The effects of abstraction levels, word position, and voice on error productions of apraxic speakers
Investigators disagree as to whether apraxia is an articulatory- or linguistically-based disorder. Since objective data supporting a language-based disorder are insufficient, the present study investigated the effects of abstraction levels, word position, and voice on error productions of apraxic speakers during a spontaneous speech task.
Ten adults, diagnosed by a speech pathologist as possessing verbal apraxia, served as subjects. Additional selection criteria included: (1) brain trauma occurring at least three months prior to this study; (2) normal hearing acuity in the better ear at 500, 1000, and 2000 Hz (ISO, 1964); (3) no associated problems (i.e., dementia or mental retardation); and (4) no significant receptive difficulties as demonstrated by their ability to correctly respond to 10 selected items from the Revised Token Test (McNeil, 1973).
Sixty stimulus sentences were constructed so that five sentences each utilized the three variables (i.e., abstraction levels, word position, and voice) under study in such a manner that the stimulus word appeared in the sentence according to one of the following conditions: (1) noun phrase one - least abstract noun - active voice, (2) noun phrase one - least abstract noun - passive voice, (3) noun phrase two - least abstract noun active voice, (4) noun phrase two - least abstract noun - passive voice, (5) noun phrase one - most abstract noun - active voice, (6) noun phrase one - most abstract noun - passive voice, (7) noun phrase two most abstract noun - active voice, (8) noun phrase two most abstract noun - passive voice, (9) noun phrase one nonsense word - active voice, (10) noun phrase one nonsense word - passive voice, (1l) noun phrase two- nonsense word - active voice, or (12) noun phrase two - nonsense word - passive voice. Each stimulus sentence was read twice to the subject. For the first reading, the sentence was read in its entirety; during the repetition of the sentence, the stimulus word was omitted. The subjects' task was to provide orally the missing word. Total errors for all subjects were tabulated and analyzed by a treatment x treatment x treatment x subject analysis of variance.
Subjects x position, the only significant two-or three-way interaction, suggested that performances of apraxic speakers were differentially effected by word position in a sentence. Significantly more errors were produced when stimuli appeared in the first rather than in the second noun phrase of a sentence. Apraxic speakers made significantly more errors on nonsense words, fewer errors on most abstract nouns, and the least number of errors on least abstract nouns. A significantly greater number of the apraxics' errors were produced on words in passive, as opposed to active, sentences.
The findings suggest that apraxia may involve an impairment of linguistic processing such that factors such as memory, storage and retrieval, and encoding processes may effect the errors produced by apraxics. The results also indicate that impaired phoneme production may involve several aspects of linguistic processing, rather than an impairment at a specific level.
Thesis76.H273.pdf
3.15 MB
Adobe PDF
e0a12612a4f8ca8966de11cd49e40910