20 likes | 118 Views
Fig. 1. display to read barcodes. barcode. microphone. 4 buttons to start replay of spoken language Independently of barcodes. speech recording button. volume control. power supply. Communication Disorders Program – University of Virginia Woodrow Wilson Rehabilitation Center.
E N D
Fig. 1 display to read barcodes barcode microphone 4 buttons to start replay of spoken language Independently of barcodes speech recording button volume control power supply Communication Disorders Program – University of Virginia Woodrow Wilson Rehabilitation Center Experimental Aspects of a Speech-Generating Barcode Reader Among Non-Disabled Students A. England1, T. Haley 1 , V. Maddocks 1, F. Loncke 1 & L. Meyer 2 1 Communication Disorders Program – University of Virginia, 2205 Fontaine Avenue, Charlottesville, VA 22903 2 Communication Services, Woodrow Wilson Rehabilitation Center, Fishersville, VA 22939 The Electronic Language Trainer B.A.Bar Theoretical importance Within AAC, we need devices that approximate the ease and directness of natural speech, and allow face-to-face interaction with the environment. A hand-held, portable design may allow an individual to integrate use of the communication device more readily in daily situations. B.A.Bar is an electronic device which is able to record, store and replay spoken language. Words, phrases or sentences are digitally recorded and linked to a barcode. When scanning a barcode, previously recorded language is replayed. Clinical and educational importance This device may have the potential to combine the advantages of (1) aided, (2) unaided, and (3) speech generating AAC devices. The mobility and the portability of the device may make it possible for a person to build up and extend a lexicon while directly interacting with the environment. The use of the device is a variant of pointing to objects, pictures or other meaningful elements in the direct environment of the person. Fig. 2 Underlying hypotheses • Questions • What processes are facilitated by a hand-held speech generating device? • Is device-generated speech equivalent to naturally generated speech? • What is the relationship between labeling, pointing, and lexical accessing? CONCEPTUALIZER discourse model message situation knowledge generator encyclopedia etc. monitoring preverbal message parsed speech A theoretical approach Most psycholinguistic models (like the Levelt model) assume that the components of natural speech generation include (1) lexical selection and decision (which words to choose from; which word do I actually choose), and (2) a phonetic-articulatory executionary program to actually speak the words. Direct Access to lexicon? FORMULATOR SPEECH- COMPREHENSION grammatical SYSTEM encoding LEXICON lemma s surface structure forms Easier lexical selection? • It is hypothesized that this device might have the potential to: • Facilitate transition from experience to lexical access • Minimize lexical access efforts • Bypass phonetic planning and speech articulation phonological encoding Bypassing phonetic planning and articulation? phonetic plan phonetic string (internal speech) AUDITION ARTICULATOR overt speech Learnability questions of a hand-held speech-generating device: Allowing insufficient time for a person and his/her environment to become familiar with a communication device increases the risk that the device will be abandoned. How easy is it to learn new internal and external programming routines for lexical and sentence utterance? References Blischak, D. M., L. J. Lombardino, et al. (2003). "Use of speech-generating devices: In support of natural speech." Aac: Augmentative & Alternative Communication 19(1): 29-35. Jaswal, V. K. and E. M. Markman (2003). "The relative strengths of indirect and direct word learning." Developmental Psychology 39(4): 745-760. Levelt, W. J. M. "Relations between speech production and speech perception: Some behavioral and neurological observations." Dupoux, Emmanuel (Ed) (2001). Language: and cognitive development: Essays in honor of Jacques Mehler. Levelt, W. J. M. (1993). Speaking: From intention to articulation. Cambridge,MA, MIT Press. Liberman, A. M. and I. G. Mattingly (1985). "The motor theory of speech perception revised." Cognition 21: 1-36. McNaughton, S. and P. Lindsay (1995). "Approaching literacy with AAC graphics." Augmentative & Alternative Communication 11(4): 212-228. Meyer, L., Loncke, F., Maddocks, V., Haley, T., & England, A. (2004). A speech-generating barcode reader: Clinical AAC and Research applications. Poster presented at the Convention of the American Speech, Language, and Hearing Convention. Philadelphia.Schlosser, R. W. and D. M. Blischak (2001). "Is There a Role for Speech Output in Interventions for Persons with Autism? A Review." Focus on Autism & Other Developmental Disabilities 16(3): 170-78. Contact:ame5f@virginia.edu
average learning on a first trial by college students 12 10 8 remembered number of pseudowords 6 4 2 0 condition 1 (silent condition 2 (natural condition 3 (barcode reading) speech) speech) conditions Communication Disorders Program – University of Virginia Woodrow Wilson Rehabilitation Center Pilot Lexical Learning study 1 Question Will learning of new lexical items (pseudo-words) be facilitated if a device speaks them for you? Will you immediately integrate the speaking device into your word storage and word retrieval strategies? Several factors may explain the low performance with the speech-generating device: (1) as college students, the participants had already developed learning preferences that do not require speech, (2) the use of the device was new for the students and distracted cognitive effort from the actual memorization process, (3) the use of the device implies a motor component (pointing the device to the barcode) which may have interfered with memory process. When used within a (re)habilitative or educational context, the motor-operational and other cognitive skills involved in language processing may prevent instantaneous restructuring of lexical learning strategies. Methods and materials 20 pseudo-words matched with similar degree of phonological difficulty. Each word was matched with an abstract geometric shape.27 college students were asked to memorize a list of pseudowords -one group learns the words while reading silently (9 participants) -one group is encouraged to learn the words while saying the words aloud (9 participants) - one group uses the barcode reading speaking device (8 participants). Results The participants performed best (and preferred) learning the words through silent reading. Memorization with natural speech mediation was slightly less performant than silent reading. Mediation of artificial speech generated through a device decreased learning. Discussion These first results indicate that substituting internal or external speech with an artificial speech generating device does not happen automatically and immediately. From the post-test interviews, it appeared that most of the students had concentrated on associative memory strategies that did not call for the use of overt speech. Pilot Lexical Learning Study 2 • Learning with the barcode reader • 20 undergraduate college students • 3 conditions: • Pseudo-word lists through natural speech • Pseudo-word lists through listening to the speech-generating barcode reader (operated by experimenter) • Pseudo-word lists through listening to the speech-generating barcode reader operated by participant Statistical Analysis: Trial 1: F[2,17] = .17 p < 1 Trial 2: F[2,17] = .28 p < 1 Trial 3: F[2,17] = .21 p < 1 Trial 4: F[2,17] = .41 p < 1 Trial 5: F[2,17] = 1.00 p < 1 • Underlying hypotheses • Participants manipulating the speaking barcode reader will initially be at a disadvantage as they have to invest cognitive and motor effort in learning to operate the device • Participants listening to (but not manipulating) the speech-generating barcode reader will do better than those operating the device • Participants using natural speech (without external device) will have highest initial outcomes • All groups will show learning gains over the different trials • As learning progresses the difference between the groups will decrease. Method for statistical analysis A repeated measures ANOVA was used (for each five probes). There were no significant differences between the three groups. Future research These studies have focused on learning by non-disabled young adult participants. The rationale for running the study with these groups of participants was that it would help to formulate hypotheses about (1) differences in learning strategies that could occur depending on modifications of the learning conditions (natural speech vs. speech-generating barcode reader, operating vs. non-operating the speech-generating barcode reader), (2) the learnability of using the speech generating barcode reader, and (3) the rate at which the use of the device would become fluent. Many questions are still unanswered. One further line of research will be isolating the effect of internal speech on non-disabled users of a speech-generating communication device. A second line of research will be an analysis of (1) the use of different modalities in communication with the speech- generating device (natural speech, barcode reading, gestures, manual signs) [research question: which communicative and informative functions are assumed by the barcode reader?], and (2) learning direct and indirect requesting through the barcode reader. Results and Discussion This study has helped to analyze the components that play a role in learning new lexical elements. Particularly, the study attempted to identify differing strategies involved in learning through a speech generating barcode reader compared with the use of natural speech as mediator. Activating natural speech is considered to increase the likelihood that the user stores and retrieves lexical elements through concomitant internal speech. But will this also be the case with the use of artificial speech? If effects on memorization are to be taken as indicators for differences between the two conditions, the present study appears to indicate that users quickly are able to use the speech generating barcode reader as an equivalent of natural speech. Additionally, no difference has been apparent between the conditions of operating the barcode reader (which could be considered the equivalent of activating speech articulation) and listening to the barcode reader (supposedly the equivalent of accessing the word through a speech-recognition system). For college students, using the speech-generating barcode reader as mediator does not significantly affect the rate of word learning. The results of the first study were not confirmed.