Listening to Children Read Aloud with a Trained Ear

You, too, can hear and interpret oral reading behaviors

Children who struggle with print—who fight it or fear it or ignore it or avoid it—usually are caught in the quicksand of decoding print into pronunciations, either from lack of experience or instruction or individual attention. The act of assessing oral reading behaviors can stretch to look at expressiveness, intonation, projection, clarity of articulation—pretty much whatever you want to examine.

In the field of reading education, particularly in the domain of experts working directly with struggling readers, such acts can be informal or formal, but they focus on oral behaviors in response to print..

When children read aloud a text, they occasionally leave a word out, put a word in, change a word, or repeat a word. Sometimes they reread and self-correct.

When you listen to a child read aloud and note the differences between what is printed and what is pronounced, the record serves as the basis of an oral fluency assessment.

For example, you can circle omissions, caret in additions, underline repetitions, note the substitution. So much depends upon whether you chalk these reading behaviors up as ‘errors’ or ‘miscues.’ ¹

Error is the preferred term if accuracy of reading the word “affection” is prized: “She looked upon her friend with infection,” read the child. Miscue is the preferred term if sensibleness of reading is prized: “She looked upon her friend with happiness.” Each deviation from”affection” changes the text; an error is an error whether or not it makes sense. A miscue is a miscue that may or may not make sense.

An error is an error whether or not it makes sense. A miscue is a miscue that may or may not make sense.

My wife, who taught third and fourth grade for many years, was trained to record ‘errors’ according to a protocol in place when Reading Recovery, an early intervention program for slow starters in reading, came online in the 1990s. She used the ‘running records’ technique as developed by Marie Clay to assess the practical match between a learner and a text. Running records is a perfect name for this technique. Teachers can use it on the run, a kind of mental math you can do in the moment without pen and paper.

All deviations from text are counted as errors, but self-correction behaviors are noted with an SC, though still counted as an error. An error rate greater than 10% regardless of self-corrections suggests that the text and others of its linguistic ilk are difficult for this reader.

Self-corrections are prized in Reading Recovery as evidence of self-monitoring, considered the engine of development.

As I’ve said in earlier posts, my earliest experience in oral fluency assessment involved the Gray Oral Fluency Test² in a clinical setting. This assessment is appropriately called a ‘test’; for one thing, you need a stop watch. Errors are definitely errors and serve to match a reader to a grade level with a decimal point—very different from running records.

It consists of sixteen passages, graded for difficulty, with comprehension questions, a standardized coding system for recording errors, psychometric data supporting its reliability and validity, a scoring system juggling points for number of seconds to read a passage, number of errors, form of the test, gender of the reader, and responses to comprehension questions. This test system has been normed empirically and has withstood the test of time.

During my first year of university teaching, I taught a graduate course in assessment of reading using a technique called ‘Miscue Analysis’ as spelled out by Kenneth and Yetta Goodman.³

The Goodmans characterized reading as a ‘psycholinguistic guessing game,’ where readers use ‘cueing systems’ to guide their mouths to make the words they pronounce from print. The spelling of a word is one cueing system. Grammar is another. Semantics is a third.

The assessor looks for explanations—did the substitution make sense? Did it resemble the spelling of the textual word? did it fit the grammatical structure of the context? What was the nexus of thought and language at the point of miscue?

The existence of robust approaches to assessment of oral fluency is evidence of its importance as a window on mental behaviors otherwise inaccessible. The work I did with struggling readers in Kurt’s clinic in the late 1970s required ongoing record keeping of errors or miscues occurring during oral reading with targeted work on patterns that emerged from the data.

ETS offers a number of automated scoring services. I’m not sure I would want to plan instruction on a machine-based oral reading assessment, though this conclusion isn’t well supported with relevant experience.

One spring I spent several days at ETS headquarters on a team writing multiple-choice questions for a reading specialist credential exam. Another year I was there on a team giving feedback on the design of the National Board for Professional Teaching Standards⁴. Maybe there are limited uses. Perhaps AI technology may help with assessment and instruction of phonological processing at an atomic level.

1. Metatheoretical Differences between Running Records and Miscue Analysis Author(s): Sinéad Harmey and Bobbie Kabuto. Source: Research in the Teaching of English , AUGUST 2018, Vol. 53, No. 1 (AUGUST 2018),

2. Bryant. (2012). Gray Oral Reading Tests—Fifth Edition (GORT-5). Austin, TX: Pro-Ed

3. https://eric.ed.gov/?id=EJ1119775

4. https://www.nbpts.org/