- Published: Wednesday, 06 November 2013 11:48
- Hits: 16671
Dr Kerry Hempenstall, Senior Industry Fellow, School of Education, RMIT University, Melbourne, Australia.
All my blogs can be viewed on-line or downloaded as a Word file or PDF at https://www.dropbox.com/sh/olxpifutwcgvg8j/AABU8YNr4ZxiXPXzvHrrirR8a?dl=0
The assessment of children’s reading progress has long been of interest to teachers, researchers, and parents. The purposes for reading assessment include comparing one child‘s progress to that of his peers, screening students for special assistance, measuring an individual’s progress over a period of time, diagnosing particular areas of strength or weakness, using information for decisions about instruction, and determining placement within a reading program or special facility. There have been many different approaches to reading assessment based partly upon these differing purposes, but also upon the conception of reading development held by the test designer or user.
Miscue analysis was initially a major whole language procedure (for a fuller discussion of whole language, see A HISTORY OF DISPUTES ABOUT READING INSTRUCTION on this site) designed to assess the strategies that children use in their reading. Goodman and his colleagues in the 1960’s were interested in the processes occurring during reading, and believed that miscues (any departure from the text by the reader) could provide a picture of the underlying cognitive processes. He used the term miscue, rather than error, reflecting the view that a departure from the text is not necessarily erroneous (Goodman, 1979). Readers’ miscues include substitutions of the written word with another, additions, omissions, and alterations to the word sequence.
Initially, he developed a Taxonomy (Goodman, 1969) which detailed 28 different types of miscues. Established for research purposes, its unwieldiness and a desire to broaden its usage led Yetta Goodman and Carolyn Burke in 1972 to develop a briefer version comprising nine questions to be asked about each miscue - a simpler system that they believed would become a useful and manageable tool for clinics and for teachers in the school system. The authors were less interested in traditional quantitative measures such as reading accuracy or reading rate, and considered that their qualitative approach provided more fine-grained and relevant information than did other approaches to reading assessment. In the Reading Miscue Inventory (RMI) a student’s incorrect response, when compared to the written word, may display a dialect variation, an intonation shift, graphic similarity, sound similarity, grammatical similarity, syntactic acceptability, semantic acceptability, meaning change, and self-correction with semantic acceptability to the text word. An inventory of a child’s miscues involves selecting text of sufficient length and difficulty to ensure that a child will make at least 25 errors, tape-recording the oral reading, and assigning the child’s miscues to one or more of the nine categories. A further step requires a retelling of the story by the student as a comprehension check. The RMI requires about 20-40 minutes with each individual child, and a further hour for analysis. Through miscue analysis, the authors argued, teachers can better monitor a child’s progress along the path to reading success, and identify the strengths and needs of students. Depending on the prevalence and type of miscue, teachers may decide whether any intervention is required and also its focus.
The value of any assessment tool depends upon the importance of the quality to be measured and the capacity of the tool to perform its task. For example, the measurement of height does not provide important information about reading. Despite the fact that height can be measured quite accurately using appropriate instruments, it has neither theoretical relevance to reading, nor does it correlate even moderately with reading development when both height and reading are assessed across the population. Thus, a consideration of the RMI involves two questions, each of which must be answered in the affirmative for the Inventory to be useful: Are the qualities that the instrument purports to measure significant as indicators of reading progress given the current knowledge about reading and its development? Is the instrument a valid and reliable indicator of the presence or absence of the targeted qualities?
The first question for the RMI involves its theoretical relevance to reading development. What is the status of the whole language view of reading development and of skilled reading? This question is crucial because miscue analysis is predicated upon the whole language conception of reading, and hence stands or falls on the validity of this conception. The significance of any reading errors is thus superimposed on the reading behaviour through the adoption of the whole language conception of reading - “ ... the model of reading makes the understanding of miscues possible” (Brown, Goodman, & Marek, 1996, p.vii). Goodman believes that the process of reading development involves the gradual integration of three cueing mechanisms (semantic, syntactic, and graphophonic), although the graphophonic system is considered a lesser contributor - even potentially disruptive if over-relied upon by readers (Weaver, 1988). He considers skilled reading to entail as little emphasis on each word’s letter-construction as possible. Rather, it is perceived as a process of ongoing prediction of target-words based primarily upon semantic and syntactic cues, followed by confirmation that the chosen word is consistent with the context (and possibly the target word’s initial letters). “In turn his sense of syntactic structure and meaning makes it possible to predict the graphic input so he is largely selective, sampling the print to confirm his prediction” (Goodman, 1973, p.9).Even in more recent times after this assertion has been discredited, they have continued to assert the primacy of context: ‘Our research demonstrates that the least proficient readers we studied in the 6th, 8thand 10th. grades use graphic information more than the most proficient readers.’ Goodman & Goodman (2004, p. 630)
Consistent with this view of skilled reading, the Reading Miscue Inventory is concerned largely with errors that cause a loss of meaning, the number of errors being less important than their immediate impact on comprehension (Weaver, 1988). There are differences in the acceptability of various miscues. Good miscues maintain meaning and are viewed as an indication that the student is using meaning to drive the reading process, and hence, is on the correct path. Bad miscues are those that alter meaning. Whether the word the student reads corresponds to the written word may not be important in this conception. “Accuracy, correctly naming or identifying each word or word part in a graphic sequence, is not necessary for effective reading since the reader can get the meaning without accurate word identification” (Goodman, 1974, p.826).
Research has demonstrated that this assertion is incorrect. Good readers, though more sensitive to context cues to elicit the meaning of unfamiliar words, do not need to use context to decode unknown words (Tunmer & Hoover, 1993). At best, even good readers can guess words only one time in every four, and then only with fairly predictable words (Gough, 1993). They soon learn that word structure more reliably supplies the word’s pronunciation than does context; unfortunately, it is poor readers who are more likely to invest attention on such context guesswork (Nicholson, 1991). The error made by whole language theorists was to confuse the desired outcome of reading instruction - a capacity to grasp the meaning of a text - with the means of achieving that end. In order to comprehend meaning, the student must first learn to understand the code (Foorman, 1995).
A teacher using the RMI will examine the nature of the errors the student has made in the chosen passage. Consider this text and a reader’s response, substituting pony for horse:
The man rode his horse to town.
Asking the nine questions reveals that the miscue (compared with the target word) has grammatical similarity, syntactic acceptability, semantic acceptability, does not change meaning, and the miscue does not involve dialect variation, an intonation shift, graphic similarity, sound similarity, or self-correction. Such an error is considered an acceptable miscue. Reading pony for horse is indicative of the student using contextual cues appropriately and a signal for satisfaction about reading progress. The teacher would be content with this error, as meaning has been more or less preserved.
Often substitutions of words like a for the, by for at, in for into, do not cause a change in meaning. ... substitutions like daddy for father, James for Jimmy ... are generally produced by proficient readers and are not reading problems (Goodman & Burke, 1972, pp.101-102).
Suppose another student reads house for horse:
The man rode his horse to town.
This response is considered an unacceptable miscue. Asking the same nine questions reveals that the miscue (compared with the target word) has graphic similarity, some degree of sound similarity, grammatical similarity, syntactic acceptability, and the miscue does not involve dialect variation, an intonation shift. Further, it does not include self-correction, is not a semantically acceptable change, and the miscue creates meaning change. The Inventory does not place a strong emphasis on graphemic and phonemic accuracy as being of primary importance in reading. “Proficient readers resort to an intensive graphophonic analysis of a word only when the use of the syntactic and semantic systems does not yield enough information to support selective use of the graphophonic system” (Goodman et al., 1987, p.26). For an extended discussion of the 3 cueing process, see the paper on this site The three-cueing system in reading: Will it ever go away?
Despite the closer graphemic similarity of the response house to the target word, children who make errors based on graphemic similarity, such as house for horse, are considered problematic and over-reliant on phonic cues. The teacher would encourage this latter child to learn to rely more on context than on letter patterns. Even in more recent times, Goodman has reiterated his belief that phonics is not an appropriate focus for instruction: ‘Our research demonstrates that the least proficient readers we studied in the 6th, 8thand 10thgrades use graphic information more than the most proficient readers.’ (Goodman & Goodman 2004, p. 630).
Because whole language theorists argue that good readers’ miscues display less grapho-phonemic similarity to the target word than do those of poor readers (Weaver, 1988), it is unsurprising that they consider readers-in-training should do likewise. Empirical research (Adams, 1991) has found the reverse to be true, good readers’ miscues display more grapho-phonemic similarity to the target word than do those of struggling readers. In 2009, Beatty and Care in Australia studied the reading miscues of a hundred children between 5-8 years of age noted that both average and above average readers relied more on graphophonic cues than their below average classmates when reading unfamiliar text.
In fact, most nascent readers’ miscues shift over time, from early errors based upon contextual similarity to those based upon grapho-phonemic similarity; however, this shift is now recognised as functional and a characteristic of progress. The student’s dawning understanding of the pre-eminence of a word’s graphemic structure encourages close visual inspection of words, a strategy that accelerates the progressive internalisation of unfamiliar spelling patterns; that is, it leads ultimately to whole-word recognition. That some teachers may unwittingly subvert this process with well-meaning but unhelpful advice to readers is an unfortunate outcome.
According to current knowledge, the house response is a preferable error to the pony substitution. It is a sign that the student is in the process of acquiring the alphabetic principle; however, corrective feedback should be provided as house is an erroneous response. Through the error correction, the student’s attention is directed toward the letters in the written word and the sound usually made by the or combination. The response recommended to teachers through the RMI - that of directing the students’ attention away from the letters in the word to that which can be predicted and which makes sense - provides an alarmingly unstable and counter-productive rule for students. Suppose the student had substituted bicycle for horse. The substitution makes sense but is far from that which the author intended. The fact that it bears no graphemic similarity to the target word may not be salient to a student whose primary decoding strategy is driven by semantic and syntactic similarity. The instructional message to the student is that, despite the errors directly attributable to the strategy of guessing, the strategy is the appropriate one. The student is encouraged to continue using a strategy that is unhelpful, and is dissuaded from attending to the major cue - the word’s structure.
The idea that miscues are acceptable if meaning is essentially preserved appears unduly reckless to many observers. However, it is consistent with the constructivism of the whole language philosophy: the meaning available to the reader cannot be the same as that which an author intended because of the different life experiences of the author and reader, and the different meanings they attach to words. Goodman (1992, 1994) refers to the construction of meaning as a transaction between the author and the reader. “There are two or more texts during reading: the published text and the reader’s text. ... the text is changed as the reader constructs it to fit expectations and world knowledge” (Goodman, 1992, p.358). Weaver (1988) describes the extreme view - “Meaning is not in the text, nor will the meaning intended by the writer ever be perceived (or rather, constructed) exactly the same by the reader” ( p. 40). This is a cavalier attitude to say the least; one would hope that the readers of a machine’s safety manual were capable of extracting a meaning closely aligned to that which the writer intended!
In order for reading to develop (according to the whole language conception of skilled reading), it is expected that students will make many miscues during the progressive integration of the cueing systems, and these errors are not necessarily a cause for intervention but simply a sign of a reader prepared to take risks. Any corrective feedback regarding errors is risky as it may jeopardise the student’s willingness for risk-taking. “ ... if these resulting miscues preserve the essential meaning of the text, or if they fail to fit with the following context but are subsequently corrected by the reader, then the teacher has little or no reason for concern” (Weaver, 1988, p.325).
Self-corrections are errors that are corrected without another’s intervention, usually because the word uttered does not fit in the context of the sentence. Within the whole language framework, self-corrections are a clear and pleasing sign that meaning and syntactic cues are being integrated into the reader’s strategies. Clay (1969, cited in Share, 1990) asserted that good readers self-corrected errors at a higher rate than did poor readers. She considered high rates were indicative of good text-cue integration, which in turn was a measure of reading progress. The significance of self-correction has been questioned by Share (1990), and Thompson (1981, cited in Share, 1990). They found that self-correction rates are confounded with text difficulty. When text difficulty was controlled in reading level-matched designs, the rates of self-correction became similar among good and poor readers. That is, when text is very difficult everyone is more likely to make errors and increase their rate of self-correction. Hence, an increased rate of self-correction can be interpreted as simply indicative of excessively difficult text rather than as reflecting reading progress. This interpretation based on difficulty levels also presents problems of unreliability for the assessment of self-correction rates. The conclusion that there is no direct support for self-correction as a marker or determinant of reading progress makes the activity of recording such ratings for students of questionable value.
How does the view of reading underpinning the RMI sit with research findings regarding the reading process and its development?
This view of skilled reading, which comes from Goodman (1967) and Smith (1978), has been rejected by the scientific community (Adams, 1990; Ehri, 1986; Goswami & Bryant, 1990; Gough, Ehri & Treiman, 1992; Just & Carpenter, 1987; Perfetti, 1985; Rayner & Pollatsek, 1989; Rieben & Perfetti, 1991; Stanovich, 1986, 1991; Vellutino, 1991). Skilled reading is not sampling features of the text on the run, it is not a psycholinguistic guessing game, and it is not incidentally visual. Rather, research has shown that ‘skilled readers process virtually all the words they encounter in connected text, and typically, all of the letters in those words’ (Vellutino, 1991, p.82). Research further indicates that skilled readers are sufficiently fast and accurate at recognising words in text to make reliance on contextual information unnecessary (Perfetti, 1985).
(Tunmer & Hoover, 1993, p.167)
The findings of individual researchers and such syntheses as provided above have been formalised through the National Institute of Child Health and Human Development (NICHD). In 1985, the Health Research Extension Act directed the NICHD to coordinate research on reading disability and learning disability such that results of research would meet a number of criteria regarding scientific rigour. The intention was to define research characteristics that would ultimately lead to methodologically unassailable findings and benchmarks of consensual knowledge. More than 100 researchers in numerous sites across the USA are involved in this cooperative multidisciplinary research employing large scale longitudinal studies, careful sampling, and replication of findings with the view of integrating their research efforts. A summary of the findings is provided by the director, Reid Lyon (1996):
The ability to read fluently for meaning depends primarily on rapid, automatic decoding and recognition at the level of the single word. The basis of reading deficits (phonological processing) should provide the focus for intervention. Efforts should be directed at explicitly and systematically teaching the connection between phonological rules and the written word. A phonics emphasis provides advantages for disabled readers over a whole language approach.
The NICHD research summary was influential, even at a political level. In 1997, the US federal Reading Excellence Act was enacted. This legislation was intended to ensure that all reading programs eligible for federal support in future would be based on reliable and replicable research. Part of the definition of reading included in the Bill provided a clear indication of consensus concerning the incompleteness of the whole language view of reading. Reading is "The process of comprehending the meaning of written text by [depending] on the ability to use phonics skills (i.e., knowledge of letters and sounds) to decode printed words quickly and effortlessly both silently and aloud”. Despite this apparent clear direction, the move was extraordinarily controversial, and evoked strong resistance from within the education field.
In a similar vein, the British National Literacy Strategy (1998) and more recently, the Primary National Strategy (2006) was released to all primary schools, requiring them to abandon the Whole Language approach to reading. Components of the former system, such as reliance on context clues to aid word reading, are discredited in the Strategy, and schools are directed to introduce explicit phonics instruction from the earliest stages of reading. This latter strategy action was supported by the Rose Report (2006), and was more forthright about implementing evidence-based approaches to instruction. Despite much opposition from within education circles, the Primary National Strategy (2006) has had substantially more impact on classroom practice.
In recent times, various reports have highlighted the central role of alphabetics in beginning reading instruction. For example, the National Early Literacy Panel’s meta-analysis (2008) noted that “Few interventions improved conventional literacy skills or the precursor skills most related to later literacy growth, the exception being code-focused interventions” (p.5).
six variables representing early literacy skills or precursor literacy skills had medium to large predictive relationships with later measures of literacy development.
• alphabet knowledge (AK): knowledge of the names and sounds associated with printed letters
• phonological awareness (PA): the ability to detect, manipulate, or analyze the auditory aspects of spoken language (including the ability to distinguish or segment words, syllables, or phonemes), independent of meaning
• rapid automatic naming (RAN) of letters or digits: the ability to rapidly name a sequence of random letters or digits
• RAN of objects or colors: the ability to rapidly name a sequence of repeating random sets of pictures of objects (e.g., “car,” “tree,” “house,” “man”) or colors
• writing or writing name: the ability to write letters in isolation on request or to write one’s own name
• phonological memory: the ability to remember spoken information for a short period of time.
An additional five early literacy skills were also moderately correlated with at least one measure of later literacy achievement but either did not maintain this predictive power when other important contextual variables were accounted for or have not yet been evaluated by researchers in this way. These additionally potentially important variables include
• concepts about print: knowledge of print conventions (e.g., left–right, front–back) and concepts (book cover, author, text)
• print knowledge: a combination of elements of AK, concepts about print, and early decoding
• reading readiness: usually a combination of AK, concepts of print, vocabulary, memory, and PA
• oral language: the ability to produce or comprehend spoken language, including vocabulary and grammar
• visual processing: the ability to match or discriminate visually presented symbols (p.3-4).
The RMI was designed to provide a “window on the reading process” (Goodman, 1973, p.5); however, the analogy with a window is a misleading one as it implies a direct and transparent medium. The picture of reading obtained through the RMI involves an interpretation of what is viewed through this window. What is really displayed by a student is reading behaviour (words, sentences) - the subsequent analysis of miscues involves making inferences about unobservable processes based upon assumptions about the reading process. With this instrument, the picture is coloured by a discredited conception of reading.
An important rationale for the choice of an assessment device resides its capacity to inform intervention (Goyen, 1992). The RMI Manual, however, provides few strategies for corrective intervention - perhaps because miscue analysis was not originally developed to inform intervention. Alternatively, it may relate to the whole language view that reading progress is natural in a strongly literate environment. “Learning is continuous, spontaneous, and effortless, requiring no particular attention, conscious motivation, or specific reinforcement” (Smith, 1992, p.432). The typical global recommendation from the RMI for students with reading difficulties involves prompting the increased use of psycholinguistic guessing - “The reading strategies of sampling, predicting, and confirming are the same for all readers. ... non-proficient readers ... need to be invited to do what proficient readers do, their attention drawn to inferential strategies” (Goodman, Watson, & Burke, 1987, p.170). Additionally, students are expected to learn about reading through their mistakes rather than through instruction from teachers; hence, the reticence of the RMI developers toward explicit intervention strategies is understandable, even if unhelpful to students.
One implication of the current understanding of the reading process is that the qualitative analysis of reading errors is largely irrelevant to instructional planning. Decoding errors of whatever type are best addressed at the level of decoding instruction (Lyon, 1996). Thus, the student who makes errors due to reliance on contextual strategies and the student who makes errors based on inadequate graphophonic analysis each requires decoding instruction and practice, sufficient to enable effortless reading at the appropriate level of text difficulty. Psychometric studies have demonstrated that it is decoding ability that predicts children’s capacity for word identification and comprehension. Measures of semantic and syntactic ability as assessed in the RMI are not strongly correlated with word identification or passage comprehension (Vellutino, 1991).
Issues of validity and reliability
Modern research has indicated that the RMI does not provide important information regarding reading, and hence is of largely historical interest. However, even if its foci were of interest, there are other difficulties that create problems for its use. An assumption implicit in miscue inventories is that oral reading reflects similar processes to those involved in silent reading, and hence errors detected while students are reading aloud are representative of errors in their silent reading. However, even Goodman (1976) expressed suspicions about the usefulness of the results of oral reading assessment. “ ... ‘poor’ oral reading performance may reflect a high degree of reading competence rather than a lack of such competence” (p.489). If poor oral reading can be interpreted so diversely, and may be simply an artifact of the assessment, its value is dramatically compromised (at least insofar as its implications for silent reading ability).
An additional problem for the Reading Miscue Inventory is its inadequacy as a psychometric instrument (Allington, 1984). Leu's (1982) review of oral reading error analyses highlighted serious problems of unreliability. Unreliability in an assessment means that the same tool can provide differing results on different occasions, or with different texts, or with different examiners, without any change in the student’s capacity. The unreliability problems arise from:
(a) vague definitions of the boundaries of the error categories. Determining when meaning has been essentially preserved may produce different decisions from different teachers for the same miscue.
(b) an absence of theoretical justification for the categories;
(c) a failure to allow for the effects of passage difficulty. When passage difficulty is controlled (i.e., similar error rates), reliance on context occurs at least as much for less skilled as for skilled readers (Allington & Fleming, 1978; Batey & Sonnenschein, 1981; Biemiller, 1970, 1979; Cohen, 1974-5; Coomber, 1972; Harding, 1984; Juel, 1980; Lesgold & Resnick, 1982; Perfetti & Roth, 1981; Richardson, Di Benedetto, & Adler, 1982; Weber, 1970; Whaley & Kibby, 1981; cited in Stanovich, 1986);
(d) the ambiguity resulting when categorising multiple-source errors.
Hood (1982) noted that other text characteristics (besides difficulty) also influenced the type of error produced by readers. Wiederholt and Bryant (1987) further point out that miscues are influenced by the reading instruction the students have received, student age, the writing style, student familiarity with the text, and the stated purpose of the reading task (e.g., reading for expression compared to reading with a view to answering comprehension questions). Such contamination of results inevitably leads to inconsistent diagnosis and similarly inconsistent instructional implications. The findings of any individual assessment cannot be relied upon to provide information about the habitual strategies used by a reader, and thus fail the basic reliability requirement of an assessment instrument. Given the difficulty in separating these various potential causes of miscues, it is difficult to accept Goodman’s characterisation of the miscue analysis as providing a clear insight into the student’s cognitive processes.
The Reading Miscue Inventory has had considerable influence in instructional texts and in classrooms (Allington, 1984), and remains influential among Whole Language theorists and continues to have traction among many teachers (Weaver, 1988).
80% of the preservice teachers and 74% of the inservice teachers agreed that the most beneficial strategy for identifying an unknown word was to use the context to figure it out. Both groups agreed that a teacher should not be concerned when early readers' miscues do not alter meaning (76% preservice; 79% inservice). Only 22% of the preservice and 36% of the inservice teachers recognized that phonological awareness involves oral language and is not a method of reading instruction (Mather, Bos, & Babur, 2001).
In summary, there has long been a significant difference of attitude toward miscue analysis and running records between the (mainly) whole language fraternity and reading researchers. The underlying theory has been thoroughly discredited (Gough, 1993; Perfetti, 1985; Rieben & Perfetti, 1991; Nicholson, 1991; Stanovich, 1986, 2000; Turner& Hoover, 1993), and the method of assessing the miscues has been shown to be unreliable (Allington, 1984; Leu, 1982).
One would imagine that miscue analysis would have slipped quietly into education’s long history of intuitively derived concepts that proved incorrect when the light of empirical study was played upon them. However, consistent with the long standing disconnect between evidence and practice in education, miscue analysis and running records remain evident in many education settings (Beatty & Care, 2009), no doubt partly due to the continued acceptance among educators of the 3 cueing process.
Even after all the research that has highlighted the central role of phonological processes, there remains a seemingly unshakeable belief in the whole language legacy.
See examples below:
“Empirical investigations may demonstrate, as miscue analysis has, that phonics is a distinctly inferior cuing system, one hardly deserving any privileged status” (Strauss & Altwerger, 2007).
“When I first learned about Miscue Analysis as a young teacher myself, and then studied with Ken and Yetta, we saw the miscues, and we were pretty sure it was representative of the fact that children are definitely constructing meaning. They’re selective, they’re intentional, they’re predicting and so forth, but to actually see it in action is remarkable” (Goodman, Goodman, & Altwerger, 2012, p.351).
“Miscue analysis continues to be popular with teachers because it helps them to understand how their pupils make sense of print. It is in wide use in teacher education programs as a means of getting teachers to revalue the reading process. It continues to be used in research because it provides a depth profile of the reading process in use. No published critique of miscue analysis has shown it to be invalid” (Goodman, 2006, p.1)
“I have found that designing miscue analysis projects based on Goodman's research helps students identify cognitive activities inherent in reading and interpretation, encouraging them to consider ways of extending reading strategy instruction to the secondary and postsecondary level (Hoge, 1983; Paulson, 2001; Worsnop, 1980). I am indebted to Weaver (2002) and Wilde (2000) for providing essential resources for conducting a miscue analysis for preservice and inservice educators and would strongly encourage teachers to consult these sources before designing miscue analysis projects for the classroom” (Eckert, 2010, p.115).
A revised version - RMI: Alternative Procedures (Goodman, Watson, & Burke, 1987) offered four analysis options of varying complexity for classroom use. The rationale is unchanged - " ... it is best to avoid the common sense notion that what the reader was supposed to have read was printed in the text" (Goodman et al., 1987, p.60) - and the Alternative Procedures are subject to the same criticisms as earlier versions.
There have also been further developments on the theme, such as Retrospective Miscue Analysis (Goodman & Marek, 1996) that involves recording the reading session and discussing the recorded miscues with the student. This newer form of analysis suffers from the same deficiencies as the traditional form, but also has an opportunity cost in that it reduces the time that struggling students have for appropriate and intensive intervention.
More specifically, RMA (Retrospective Miscue Analysis) offers a context in which parents and their children can become more aware of their strengths as readers through the process of revaluing (K. Goodman, 2003b). Revaluing readers involves shifting the focus from reading as accurate, oral production to reading as comprehension (Y. Goodman, 1996). With this shift in focus, we help readers to recognize and concentrate on their strengths by valuing what they can do (K. Goodman, 2003b). Therefore, RMA discussions address the merit of high-quality miscues, which are miscues that do not change the meaning of the sentence or story (e.g., substituting sofa for couch). Ultimately, this study aspired to help parents and children to view miscues as part of common reading behaviors (Kabuto, 2009, p.213).
Over-the-Shoulder Miscue Analysis (OSMA) shares similarities with Retrospective Miscue Analysis (RMA), but is not voice recorded. It is similar to Running Records in that it involves a written list of miscues, but the “reading conference” component occurs in real-time, perhaps at the end of each page.
Similar to the talk that takes place during retrospective miscue analysis (Goodman & Marek, 1996), these conversations provide a model for the child to become thoughtful about her own reading and a context in which the child and teacher together can ask questions and explore the child's reading strategies, perspectives, and attitudes. We end our conversation by celebrating the student's successes as a reader (Davenport & Lauritzen, 2002,p.110).
Yet another acronym is Collaborative Retrospective Miscue Analysis (CRMA). In this variant (Costello, 1992), students are expected to discuss with their peers in small groups their reading miscues, retellings, and thoughts about what they have read.
RMA/CRMA is counter to the current focus on reading instruction as a discrete set of skills focused on what the reader is unable to do rather than what s/he is capable of doing, deskilling teachers in the process. Learning from students is at the heart of RMA/CRMA and critical to “expanding our knowledge about teaching and learning” (Y. Goodman, 1996, p. 600) (Seeger, 2009, p.13).
Wohlwend (2012) introduces spider charts to spruce up the appearance of the data collected. However, the old 3 cueing system remains unapologetically central.
The focus is on the unfolding meaning making with a text rather than on word-by-word accuracy. To do this, teachers must consider how individual readers are simultaneously pulling together multiple threads as they coordinate all their linguistic resources: their knowledge of familiar cultural patterns of language use, their developing sense of story, and their awareness of sounds and print (Wohlwend, 2012, p.110).
In these RMA related analyses the same two major criticisms apply – each of which is independently fatal to their usefulness. First, they do not represent how skilled reading develops, and may in fact be counterproductive. Second, they fail the psychometric test for reliability.
It should be noted that the critique presented here does not necessarily imply that the qualitative analysis of readers’ errors can only be valueless. However, it is essential for any proposed analysis that the rationale for error categories should be well grounded in knowledge about the reading process, the function of the analysis should be clearly explicated, and the instrument must have acceptable psychometric properties. It should also be recognised that such qualitative analyses may not be very useful in informing intervention. The authors of the RMI make the claim that their assessment is authentic because it makes use of literature for the assessment task. However, even authentic assessments should meet the requirement of relevance and trustworthiness. Given the problems with the theory, design, and implications of the Reading Miscue Inventory its continued acceptance in the education community is difficult to fathom.
A possible contributor involves teachers’ knowledge of reading research, and their capacity to implement evidence-based reading programs. There have been a number of studies in various countries pointing to a paucity of teacher training in the skills necessary for effective literacy instruction, and a resulting lack of teacher knowledge (Bos, Mather, Dickson, Podhajski, & Chard, 2001; Cunningham, Perry, Stanovich, & Stanovich, 2004; Fielding-Barnsley, 2003; Mather, Bos, & Babur, 2001; Moats & Foorman, 2003). Further, inservice education even if directed at evidence-based programs may lead only to increased knowledge without any discernible impact upon classroom practice (National Center for Educational Evaluation and Regional Assistance, Institute of Education Sciences, 2008; Roehrig, Duggar, Moats, Glover, & Mincey, 2008). Part of the answer clearly lies in the reform of teacher training, an initiative now recognised across a numerous counties, and the political will to ensure reform actually occurs.
For further discussion of evidence-based education, see on this site: Evidence-based practice in education: Why not?
Download the full text PDF here.
Allington, R.L. (1984). Content coverage & contextual reading in reading groups. Journal ofReading Behaviour, 16, 85-96.
Bos, C., Mather, N., Dickson, S., Podhajski, B., & Chard, D. (2001). Perceptions and knowledge of preservice and inservice educators about early reading instruction. Annals of Dyslexia, 51, 97-120.
Brown, J., Goodman, K.S., & Marek, A.M. (1996). Studies in miscue analysis: An annotated bibliography. Newark, DE: IRA.
Costello, S. (1992). Collaborative retrospective miscue analysis with middle school students. Ann Arbor, MI: University Microfilms International. (97-23547).
Cunningham, A. E., Perry, K. E., Stanovich, K. E., & Stanovich, P. J. (2004). Disciplinary knowledge of K-3 teachers and their knowledge calibration in the domain of early literacy. Annals of Dyslexia, 54, 139-167.
Davenport, M.R., & Lauritzen, C. (2002). Inviting reflection on reading through over the shoulder miscue analysis. Language Arts, 80(2), 109-118.
Eckert, L.S. (2008). Bridging the pedagogical gap: Intersections between Literary and reading theories in secondary and postsecondary literacy instruction. Journal of Adolescent & Adult Literacy, 52(2), 110-118.
Fielding-Barnsley, R. (2003). Teachers' knowledge of metalinguistics in the process of learning to read: Are universities providing adequate training? In B.Bartlett, F.Bryer & D. Roebuck (Eds.), Reimagining practice: Researching change, pp.209-215.
Foorman, B.R. (1995). Research on "the Great Debate": Code-oriented versus Whole Language approaches to reading instruction. School Psychology Review, 24, 376-392.
Goodman, K.S. (1969). Analysis of oral reading miscues: Applied psycholinguistics. Reading Research Quarterly, 5, 9-30.
Goodman, K.S. (1973). Miscue analysis: Applications to reading instruction. Urbana, Ill: National Council of Teachers of English.
Goodman, K.S. (1974, Sept). Effective teachers of reading know language and children. Elementary English, 51, 823-828.
Goodman, K.S. (1976). Behind the eye: What happens in reading. In H. Singer & R.B. Ruddell (Eds.), Theoretical models and processes in reading (pp. 470-496). Newark, DE: International Reading Association.
Goodman, K.S. (1979). Reading: A psycholinguistic guessing game. In H. Singer & R. B. Ruddell (Eds.). Theoretical models and processes of reading (pp. 259-271). Newark, DE: International Reading Association.
Goodman, K.S. (1992). Why whole language is today's agenda in Education. Language Arts, 69, 354-363.
Goodman, K.S. (1994). Reading, writing, and written texts: A transactional sociopsycholinguistic view. In Robert. B. Ruddell, Martha Rapp Ruddell, Harry Singer (Eds.). Theoretical models and processes of reading. (pp. 1093-1130). Newark, DE: International Reading Association
Goodman, Y., Goodman, K., & Altwerger, B. (2012). The best of times, the worst of times? Language Arts, 89(5), 349-353. Retrieved from http://www.ncte.org/journals/la/issues/v89-5
Goodman, Y.M. & Burke, C.L. (1972). Reading Miscue Inventory: Manual and procedures for diagnosis and evaluation. New York: MacMillan.
Goodman, Y.M., & Marek, A.M. (1996). Retrospective miscue analysis: Revaluing readers and reading. Katonah, NY: Richard C. Owen.
Goodman, Y.M., Watson, D., & Burke, C. (1987). Reading Miscue Inventory: Alternative procedures. New York: Richard C. Owen.
Gough, P. B. (1993). The beginning of decoding. Reading and Writing: An Interdisciplinary Journal, 5, 181-192.
Goyen, J.D. (1992). Diagnosis of reading problems: Is there a case? Educational Psychology, 12, 225-237.
Hood, J. (1982). The relationship of selected text variables to miscue scores of second graders. Journal of Reading Behavior, 14, 141-158.
Kabuto, B. (2009). Parents and children reading and reflecting together: The possibilities of family retrospective miscue analysis. The Reading Teacher, 63(3), 212-221.
Leu, D. J. (1982). Oral reading analysis: A critical review of research and application. Reading Research Quarterly, 17, 420-437
Lyon, G.R. (1996). Learning disabilities. The Future of Children: Special Education for Children with Disabilities, 6(1), 54-76.
Mather, N., Bos, C., & Babur, N. (2001). Perceptions and knowledge of preservice and inservice teachers about early literacy instruction. Journal of Learning Disabilities, 34, 472-482.
McCutchen, D., Green, L., Abbott, R. D., & Sanders, E. A. (2009). Further evidence for teacher knowledge: Supporting struggling readers in grades three through five. Reading and Writing: An Interdisciplinary Journal, 22, 401-423.
Moats L. C., & Foorman, B. R. (2003). Measuring teachers’ content knowledge of language and reading. Annals of Dyslexia, 53, 23–45.
National Center for Education Evaluation and Regional Assistance of the Institute of Education Sciences (2008). The impact of two professional development interventions on early reading instruction and achievement. Washington, DC: National Center for Education Evaluation and Regional Assistance, Institute of Education Sciences, U.S. Department of Education.
National Early Literacy Panel. (2008). Executive Summary. Developing early literacy: Report of the National Early Literacy Panel. Washington, DC: National Institute for Literacy. Retrieved from http://lincs.ed.gov/publications/pdf/NELPReport09.pdf
National Reading Panel. (2000). Report of the National Reading Panel. Washington, DC: National Institute of Child Health and Development.
Nicholson, T. (1991). Do children read words better in context or in lists? A classic study revisited. Journal of Educational Psychology, 82, 444-450.
Nicholson, T., Bailey, J., & McArthur, J. (1991). Context cues in reading: The gap between research & popular opinion. Reading, Writing & Learning Disabilities, 7, 33-41.
Nicholson, T., Lillas, C., Rzoska, M.A. (1988, Oct). Have we been mislead by miscues? The Reading Teacher, 6-10.
Perfetti, C. A. (1985). Reading ability. New York: Oxford.
Primary National Strategy (2006). Primary framework for literacy and mathematics. UK: Department of Education and Skills. Retrieved from http://www.standards.dfes.gov.uk/primaryframeworks/
Ramadiro, B.L. (2012). Reading in two languages: Evidence from miscue analysis. South African Journal of Childhood Education, 2(2), 74-92.
Rieben, L. & Perfetti, C. A. (Eds.) (1991). Learning to read: Basic research and its implications. Hillsdale, NJ: Lawrence Erlbaum
Roehrig, A. D., Duggar, S. W., Moats, L. C., Glover, M., & Mincey, B. (2008). When teachers work to use progress monitoring data to inform literacy instruction: Identifying potential supports and challenges. Remedial and Special Education, 29(6), 364-382.
Rose, J. (2006). Independent review of the teaching of early reading. Bristol: Department for Education and Skills. Retrieved from www.standards.dfes.gov.uk/rosereview/report.pdf
Seeger, V.N. (2009). Collaborative retrospective miscue analysis: A pathway to self efficacy in reading. Ph.D. Dissertation, Kansas State University.
Share, D.L. (1990). Self correction rates in oral reading: Indices of efficient reading or artifact of text difficulty? Educational Psychology, 10, 181-186.
Smith, F. (1992). Learning to read: The never-ending debate. Phi Delta Kappan, 74, 432-441.
Stanovich, K.E. (1986). Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy. Reading Research Quarterly, 21, 360-406.
Strauss, S.L., & Altwerger, B. (2007) The logographic nature of English alphabetics and the fallacy of direct, intensive phonics instruction. Journal of Early Childhood Literacy, 7(3), 299-320.
Tunmer, W.E. & Hoover, W.A. (1993). Phonological recoding skill and beginning reading. Reading & Writing: An Interdisciplinary Journal, 5, 161-179.
Vellutino, F.R. (1991). Introduction to three studies on reading acquisition: Convergent findings on theoretical foundations of code-oriented versus whole-language approaches to reading instruction. Journal of Educational Psychology, 83, 437-443.
Weaver, C. (1988). Reading process & practice: From socio-psycholinguistics to whole language. Portsmouth, NH: Heinemann.
Wohlwend, K.E. (2012). A new spin on miscue analysis: Using spider charts to web reading processes. Language Arts, 90(2), 110-118.
Wiederholt, J.L. & Bryant, B.R. (1987). Assessing the reading abilities and instructional needs of students. Austin, TX: ProEd.