A new spin on miscue analysis: Using spider charts to web reading processes

Wohlwend, Karen E. (2012). A new spin on miscue analysis: Using spider charts to web reading processes. Language Arts, 90(2), 110-118.


As a miscue analysis practitioner, my eye is always caught by articles like this that outline new ways of looking at the procedures miscue analysis involves. If you are not already familiar with miscue analysis and how it works, you may want to put the term into a search engine and familiarize yourself with how the procedures work before either reading on in this annotation or reading the article itself. Wohlwend does provide information on miscue analysis in her article, but the reader might be better served by some of the other sources that can be readily found online. I say this because although I find Wohlwend’s “spin” on miscue analysis to be an interesting touch that I may consider adding to my own practice of miscue analysis, there are some areas where I am not sure that Wohlwend’s explanations of miscue analysis fit well with what those who pioneered miscue analysis intended. More on that in a minute though.

Wohlwend here proposes a way to graphically represent the percentages that result from the various procedures involved in miscue analysis. Miscue analysis practitioners are interested in the balance among the various types of cues (graphophonic, syntactic, and semantic) they believe readers use when transacting with unfamiliar texts. Wohlwend wanted to show that balance (or imbalance, as the case may be) in a visual way. She found that the students in her university literacy methods course tended to focus too much on the percentages as numbers and not on the more holistic, overall sense of balance among the cueing systems. She devised a method that in effect creates a “picture” of how a reader uses cueing systems. Using Microsoft Office’s Excel, she first created a table for each of the four “big questions” we ask in miscue analysis, which I see as: 1) How much do the reader’s substitution miscues look and sound like what is printed in the text (graphophonic similarity)? 2) How many of the sentences a reader reads actually sound like the language that reader uses in speaking (syntactic acceptability?) 3) How many of the sentences a reader reads make sense (semantic acceptability)? 4) How often do the reader’s miscues significantly change or disrupt the meaning of sentences (meaning change)?

Wohlwend made an eight-cell Excel table with the big questions listed in one column and the child’s percentages listed in the second column. In this case, the percentage for “No Meaning Change” was put into the column rather than the percentage that the meaning was changed. Then Wohlwend simply clicked on “Insert” and then “Other Charts” and then “Radar”, which plots the data as a geometric shape that she calls a “tilted square” but that could also be called a diamond shape. Ideally, if all the cueing systems are balanced, the square or diamond should be symmetrical. If there is an imbalance, the shape will be skewed or distorted. I tried this with some data from a child I had done miscue analysis procedures with, and it was easy to do, and did provide a visual that helped to clarify what was going on with that child. It did not really add materially to the conclusions, but it did make them a bit more visible and might be useful when explaining miscue analysis results to teachers and parents, and maybe even older children, who should be involved in the interpretation process in miscue analysis.

Wohlwend provides a link to a web site where you can simply plug in your percentages and get the visual representation. It works fine, but if you have reasonable Excel skills you can follow the instructions in the article’s sidebar almost as easily. As a recent upgrader to the Windows 8 operating system (I have Office 2010), I noticed that Wohlwend’s instructions refer to Excel 2007 on the Vista operating system (she also provides instructions for Excel 2008 in Apple), which are already somewhat outdated. Almost everyone I know is on Windows 7 or maybe still on XP, with many of us already beginning to use Windows 8. I did not have any trouble using the instructions in Windows 8/Office 2010, though. Wohlwend’s work did wake me up to some of the display options we might use in Excel, which was helpful.

The new “spin” is useful as far as it goes, and it is a good integration of technology with miscue analysis, which anyone who has used the procedures probably finds welcome, because the sheer amount of information a good miscue analysis provides can be somewhat overwhelming and difficult to interpret and represent. As stated above, I will probably try using Wohlwend’s idea, and will present it at least as an option for my students in literacy methods courses where we do miscue analysis. The only daunting thing for me was my dislike of Excel, which I have always found more fiddly than I liked, and I tend to try to find other ways of displaying data rather than tackling Excel (so much so that I had to review the Excel tutorials before I could even begin to make up a table to attempt the “spider charts” Wohlwend suggests). However, I like the idea of making a graphic representation of the data to clarify and share our findings.

I’ll conclude by expressing my worries about something that may be more important than how we display the data, and that is the background information Wohlwend provides on miscue analysis. My main worry is in the bullet lists on page 111, where Wohlwend lists some “signs” that readers are using the various kinds of cues. All three lists begin with “Makes no miscues.” The fact that this was at the top of each list tells me that Wohlwend thinks it is important to make no miscues, or few miscues, which heightens my concern. I believe most miscue analysis researchers would tell you that everyone miscues, all the time, and that miscues are not always undesirable. In fact, some miscues I have seen actually improve a reader’s meaning construction. For example, I have seen readers insert words that enhance the text and show use of semantic cues. I have also seen readers “edit” text to fit their own dialects, which shows good use of syntactic cues. In both cases, the readers were “miscuing”, but in desirable ways. Miscues have to be looked at qualitatively, always. It is not whether or not we miscue that is the main thing. It is whether we miscue in ways that help us make sense of text or that hinder that sense-making. Finally, the bottom line is always comprehension! I have worked with readers who had large numbers of miscues, but their miscues always made sense and did not disrupt meaning-making. Their retellings showed they “got” the text. I have also worked with readers who made fewer miscues, but they were more serious, involving key text vocabulary and interfered with understanding (e.g., substituting non-words that were graphophonically similar, such as when I observed a child substituting “rellafits” for “relatives” in a reading of Cynthia Rylant’s The Relatives Came). In short, the item “Makes no miscues” should not be on any list involving use of cueing systems, and certainly not at the top of those lists. As I understand miscue analysis, valuing no or few miscues seems like an emphasis on text reading accuracy, which seems inconsistent with what miscue researchers intend. That won’t stop me from utilizing Wohlwend’s “spider chart” approach as an option and sharing that with others, but I will need to share my reservations on this point before recommending the article.