Monday, August 4, 2014

Don't just assist; instruct, III

I've been working on my part of a forthcoming book on linguistic technologies for people with language impairments, and have just finished writing up a section on assistive technologies. There a lot of new devices in the wings that raise various new concerns, and I thought I'd share some of what I've written here:

Currently, the most typical assistive communication devices for classroom communication involve buttons designating certain words and phrases which the user pushes to communicate certain basic thoughts—typically basic desires:



More ideal would be to go beyond preprogrammed buttons without making the device so open-ended as to offer little or no assistance--as would be the case if you replaced an array of word buttons with an array of letters, as in a standard alphanumeric keyboard. Furthermore, to be truly assistive the device would have to somehow know what the child wants to say before he or she starts saying it.

There is, in fact, one type of software that actually approaches this capability. Known as word cuing software, these are programs that one can plug into a word processing program, either on a desktop or a tablet, and that anticipate the one’s likely next word from the words and letters that one has already typed in. These programs go far beyond the typical autocomplete, which waits until the word is almost complete and determined, and, accordingly, only gives you one choice.

Here’s are two screenshots from one popular program--Read&Write Gold:




In these programs, the word choices in the pop-up window are selected based on what is syntactically and morphologically appropriate given what the user has typed so far, and also what is likely given general statistics about word occurrence as well as the individual statistics that it collects about particular users and their word selection habits.

Some of these programs have additional features: they watch for common spelling errors, providing a pop-up list of alternative options, or they offer examples of how a given word might be used. Some allow users to select particular topics--topics as specific as Polish architecture--and then invokes a dictionary for that topic, and presents word choices (nouns, verbs, adjectives, etc.) specific to this topic. And, to help students who have no idea what to write about on a particular topic, there are a word banks (with words sized by frequency) to help you get started:



For language-impaired students, word prediction software is a powerful tool. In particular, it can help students who struggle with grammar produce grammatical sentences. For example, a student who types “I want” will see the word “to” as one of his choices, which steers him away from errors like “I want go.” A student who types a plural subject like “my friends” will only see verbs in the plural form (for example, “are” rather than “is”). A student who types the article “an” will only see noun choices that begin with vowels—as we see in the first image from ReadWrite Gold. Or, as we see in the second image, a student who types in “an example” will see, as his or her first choice, the preposition that most often follows the word example “example” (“of”).

These programs are powerful, but they are also problematic. In general, the more efficacious the assistive technology, the more it potentially reduces the urgency of teaching the skills that are being assisted. It is essential that assistive technology be only treated as such—namely, as assistive—and not used as a reason to adjust teaching priorities and eschew necessary remediation.

When it comes to autistic children in particular, another concern is the extent to which technology deprives them of the face-to-face social interactions on which they may be especially dependent for their social development. Too often, whether the students are autistic or not, one finds classrooms and other settings in which students are mostly looking at and interacting with screens rather than with one another.

The proliferation of technology in the classroom heightens both of these concerns. So do the unprecedented pressures that today’s teachers are under. Most language impaired students are included in regular classes with same-aged peers, and their teachers are increasingly under pressure to teach to the new Common Core State Standards and tests. These standards set high expectations for reading and writing and take a one-size-fits all approach to students at a given grade level. In light of this, fewer and fewer teachers, even special education teachers, feel that they have time to remediate basic skills--the more so when assistive devices make remediation seem increasingly less urgent.

Remediation and accommodation shouldn't be at odds ideally, they go hand in hand. The ultimate goal, after all, is to optimize the learning environment such that students reach their potential, and, ultimately, are liberated from assistive technology to the largest extent possible.

No comments: