students' phonics knowledge. The Elementary School Journal, 115(1), 49-72. doi.org

7. Park, Y., Benedict, A. E., & Brownell, M. T. (2014). Construct and predictive validity of the CORE Phonics Survey: A diagnostic assessment for students with specific learning disabilities. Exceptionality, 22(1), 33-50. doi.org

Hero photo: A classroom alphabet strip above a blackboard. Photo by Daniellemichel, via Wikimedia Commons, CC BY-SA 4.0 (cropped).