This study investigates the relationship between phonological reading processing and poor spelling. The hypothesis investigated is the following: individuals who utilize the phonological pathway (speech encoding) in word comprehension tasks store words in phonetic representations and therefore are likely to be subjects with poor orthographic pricing abilities.
Results from a naming latency experiment investigation that there is something characteristic about the reading processes applied to misspelled and non-words that creates an effect of slow naming times for subjects with poor spelling. The study identified a significant difference in naming latency for poor spellers and good spellers between the non-word and misspelled word classes.
Authors:
Laurna Compagne, Ph.D.
Sonia Iverson
Introduction
Comprehension of a written word, lexical encoding, requires that the orthographic representation of the word be encoded in such a way that the meaning of the word can be accessed in the reader's internal dictionary, the lexicon. Presently, it is believed that the asynchronous process of looking-up words in our internal dictionary (lexical access), can be modeled by a schema involving two quasi-parallel pathways: the visual and the phonological encoding pathways. Although it has been proven by linguist, T.G. Bower, that skilled adult readers normally utilize the more direct, visual based pathway (Coltheart, 1978; Banks, Oka & Sugarman, 1981). It is generally accepted that the phonological code provides an alternate, concurrent pathway to lexical access (Jorm & Share, 1983) when no shadowing tasks are required.
**** Insert Citation Here ****
Because the phonetic representations of English words regularly do not constrain unique orthographic representations (Just & Carpenter (1987) p. 297), phonological readers are less likely to generate the correct orthographic representation, i.e. spelling, of a word when given a word production task. This study investigates the relationship between phonological processing and spelling ability.
Hypothesis: Individuals who utilize the phonological pathway (speech recoding) in word comprehension tasks (lexical access) store words in phonetic representations and therefore are likely to be poor spellers.
Neuroscientific data regarding dyslexia support the existence of two distinct pathways in reading (Kolb & Whishaw, (1980). p 528). The term "dyslexia" refers to several distinct reading problems. Classification of the different types of dyslexia focus on the distinction between a dysfunction in visual versus the verbal aspects of reading (Just & Carpenter p 367).
Methods
The paradigm used in this experiment is a word naming task. This task requires the subject to articulate words, one after the other, immediately after perceiving them on a computer screen . The experiment compares the performance of two samples of individuals, good and poor spellers, on a word naming task.
Subjects: Subjects include twenty six UCSD undergraduates volunteering their time in exchange for extra credit for the term. Subjects were screened for reading disorders of type D1 and D4.
Experiment: One list of ninety words (see page ) was stored in a data file. The list was composed of 10 correctly spelled target words, 10 isophonically misspelled target words and 10 non words, additionally there were 60 correctly spelled, distractor words which bore no obvious relation to the targets but had similar frequency ratings and length.
Materials
An Apple Macintosh computer was used to print bold, 50 point, stimuli words on a 14" monitor.
Procedure
Phase 1- Subjects view the words at a comfortable reading distance from the screen in a sound proof room. The experimenter positions the free-standing microphone in front of the computer and adjusts its height according to the height of the subject. When the subject perceives the word on the screen s/he pronounces it into the microphone. This in turn activates a voice onset detector that stops the timer and relays a message to the computer to remove the stimulus from the screen.
Sample Stimuli
Subjects run through one practice session consisting of five trials before beginning the one session of ninety words. The words are presented in random parsing order with a delay of 500 ms between the screen refresh and the presentation of the next word. Reaction times are recorded automatically. False responses ("uhhs", stuttering, etc.) that the voice onset detector records as correct responses, are recorded by the computer as NULL responses and are dealt with accordingly.
Phase 2- After a short break, subjects are given a pen and paper and are asked to write the spellings of thirty-five words presented orally with a pause between each word as in Table 1: Incidence table for Speller Groups located on page .
Table 1: Incidence table for Speller Groups
Ten of the words are the target words taken from the experiment, and ten are randomly selected distractor words, and fifteen are selected from a list of commonly misspelled words (Dillingham, p.31). The two groups of spellers, good and poor, are established based on the subjects' ability to spell the thirty-five words correctly. The classification schema of the stimuli appears below (see Table 1, page ).
Results
Out of the 35 words on the spelling list, the mean number of misspelled words for the entire sample (26 subjects) was 8.2 with a standard deviation of 3.8 words. Consequently, the good spellers group included 13 subjects with spelling scores of eight. The poor spellers group also included 13 subjects, all with scores greater than eight.
Table 2: Bar Graph for Spelling Group at Word Type
It is evident from the graph of these results, Table 2, that poor spellers are consistently slower at naming over all the stimuli presented; differences between the group means at the correct word, misspelled word and non-word conditions are 28 ms, 53 ms and 116 ms respectively.
A two-way analysis of variance test shows that there is a significant speller by naming latency interaction, ( F(258, 2) = 10.8, p = .0011). This confirms that the post hoc construction of speller groups by spelling score is valid.
References
1. Banks, W.P., Oka, E. and Sugarman, S. (1981). Does recording come before lexical access? Reading Processes Hillsdale, NJ: Erlbaum.
2. Barron, R.W., (1980). Visual and Phonological Strategies in Reading and Spelling. Cognitive Processes in Reading. New York, NY: Academic Press.
3. Coltheart, M. (1978). Lexical access in simple reading tasks. On Strategies of Information Processing. New York, NY: Academic Press.
4. Dillingham, W. (1986). Practical English Handbook, Seventh Edition. Boston, MA: Houghton Mifflin Co.
5. Jorm, A. F. and Share, D. L. (1983). Phonological recoding and reading acquisition. Applied Psycolinguistics, 4, 103-147.
6. Just, M. and Carpenter, P. (1987). The Psychology of Reading and Language Comprehension. Boston, MA: Allyn and Bacon, Inc.
7. Kolb, B. and Whishaw, I. (1980). Fundamentals of Human Neuropsychology, (pp 528-531) Ney York, NY: W.H. Freeman and Company.
8. Taft, M. (1979). Lexical access via an orthographic code. Journal of Verbal Learning and Verbal Behavior, 18, 21-29.
