This is an excellent approach to meaning-making construction in authentic oral reading. Since behavior is relative to one's actual proficiency, miscues act as a window into the proficiency of the reader under scrutiny.
There are a number of ways that oral miscues in reading can be assessed. Here we will look at the concept of the "Miscue" itself and then describe a favored analysis of miscues designed within the behavioral sampling technology – The Reading Miscue Inventory. This is based on the work of Ken Goodman – who did some of the earliest work on miscues.
Miscue analysis overall can be employed assist professionals in gaining
insight into the reading process. It involves both a quantitative
and
a qualitative component. Miscue Analysis has other purposes:
1. To analyze the oral reading of individual students
to gain insight into the
linguistic knowledge and strategy
use of readers while reading and "meaning
making".
2. To help professionals evaluate reading material.
It provides an objective
basis for determining whether
a given selection should be used in a reading
program and for determining material's
suitability for use by students.
The focus is on the unexpected responses or miscues that are produced by the readers under scrutiny. These are not considered errors or mistakes........they are considered non-random indices of the individual's underlying reading ability, linguistic knowledge, background knowledge, and reading strategies. The interest is both on the miscues and how the miscues change, disrupt, or enhance the meaning of a written text.
What is a miscue?
A miscue is defined as an observed
response that does not match what
the person listening to the
reading expects to hear (Goodman).
Overall Description of Miscue Analyses
The study of readers' miscues provides insights into how they integrate the language cueing systems during the reading process in order to construct meaning. This "realtime" and "on-line" meaning construction is referred to as "ongoing comprehending". Problems with meaning construction can be noted by collecting and analyzing the miscues which may serve as clues to students' misinterpretations after reading.
The miscue analysis enables you to comment on the LANGUAGE CUEING SYSTEMS that the reader controls and the READING STRATEGIES that he/she employs.
The quality of one's reading typically depends upon their ability to use the four major cueing systems in an integrated fashion.
Language Cueing systems of interest:
-- The Graphophonic System
** Relationship between sounds and
the written forms of
the language.
** Typically this (as all of the cueing systems)
is just sampled
while reading.
** Includes spelling, phonology and punctuation.
-- The Syntactic System
** Focuses on the grammar of the system.
** Much predictability in English based on
the NP VP
arrangement.
** Syntax may be thought of as occurring
at higher levels
of text organization than just the sentence.
** The genres may provide differences in
syntactic
organization
-- The Semantic System
** System of meanings in a language
** Strongly related to cultural knowledge
** This system is the "heart" of reading
since it is
actually a MEANING MAKING process.
** Semantic knowledge provides the reader
with the
pragmatic maps they can use to create new maps
or new information based on the text so the information
Can be mentally organized so they can remember it.
** If a reader's semantic knowledge in an
area is weak, it
Can interfere with their "on-going comprehending".
-- The Pragmatic System
** The linking of the other three systems with the CONTEXT.
ALL FOUR SYSTEMS INTERACT TO CREATE THE PROCESS OF READING
Reading Strategies of Interest:
--
Initiating Strategies
** Activated as soon as the readers recognize that there is
something they want to read.
** Involves the scripts for the activity of initiation and
determining what will be expected overall.
--
Sampling Strategies
** In perception or reading (or anything with meaning)
we don't focus on all aspects of the signals or text we
only select or focus on enough to make meaning.
** This increases efficiency and speed.
--
Predicting Strategies
** On the basis of the information that an individual has,
he/she starts to predict.
** The knowledge used to predict is based on experience
with the four cueing systems and the knowledge of the
world (which really is part of the pragmatic cueing system)
--
Confirming Strategies
** The active process of construction requires a constant
"testing" of the meaning-making as it occurs.....the
predictions are confirmed or disconfirmed as one goes
along in language (whether reading, speaking, listening,
or writing)
** These strategies will often result in self-corrections or
attempts at self-correction.
When Analyzing an Individual's Miscues, here are some relevant issues and questions (After Rhodes, 1993)
– You should focus
on every miscue and ask a series of questions that
address the following:
** How are the four cueing systems employed
** Are any corrections of miscues made
** Are the miscues of "high quality" or "low
quality"
** Are any of the other (minor) cueing systems
employed
** How do the miscues affect comprehension
of the text
* These questions are detailed to enable you to know what kinds
of
questions or reflections to use on each of these issues with each
miscue. Obviously, with experience you learn to key into certain
aspects – but the whole generic group is detailed here for your
information.
– With Regard to
the SEMANTIC CUEING SYSTEM
* Does the reader indicate through the miscues that he/she
does
not understand certain concepts or ideas presented by the author?
* Does the reader make omissions that disrupt the ongoing
construction of meaning and comprehension?
* When the reader comes to an unknown word, does he/she
put in a substitute word to maintain meaning?
* Is the reader able to understand certain concepts and
ideas
even if he/she mispronounces words or phrases related to the
concepts?
* Does the reader habitually associate the same words with
each
other? Are these associations disruptive to meaning?
(Was/saw, were/where, who/what, with/which, of/for/from)
Make the point that if the readers are using semantic cueing
these miscues typically won't occur.
* To what extent is the reader able to understand what the
antecedents are for pronouns in the text?
* A few considerations for judging semantic unacceptability:
- Proper names that are miscued are not semantically unacceptable
- Corrected miscues are not semantically unacceptable
- Grammatical miscues that don't change actual meaning are not
semantically unacceptable
- When a sentence is broken by punctuation (reading through
or inserting a period) it is best to look at the semantic
acceptability of the "reformulated" sentence than the original
one
- Often the reader gives you additional cues about his/her sense-
making. If they have lost the meaning, you may often see more
repetition, quizzical looks, change in voice inflection, or change
in body language.
- Use the criterion of predictability when puzzled. Read prior to
the miscue in question and knowing what you know of the text
and given your background information, would this miscue be
one that you could predict....one that does not distort the
essential meaning of the text
– With Regard to
the SYNTACTIC CUEING SYSTEM
* Does the reader read certain words or phrases in his own
dialect?
* Are the reader's insertions plausable given the syntax of his
native language or dialect? Do they show that the reader is
making predictions based upon his knowledge of syntax.
* Does the reader attend to punctuation as a clue to meaning?
* What sentence structures seem to give the reader difficulty?
*
What types of corrections or adjustments does the reader
make when his sense of syntax begins to breakdown?
*
Is the reader able to recognize and use common patterns
of text structure to construct meaning as he is reading?
– With Regard to
the PRAGMATIC CUEING SYSTEM
* To what extent does the reader expect text to make sense?
*
In what ways does the reader use cues from the environment
to figure out what a text might say?
*
To what extent is a reader able to understand what each
character is saying in a written dialogue?
*
In what ways does a reader converse with others about what
a text might mean?
– With Regard
to the GRAPHOPHONEMIC CUEING SYSTEM
* Does the reader rely on the sound-letter similarity to the
exclusion of thinking about constructing meaning?
*
Does the reader focus only on the first letter and then say
a word beginning with that letter without necessarily making
sense of the story?
*
To what extent does the reader overcorrect miscues that make
no difference to the meaning?
*
What consistent habitual associations does the reader make?
(Making errors like "was/saw" or "with/which"
*
How much attention is paid to "sounding out" words and not
Correcting via intonational and phrasing differences?
* Does the reader omit unknown words?
– With
Regard to the ANALYSIS OF CORRECTIONS
*
Corrections give you further insight into the dynamic process
since the tendency for appropriate correction depends on
"ongoing comprehending". That is, determining if meaning
has been compromised and what to do about it.
*
These corrections and their interpretation are related to the
quality of miscues and provide insight into the reader's
confirming strategies.
*
Focuses on success in self-correcting and in the reader's
judgments concerning which miscues should be corrected and
why
*
It is important to recognize that the percentage of corrections
themselves provides little insight. So you have to compare
comprehension with the corrections.
*
CORRECTIONS are considered EFFICIENT if:
– High Quality miscues (see below) are typically not
corrected.
– Low Quality miscues are corrected
– Low Quality miscues are often immediately corrected – so
that they become "partial miscues".
– Efficient corrections are indications of MORE
PROFICIENT READERS
*
CORRECTIONS are considered LESS EFFICIENT if:
– High Quality miscues are often corrected
– Low Quality miscues are typically not corrected
– When there is a tendency to correct high-quality miscues
and these corrections make little change in meaning,
it is often an indication of too much focus on the graphic
information in the text.
– Inefficient corrections are indications of LESS
PROFICIENT READERS
*
The APPROPRIATE corrections suggest the reader's awareness
and concern about the Cueing systems
– The appropriate correction of syntactic miscues
suggests a concern about linguistic structure.
– The appropriate correction of semantic miscues
suggests a concern about reading for meaning and
"Sense-making".
*
Beware of a simple "counting of corrections". The lack of correction
may be due to a number of factors.....so you have to use
EXPLANATORY analysis to consider:
– Readers DO NOT correct all miscues
– Miscue correction must be understood in relation to the
quality of the miscues produced
– Proficient readers may not ALWAYS correct low quality
miscues if they don't consider them significant to the
development of meaning. (They are more selective)
– At times, proficient readers may do "silent corrections"
so
that you can't see the corrections but they have strong
comprehension
– Many readers may not correct if there are too many
low-quality miscues clustered in the same phrase.
Often, rethinking or re-construction of meaning as they
continue occurs
– Retelling give a good indication of this.
*
This information should be considered along with the uncorrected low
quality miscues
– With Regard
to the QUALITY OF THE MISCUES
* Miscues can be generally rated as follows:
– High Quality Miscues are semantically
acceptable or
represent miscues that are subsequently corrected
because the text has stopped making sense.
– Low Quality Miscues are semantically
unacceptable
and there are few or no attempts to self-correct.
*
Proficient Readers tend to correct low quality miscues because
they interfere with meaning within the sentence and to leave
high quality miscues alone because they don't interfere with
meaning.
*
Non-proficient Readers tend not to correct low quality miscues
as much and/or correct high quality miscues. This is nearly the
reverse of the Proficient Reader.
*
The difference between the two is that the Proficient Reader is
Involved in "on-going comprehending" and is reading for meaning
while the Non-Proficient Reader does not read for meaning as well
– or at all.
– With Regard
to the OTHER MINOR CUEING SYSTEMS
* How does the reader use pictures in the text to figure out unknown
words?
*
How does the reader use pictures to confirm or disconfirm her
understanding of the text?
*
Is the reader confused by where and how pictures/figures are
placed in related to printed text?
*
Is the reader able to read maps, bar graphs, and similar figures
that are part of a text?
*
In nonfiction materials, does the reader understand the function
of main headings and subheadings and how those can be used
to aid comprehension?
*
In nonfiction materials, can the reader explain how bold or
other special print is used, especially in relation to new
vocabulary?
.
– With
Regard to the DETERMINATION OF COMPREHENSION
*
Although there is a definite relationship between miscues and
comprehension, it is not always a direct or easy relationship.
– It is necessary to look at the quality of the
miscues
– This is dependent on how the miscue affects the
meaning of the passage.
*
The following relationships should be considered
High Quality miscues — Good Comprehension
High Quality miscues — Poor Comprehension
Low Quality miscues — Poor Comprehension
Low Quality Miscues — Good Comprehension
*
Since High Quality Miscues are the result of good on-going
comprehending, it stands to reason that overall good
comprehension frequently co-occurs with HQMs.
*
When you note High Quality Miscues but poor comprehension,
then you have a mismatch between on-going comprehending
and overall comprehension. This is typically due to the reader
having poor background information about the topic at hand.
*
Low Quality Miscues suggests that on-going comprehending
is poor so it is not surprising that overall comprehending will
be poor also. When a reader is unable to construct meaning
as he/she is reading, then the final comprehension of a text will
also be poor.
*
Low Quality miscues and good comprehension is a bit puzzling.
It may be that the reader does correct miscues internally so that
the on-going comprehending is occurring but it is not evident
to us. Since it is going on, however, good overall comprehension
ensues. Also check to see whether there is a difference between the
beginning of the passage and the end of the passage. It may be that
the on-going comprehending got better as the reader got "involved"
in the passage and so the majority of miscues are at the beginning
but then improve so that there is good overall comprehension.
– It is
not a bad idea – at times – to wait longer before interrupting a
student who has made a miscue and is struggling with it than you
feel comfortable doing. This is to see whether the reader will figure
out a word through the context of the text independently.
READING MISCUE INVENTORY
(Goodman, Watson & Burke, 1987)
The procedure uses a Taxonomy of Reading Miscues to evaluate, categorize, and explain the miscue phenomena found to be prevalent in every reader. The procedure evaluates each of the reader's consecutive miscues through a series of questions designed to gain the greatest amount of information about the causes of miscues and their influences on reader comprehension.
Within this procedure, the decisions about the specific miscues analyzed are actually dependent on specific forms and interpretations formed by the authors of this procedure
As with all behavioral sampling procedures, the RMI requires the taping of the student's reading aloud so in-depth analysis can occur at a later time.
All deviations from the actual print are marked – whether they are high quality or low quality -- and whether they are corrected or not.
There are several forms that can be used for the RMI. The differences
between them pertain to what the focus of the coding and interpretation
happen to be:
– Procedure I
* The most detailed analysis
* Each miscue is analyzed on the basis of six questions
- Syntactic acceptability
- Semantic acceptability
- Degree of meaning change
- Whether there was self-correction
- Graphic similarity
- Sound similarity
* There is also a way (once the questions are answered) to
calculate
a student's attempts at meaning construction and creating
grammatical relationships for each miscue
* This form does not focus on the semantic acceptability of
each
sentence. Rather, the focus is on each individual miscue.
– Procedure II
* Very useful for classroom teachers
* Some counting of behaviors does occur
* Coding in three areas
- language sense
- graphic and sound similarity of word substitutions in context
- retelling
* Language Sense
– focus on sentence and decide if each one is
Syntactically acceptable
Semantically acceptable
How much meaning change has occurred
– patterns of strength, partial strength, and weakness is then
determined
* Similarities
– word substitutions are marked as highly, somewhat, or not
graphically similar or similar in sound.
* Retelling
– scoring can be done anecdotally or via a holistic score
The Procedure for Obtaining a Reading Sample
Although these procedures are taken from Goodman, Watson,
& Burke,
they are fine data collection procedures overall.
1. You must collect and examine a single and complete
oral reading
experience followed by a retelling.
2. Selecting Materials
-- should be
unfamiliar and unpracticed to the reader
-- but the concepts
should be known to the reader
-- must be difficult
enough to challenge the students strategies
but not so difficult that the reader cannot proceed
independently.
-- Selection
of material one grade level above the student's
assigned reading level and one or two alternative selections
above and below that are usually sufficient.
-- Need at least
25
miscues for an in-depth, well-rounded analysis.
-- Choose an
entire cohesive text that is both of interest to the student
and well-written.
-- The kind
of text (trade book, textbook, newspapers, basal reader)
depends on your objectives.
-- Passage
should typically be longer than 500 words.
-- Try
to get texts that strike a balance between the language or the
format of the text and predictability.
3. Prepare a typescript
-- You
use this to follow along
-- You
use this to record miscues, verbal asides, and any
significant nonverbal actions.
-- Should
make this typescript and the text look as much alike
as possible.
Same line length.......spelling....punctuation...
-- Triple
space so you have room for writing the miscues.
-- Number
with the page/line system (0110).
4. Getting ready
-- A tape
recorder in good working order with tape labeled
and dated.
-- Suitable
reading selections
-- Type
scripts of the selections
-- Retelling
guides
-- Pencils
for writing miscues
5. Informing the Reader
-- Make
the reader as comfortable as possible
-- Can
sit beside or across
-- Atmosphere
is informal and friendly
-- Chat
briefly before beginning
-- Briefly
tell reader why they are being asked to read and let
them know what is expected of them (e.g., "read this story
aloud....if you have trouble with it, try to figure it out like
you would if I wasn't here........after you read it, I'm going to
ask you to retell the story for me")
-- Let
the reader leaf through to see how long it is and ask if
they have read the passage before.
--
Remind them to read aloud if they forget during the reading.
--
NEVER
stop them except when they are making very few
miscues (and go to a more difficult passage) or when they
are extremely uncomfortable and are not able to continue
independently.
6. Oral retelling Procedure
--
Good idea to have a written outline available
--
You can compare...but be willing to be flexible with this.
--
Prepare a retelling guide by carefully analyzing the selection in
order to become familiar with and to determine all its features
(characterization, setting, events, plot, theme, concepts,
organization and stylistics).
--
You want to obtain both an unaided and an aided retelling.
Suggestions from Goodman on the retelling portion:
* Get to know the readers
* Become familiar with the story
* Avoid giving the reader the information from the text
* Include in questions and comments only information
introduced by the reader.
* Don't rush yourself or the reader. Think through
your questions and patiently wait for the reader's reply.
* Make your directions and questions very clear and
avoid giving more than one question at a time.
* Don't take "I don't know" for an answer. Rephrase
questions to get information another way. At the
same time, don't exhaust the reader with too great a
focus on any one topic.
* Let student's develop a topic and reach their own
conclusions before changing the subject.
* Ask open-ended questions. Questions that can be
answered with yes or no or with single words often
limit the reader's presentation potential.
* Retain any nonwords or name changes given by the
reader unless they self-correct them.
Interpretation of Miscue Analyses
1. Collect the data
and code the miscues as described in the
procedural description.
2. This procedure has
a special interpretation guide – but it is
best to combine the guide with both the earlier series of
questions and the following considerations.
3. Engage in the analysis
and interpretation activities
-- The focus is on the actual questions listed earlier and in
this
section
-- The questions are
** Asked about each miscue
** Patterns of miscues in relationship to other
miscues is a focus
** The questions evaluate the relationship between
miscues and linguistic systems of the text and of
the reader, Language of the reader and the
author, Concepts of the reader and the author,
The Reader's use of sampling, predicting, and
confirming strategies
-- Rarely do miscues involve only a single system of language
or an isolated reading strategy so all must be assessed.
The earlier set of questions and the analyses are systematic
ways of approaching this task.
Brief considerations for each major question:
Syntactic Acceptability and Semantic Acceptability
-- These questions reveal how proficiently a reader uses
prediction and confirmation strategies.
-- Proficient readers tend to produce miscues that are
semantically and syntactically acceptable within the context
of the written text.
-- When proficient readers do produce unacceptable miscues,
they make use of their confirming strategies. (They tend to
self-correct)
-- Less proficient readers are not so consistent in
their patterns
and may produce acceptable sentences less than half of the time.
-- The ability to produce semantically and syntactically acceptable
structures, or if the structures are unacceptable, to correct them,
provides evidence of a reader's predicting and confirming strategies.
Syntactic Acceptability
* A sentence can have an acceptable syntactic structure
without having acceptable meaning.
* The Goodmans use "syntax" but they mean "grammar"
* Usually the focus is on the syntactic structure or the
morphological units that give the elements their syntactic
"slots" and not on the actual words.
Colorless green ideas sleep furiously
The flommus kepbocked with the netwok
The girl out-ran the rabid raccoon
* Syntactic acceptability creates a pattern within which
appropriately ordered words, phrases, and clauses
support meaning.
* This reveals the degree to which the sentence sounds like
language.
* It reveals the success with which the reader is controlling
the structure of sentences as well as the relationship
of the sentences to the whole text
Semantic Acceptability
* Focuses on the success with which the reader is producing
understandable structures
* Ask the question, "Does this sentence fit into the story as
a
whole?"
* Semantic acceptability is judged after syntactic acceptability,
and a sentence is always considered to be semantically
unacceptable if it is syntactically unacceptable.
* Sometimes sentences are acceptable on the sentence level but
not acceptable within the context of the whole text.
We'll do it tomorrow. So the next month the wife went
off to the forest.
* The questions of semantic acceptability are always
considered within the complete text. This requires
not only awareness of the syntactic and semantic
coding systems.....but of the pragmatic coding system
as well.
* You may, of course, examine acceptability in sentences
alone and relate acceptability to self-correction......this
will allow you to focus on the reader's use of acceptance
or confirmation strategies.
* Keep in mind the idea of Expected Response when
focusing on meaning and remember that it is influenced
by your cultural and experiential background.
Pragmatic Acceptability
* Focus is on the amount of background information the
reader has to build "on-going comprehending"
* This is partly subsumed under "semantic cueing"
but it requires the additional knowledge of the world
* The reader is always using the coherence metric
–
that is, "does this make sense based upon what I know
of the world".
* Readers with problems with background information
often exhibit problems such that they have fewer miscues
but relatively poorer comprehension.
Graphic Similarity and Sound Similarity
* These provide data about the degree to which readers
use
the graphophonic system.
* The two systems (graphic and sound) are considered
separately
* Only word for word substitutions are examined
for graphic
and sound similarity.
* If it is not possible to determine which word is substituted
for which (complex miscues, insertions or omissions) the
questions are not asked.
* Even proficient readers do not use the graphic and
sound
systems to a high degree all the time.
* Only about half the substitution miscues show some
similarity to the text item.
* Less proficient readers generally have higher degrees
of
graphic and sound substitutions.....they rely on it and
not meaning making more. (Aspectual readers)
Graphic
Similarity
* the sequence and shape of the written miscue and the
text word must be examined with no concern for
pronunciation. There are many varieties of pronunciation
for the same graphic item....... particularly with dialectal
speakers.
* Divide into graphemes (three parts) See criteria below.
Sound
Similarity
* The graphic shape is not considered. Close your eyes
and say the two words. Ask: "How much do the words
sound alike?"
* Divide the words into three parts (beginning, medial,
and ending sounds). If two are alike and in the correct
position, then there is a high degree of similarity
* If only one is alike.....then there is some similarity
Intonation
* These are significant clues to a reader's processing of
language units.
* Generally caused by anticipation of a different structure
or because they are not familiar with the author's style
or structures.
* The key to intonation is not the "terminal juncture" but
the overall intonation contour of the sentence.
* The degree of stress and pitch that "flows" from the
beginning to the end.
* Intonation shifts due to pitch, stress, and pauses are
considered miscues when they change meaning or
the syntax.
Corrections
* These do relate to the ways that the actual cueing systems are
woven together with various strategies.
* Corrections tell you how well the individual is doing "on-line"
and with "on-going" comprehending.
* At least three activities are going on during the process
that
results in miscue corrections
– The various cueing systems are employed to determine
what the meaning of the visual text happens to be
– The meaning-making strategies of predicting, sampling,
and confirmation are used in real time to create
"pragmatic maps" or on-going comprehending
– The reader demonstrates both their facility with
self-correction, their awareness of the needs to
ensure effective meaning-making, and gives entre
into their judgments concerning which miscues
should be corrected.
Peripheral Vision
* To what degree do readers pick up graphic information
in the periphery of the text being processed?
* Near – defined as line immediately above or below
Far – second or third line above or below
* Only COMPLETE WORDS OR PHRASES ARE
CONSIDERED HERE.
Additional Cueing Systems
* Additional cueing is provided by pictures, captions,
titles
and subtitles, charts and graphs.
* If there is a good match between the text and these
other
visuals, students will frequently use them to figure out
unknown words.
* Moving away from using pictures as cues is one of the
challenges children face in first learning to read the
more extended text of chapter books.