[Technology 589] Re: Cognitive Rescaling

Archived Content Disclaimer

This page contains archived content from a LINCS email discussion list that closed in 2012. This content is not updated as part of LINCS’ ongoing website maintenance, and hyperlinks may be broken.

Denis Anson danson at misericordia.edu
Fri Sep 22 13:55:58 EDT 2006


Dave,

This is, indeed, a potentially powerful means of providing a summary of
a long and complex document. And, for a well-formed document, it works
quite well.

The Autosummary tool appears to work by selecting topic sentences from
paragraphs, and then identifying supporting sentences. As such, it
works pretty well for declarative documents. It doesn't work very well
for less well structured documents. I have seen Word produce summaries
that are a mish-mash of sentences that make little sense.

There are two important limitations to this approach to scaling
documents. While the overall length of the document changes, the
inherent readability does not. Readability is a function of number of
syllables per word, words per sentence, and sentences per paragraph.
While autosummarize may reduce the number of sentences per paragraph,
the readability of individual sentences remains a function of word
selection and sentence length, and these do not change. Hence, a person
who is having difficulty with a document because they cannot parse the
sentences to obtain the underlying thought remains unable to parse the
sentence.
The second weakness actually exacerbates the first. A summary, by its
nature, discards much of the detail of a document. And, as has been
observed, the devil is in the details. In a complex document (and even
in a well formed document), a paragraph begins with a topic sentence
which may not be entirely clear to the reader. As the paragraph
continues, the supporting sentences provide the nuances to the topic
sentence that clarify and explain the thoughts of the author. When
using a summary, the explanation of the topic sentence may be lost, so
that the reader must attempt to interpret, then accept or reject the
topic sentence as a bald insertion, bereft of the supporting language.
For a person who is already familiar with the arguments, this may help
absorb the document more quickly. For a reader who has difficulty
interpreting language, and the particular language of the document, the
summary may, in fact, be less understandable than the longer document.

To assert that the summary provides the same content as the full
document would be to assert that Cliff's Notes for Shakespeare free the
reader from the tedium of reading the entire play. The flow of ideas
and of language, which is what the play is "about" is simply absent from
the summary.

Denis Anson, MS, OTR
Director of Research and Development
Assistive Technology Research Institute
College Misericordia
301 Lake St.
Dallas, PA 18612
Phone: 570-674-6413
Fax: 570-674-8054



Dave L Edyburn wrote:

> In a previous posting I discussed the conceptual foundations associated

> with digital sliders or equalizers. The issues involve dynamically

> adjusting the difficulty of a task along any one of several dimensions

> that make corresponding changes in a task or the supports provided to

> complete a task. This posting seeks to bridge that theory with a practical

> experience in using a slider that has already been built into software

> that is on your desktop.

>

> The Need for Cognitive Rescaling

>

> As a teacher of students with disabilities, I've dreamed of tools that

> would enable me to take an assigned reading, manipulate the text in ways

> that alter (increase or decrease) the cognitive challenge involved in

> understanding the information, and produce cognitively rescaled versions

> of the original assigned reading that I could distribute to individual

> students. For many years, these ideas were simply dreams.

>

> Recently I discovered several software tools that now make it possible to

> cognitively rescale text-based information. Simply. In just seconds. Sound

> too good to be true? Read on.

>

> Getting Started

>

> Before we can cognitive rescale text-based information, we must ascertain

> the source and format of the original text. Obtaining a digital version of

> the text is a prerequisite for cognitive rescaling.

>

> 1. Information Sources and Formats

>

> When students are required to learn from text, any number of sources and

> formats may be involved: (a) web pages, (b) textbooks, (c) teacher-created

> handouts, (d) periodicals, or (e) reference books. Assistive technology

> and universal design interventions must recognize the array of information

> sources and formats commonly used in classrooms.

>

> 2. Digital Text

>

> Regardless of the source of the original text, to make the curriculum

> accessible, it is essential to obtain the information in a digital format.

> In the case of the teacher-created handout, it should be relatively easy

> to obtain the word processing file that was used to create the document.

> Obviously, web sites, some periodicals, and online ready-reference

> documents are already in digital format. Many literature texts

> (http://promo.net/pg/) and historical documents (http://www.loc.gov) are

> in the public domain and the full-text can be downloaded.

>

> Persuading commercial publishers of the necessity of digital text for

> struggling readers is an on-going effort. Ideally, textbook publishers

> will provide digital copies of reading materials as part of the textbook

> adoption process. A new development seeking to exchange digital texts is

> an extremely promising initiative (http://www.bookshare.org). However,

> without systematic support to access to digital text, print documents must

> be scanned using optical character recognition (OCR) software in order to

> obtain a digital version of the information.

>

> Once we have a digital version of a reading assignment, we can utilize one

> or more cognitive rescaling strategies: executive summaries and

> rebus-enhanced texts.

>

> Step-by-Step Directions

>

> Below is an abbreviated list of directions for creating executive

> summaries using the AutoSummary Feature in Microsoft Word. To download an

> article reprint (Edyburn, 2002) with step-by-step directions for two

> cognitive rescaling strategies and illustrations, visit: http://

> www.uwm.edu/~edyburn/Cognitiverescaling.pdf

>

> Cognitive Rescaling Text Using the AutoSummary Feature in Microsoft Word

>

> Microsoft Word includes a feature for creating executive summaries. As a

> result, this tool can be used to create summary versions of of any text

> document. The following description will highlight step-by-step directions

> for accessing and using this tool. Note: The following procedures work on

> both Windows and Macintosh computers and on all versions of Word (no

> matter how old your copy is!).

>

> 1. To begin, you'll need a text document. This can be information you have

> copied and pasted, a document that you typed or scanned into a word

> processor, or an etext that you have downloaded. Personally, I like to go

> to Windows to the Universe (http://www.windows.ucar.edu) and copy the text

> on an entire topic at the advanced level. You can go to Wikipedia or any

> other text-based web page to copy text for this example.

>

> 2. Open Microsoft Word and paste the desired text into a new document

> file. Note: You may wish to Select All and adjust the size, color, and

> font of your text before proceeding.

>

> 3. Locate the Tool menu at the top right center of your screen, and

> select: AutoSummarize. Note: In Windows XP you may need to expand the

> pop-up window since items are prioritized on their use; when you do this,

> AutoSummary will be added to the menu.

>

> 4. A window will appear in the center of the screen when AutoSummarize is

> selected. The first section, "Type of Summary," offers four possibilities

> for viewing your executive summary. The first option provides a slider for

> users to interact with the percentage of the summary and view highlighted

> information that will be included (non-highlighted text will be left out

> of the executive summary). The second option creates a new document with

> the summary. The third option inserts the summary at the beginning of the

> document. The fourth option hides the text of the original document and

> includes the summary. The second section, RLength of SummaryS allows user

> to select three types of lengths for their executive summary:

> predetermined number of words (100 or less, 500 or less), predetermined

> number of sentences (10 or 20), or a % of the original text (10%, 25%,

> 50%, 75%). Select the first type of summary and using the pop-up, select a

> 10% summary, then click ok.

>

> Note: If you get an error message that indicates there is no text to

> summarize, this probably means that the text you copied included html code

> and that when you pasted the text the information came into Word as a

> graphic. To remedy this situation, open a new document. Then, use the Edit

> Menu to select: Paste Special. Then select the option to paste text only.

> This will remove the html code. Now, repeat Step 4 and things should work

> normally for Step 5.

>

> 5. You will now see your text with the summary elements highlighted in

> yellow. Notice the floating palette with a real life slider on your

> screen. Experimenting with the slider will visually reveal how increasing

> or decreasing the percentage of the summary will add or delete text.

> Continuing interacting with your text until you get a summary that you

> feel is adequate.

>

> 6. After you have interacted with the slider and determined the optimal

> size for the executive summary, return to the Tool menu and again select

> AutoSummarize and then select the second option: create a new document and

> put the summary there. After you click ok, your executive summary will now

> appear in a new file. Save this file containing your executive summary of

> the text or print and distribute to individual students as needed. Check

> your watch: this entire process has only taken a few minutes!

>

> Implications

>

> --The concept of a slider is not just theory. A real-life working slider

> has been built into Microsoft Word.

>

> --Anyone can now make executive summaries of text information that they

> are not able or not willing to read (e.g., very long documents).

>

> --Consider how this intervention may work in conjunction with others we

> have discussed this week. Obtain a digital copy of a reading assignment,

> copy and paste it into Word, use AutoSummary to make a summary of the most

> important information in a document at a % that offers optimal challenge

> for a reader ("just right," not too hard [100%] but not too easy [10%],

> and then copy and paste the summary into a text to speech program like

> ReadPlease (http://www.readplease.com) so the student can listen to the

> information that they could not have read independently, or paste into

> Babelfish (http://babelfish.altavista.com) for second language

> translation.

>

> --Consider how teaching our adult struggling readers to cognitively

> rescale text might engage them with print in ways they have never

> experienced in their life. Consider how three forms of technology converge

> within this powerful application: instructional technology offers digital

> text; universal design for learning feature of AutoSummarize in Word was

> designed to help users of all ability levels by manipulating text to get a

> summary that is "just right" for their immediate needs, and assistive

> technology in the form of text-to-speech provides another level of access

> and support for those that need it.

>

> --Our subsequent research on cognitive rescaling has demonstrated that

> this technique does not appear to work on fiction. However, it does seem

> to be an ideal text modification strategy for expository text (e.g.,

> textbooks, reference materials). We have also learned more about the

> algorithm that is used such that the readability of the text is not

> significanly altered but rather the length of the text is reduced.

>

> --Beyond the initial publication on the discovery of cognitive rescaling

> interventions (Edyburn, 2002), research on classroom use of cognitive

> rescaling is limited. One interesting study of its use by a secondary

> student with behavior and learning disabilities has been reported by

> Mortensen (2002) who found significant gains in writing after using

> cognitive rescaling as a means of gathering information for research

> reports.

>

> --Anecdotal evidence suggests that the power of the slider allows many

> struggling readers a level of control they have never experienced when

> interacting with text. Whereas we thought we discovered a tool for

> teachers, the real value appears to be to give this tool directly to

> students. Further, while we anticipated that they would create 10%

> summaries, we routinely found they exceeded our expectations by making

> summaries that were longer than the minimum because they want to know more

> about the topic.

>

> Concluding Thoughts

>

> To-date, most efforts to make the general education curriculum accessible

> have focused on enabling students to listen to information when they are

> unable to read it (i.e., text-to-speech). The traditional method of

> creating cognitively simplified versions of text documents, like textbook

> chapters, has required teachers to first read an assigned reading and then

> create summaries in their own words. Obviously, the traditional process

> has been labor and time intensive given the lack of powerful tools.

>

> The cognitive rescaling example described above suggests ways in which

> technology can be used to automate and facilitate the process of altering

> the cognitive difficulty of text so that information can be made

> accessible for students who lack grade-level reading skills. These types

> of interventions have significant potential for expanding the palette of

> tools available for responding to academic diversity.

>

> The benefit of cognitive rescaling for teachers and curriculum developers

> is that it is efficient to make instructional materials which anticipate

> the diverse cognitive abilities found in every classroom. The

> metacognitive value for students, as they are taught to manipulate text

> using cognitive rescaling strategies, needs further study but appears

> promising. The value and full potential of the cognitive rescaling

> strategies outlined in this article remains to be demonstrated through

> research and classroom application.

>

> References

>

> Edyburn, D.L. (2002). Cognitive rescaling strategies: Interventions that

> alter the cognitive accessibility of text. Closing the Gap , April/May, 1,

> 10-11, 21. Reprint available online: http://

> www.uwm.edu/~edyburn/Cognitiverescaling.pdf

>

> Mortensen, S. (2002). Action research on cognitive rescaling. Journal of

> Special Education Technology, 17(4), 53-58. Available online:

> http://jset.unlv.edu/17.4/asseds/edyburn.html

>

>

> ----------------------------------------------------

> National Institute for Literacy

> Technology and Literacy mailing list

> Technology at nifl.gov

> To unsubscribe or change your subscription settings, please go to http://www.nifl.gov/mailman/listinfo/technology

>

>

>