Testing Reading and Writing of Students with Speech and Physical Challenges
by Hester Brooks, Ed.D.
My first in-depth encounter with a youngster with severe speech and physical impairment came in 1988 when I met Ricky (not his real name.) He was four years old, and already an experienced AAC (Augmentative and Alternative Communication) user, starting with a Light Talker when he was three. Ricky presented as a bright and sociable child. He was a client at CAST (Center for Applied and Special Technology) in Peabody, Massachusetts, where I had just begun working. CAST was a recently formed organization with the primary mission of making education more accessible to individuals with disabilities by applying evolving technology. Computers had begun opening up opportunities for people with physical, sensory and intellectual deficits to participate more fully in academic settings.)
Ricky's mobility was severely limited. He could voluntarily move only his eyes and chin, nothing else. He used his chin to operate a lever affixed to an eye glasses frame. The switch activated cells on his Light Talker. He quickly introduced me to his method for saying yes and no/yes was both eyes firmly shut, no was looking upward to the left. Using the Light Talker, he then demonstrated his sense of humor, asking "Why was six afraid of seven?" When I said I didn't know, he beamed as he tapped out the answer, "Because seven ate nine!" In an instant, I was captivated by Ricky's engaging personality.
At the time, children with significant special needs were not ordinarily enrolled in regular education classes. Ricky had spent the two past years in an integrated preschool, and his teachers anticipated that he would be ready to attend kindergarten in a mainstream classroom the following September. His local school district agreed on the condition that he could demonstrate the intellectual ability to do regular schoolwork?a condition that would not apply today. Testing Ricky presented a challenge: how to test student who could not speak, write, or hold a pencil. On the computer, an Adaptive Firmware Card (AFC) was used to modify some conventional software so choices could be activated with the single switch. Using this method, Ricky played computer gam'es such as Old MacDonald and Pictures, Letters and Sounds. We knew from this that he could reliably identify letter names and some letter sounds.
David Rose, one of the founders of CAST, determined a way for Ricky to independently take a standardized test. Using the AFC, a four quadrant display was set up on an Apple IIe computer in such a way that the quadrants were highlighted in sequence. Using the Peabody Individual Achievement Test (PIAT), transparencies were made of the test items. Velcro was used to fasten the transparencies, one item at a time, over the computer monitor. The examiner gave the user a prompt such as, "which of these pictures shows a word that begins with the 'buh' sound?" and as the choices highlighted, Ricky responded by clicking his chin switch. In this way he identified letter names and initial letter sounds, numbers and basic numeric concepts. Based on this technique, Ricky showed advanced knowledge of letter recognition and initial sounds as well as solid understanding of basic numeric concepts compared to age expectations.
When they received the test results, the school staff made plans for Ricky to enter kindergarten in September. The plans included hiring a 1:1 teacher to support him in the classroom as well as signing a contract with CAST to provide assistive technology services. His teacher spent time at CAST prior to the beginning of the term to familiarize herself with Ricky's technology. Thus he began school with a substantial support network. In addition to his parents, his classroom and 1 on 1 teachers and a staff member from CAST, there was a speech and language therapist experienced in using AAC devices, a physical therapist to continue working on range of motion, and a nurse who attended to his personal needs and feeding. Behind the scenes, a gifted and creative CAST programmer wrote programs using HyperCard, to provide him with computer access to reading and writing activities. All of these resources were provided to enable Ricky to attend school in mainstream classes, but in addition, the undertaking became a research project in the development of technology to enable students to have access to curriculum, a topic that CAST continues to pursue today.
We all felt that literacy was of key importance, not only because it is central to every child's education, but because, as an AAC user, Ricky would be able to communicate more fully using text than would be possible with a picture or symbol-based system. Although he sometimes used his Light Talker creatively, he rarely initiated interactions using the device, preferring eye gaze, facial expression and vocalizations. His attendants, teachers and classmates often played 'Twenty Questions,' to extract meaning.
Ricky's literacy development progressed in fits and starts. He continued to demonstrate quick understanding of concepts presented orally and he loved attending school. Over the course of his schooling, it was important to assess Ricky's progress, both to adjust his curriculum according to his needs and to evaluate the effectiveness of the means we were devising. This presented an on-going challenge. Like many youngsters with significant deficits, he compensated by developing other skills. While Ricky was not great at reading text, he was a superb reader of people. He watched his surroundings intently and picked up the slightest cue to respond in the desired way. Reading instruction in his primary grades was for the most part 'whole language' as opposed to being phonetically structured. This meant exposure to a wide variety of texts, with repetitive patterns, choral reading and emphasis on meaning rather than precise decoding. Although he couldn't speak, as Ricky sat with his classmates, his expression indicated that he was following right along. One spring day, the children read aloud the story of fireman Mike, going through the motions as they heard the alarm go off, jumped into their clothes, slid down the pole, stamped across the floor, leapt into the fire truck and raced off to the fire. Ricky's partner held his hand as he read the text and went through the actions, Ricky extended his arms and legs in excitement and as the fire truck siren shrilled, Ricky's wail could be heard along with the rest.
Early in his first grade year, Ricky was tested using the Peabody Picture Vocabulary Test (Revised) [PPVT-R] and the Test of Auditory Comprehension of Language [TACL].
Both tests were prepared on a Macintosh Computer using HyperCard. Using his chin switch, Ricky could activate a scanned sequence that highlighted the choices as well as buttons to choose 'go next' and 'go back.' When he clicked his switch, the highlight flashed to indicate his choice. Another click and the scan resumed. We established a protocol that if he wanted to change his selection, only the final choice would be recorded. The examiner recorded his responses, but observed the usual testing protocols prescribed in the manual insofar as possible, differing only in terms of mode of presentation.
I am describing this procedure in some detail to emphasize the importance of finding a valid way to evaluate independent mastery of subjects for a student with SSPI As we have learned from experience with Facilitated Communication, it is easy to be misled by the appearance of understanding when there is close contact between the subject and the examiner or instructor. [For those readers unfamiliar with Facilitated Communication, a source for a brief explanation can be found at http://en.wikipedia.org/wiki/Facilitated_communication]
Results on the PPVT-R and TACL suggested that Ricky had excellent auditory vocabulary and language comprehension skills. Although it is important to note that the procedure differed from the usual administration, there was no examiner intervention, and based on the conventional scoring procedure, Ricky's score fell in the high eighties on the TACL and the ninetieth percentile on the PPVT-R. This was consistent with our observations, and encouraged us to think that, given the tools he needed to access curriculum, Ricky would make normal academic progress. At the end of his first grade year, judging by Ricky's attentiveness and responsiveness, we assumed that he was doing well. In order to confirm this, I converted the Stanford Diagnostic Reading Text-Red Level (SDRT) to our HyperCard format. I chose the Stanford (the Red level is designed for students in the first and second grades, and for third graders who appear to be lagging) because it is a norm-referenced multiple-choice text that assesses vocabulary, word attack skills and reading comprehension. Also, it did NOT include oral reading, a common way to evaluate a beginning reader's skill. Using the HyperCard design , one item was presented per screen. Sitting beside Ricky, I recorded his responses as he went through each subtest, noting how many scan cycles he went through before responding, and whether he changed any answers. I used a code to avoid extensive writing, because I had already learned that Ricky was acutely aware of my presence, and if I did much scribbling, he was likely to think he'd made an error and would change his answer?that familiar bugaboo, reading the examiner instead of the question. His body language was informative. When he was confident of an answer, his body would tense as soon is the item was displayed as he prepared to click his switch. At other times, he allowed the highlight to go through several cycles while he pondered. Once, when the word 'embroider' came up, he rolled his eyes and made what we referred to as his 'squirrel face.' It was obvious that he had no idea what the word meant. When I mentioned this to his mother later, she said it was certainly a word he would never learn from her, as she hated sewing.
When all subtests were complete (a process that took three sessions) the result was a completely independent evaluation of Ricky's reading status. These results were very informative. His auditory vocabulary continued to be very strong. (In this subtest, the examiner reads aloud: "To linger is to wait, hurry or agree," as the words 'wait,' 'hurry,' and 'agree' are highlighted on screen. Thus, the subtest actually requires no reading.) He had no difficulty identifying letters in upper or lower case. He was able to match pictures to words by initial consonant, except when all the choices had the same initial consonant (for example 'bell,' ball' and 'bull'). Then his performance fell to chance level. Finally, he did not seem able to decode even the simplest sentence. To say we were surprised is an understatement. During the course of the school year, he had given every appearance of understanding text, but he was actually following the cues emanating from classmates and teachers, while still using the same strategy he had learned in preschool?initial letter-sound recognition instead of actual word identification.
Accordingly, we began to construct a more individualized curriculum, with emphasis on phonemic awareness and phonetic rules. Gradually Ricky's reading improved. Over the course of the next several years, Ricky and I continued to meet, and I continued to evaluate his progress. The method was basically the same, though we moved from using HyperCard to converting the tests to a Speaking Dynamically format. His auditory comprehension continued to outstrip his ability to decode, but he did become a reader and writer. For AAC, he transitioned from the Light Talker to Liberator, and eventually to an entirely text based system using Words +. He continued to depend primarily on eye gaze and facial expression for casual communication, but as his need to convey meaning became more complex and less context-bound, he used his AAC device more frequently.
We encountered new issues as reading passages advanced from simple sentences to paragraphs, and eventually to formats that included extensive text, tables and charts. Crowding of text on the page and visual tracking presented problems. By the time Ricky was in high school, a dedicated software engineer adapted Kurzweil reading software to a single switch format for his use. In a trial, I gave Ricky two versions of the Stanford (by this time he had progressed through the levels from Red to Green to Brown to Blue for grades 9 through 12). We used only the reading comprehension subtest for this trial, one version presented in Speaking Dynamically, which did not include a text-to-speech option, and the other in Kurzweil, where Ricky could hear the text read aloud. In the silent reading version, it was clear that Ricky was reading the text, but with low average comprehension according to the standardized norms. With Kurzweil, he controlled the mode, and I acted as observer using our usual protocol. When he wasn't sure of the meaning, he chose the 'read aloud' option. His final comprehension score was in the 88th percentile.
What I have written here is a brief summary of some of the methods used to support Ricky in his education. It was a collaborative effort as the tools evolved, and, by trial and error, we applied them. Assessment was only a small part of the intervention, and, while it was not the most important part, it mattered to us because it enabled us to understand what techniques were working. We are indebted to Ricky for all he helped us to learn as he himself was learning.
Sign in to participate in the discussions.