Sequence Learning and Audiological Outcomes

The present study extends earlier research carried out by Cleary and Pisoni (2001) who found that measures of auditory sequence learning, using the Simon memory game procedure, were related to open-set spoken word recognition and language comprehension. Sequence learning scores were obtained from a new group of profoundly deaf children with cochlear implants. Two different measures of sequence learning were computed. One measure, Simon auditory redundancy gain, was used to assess the benefit of redundant auditory information on the reproduction of visual sequences of colored lights. A second measure, Simon learning improvement, was used to assess the increase in sequence learning observed over time after a period of implant use. Both learning measures were found to be correlated with traditional audiological measures of outcome and benefit. The auditory gain measure was significantly correlated with two measures of language comprehension on the Common Phrases test. Vocabulary knowledge on the PPVT was found to be correlated with the improvement in sequence learning over time. Taken together with the earlier findings reported by Cleary and Pisoni (2001), the present results suggest that differences in learning may contribute an additional source of variance to traditional measures of speech and language outcomes in this clinical population. Measures of learning and memory may therefore provide important new insights into the underlying cognitive and neurobiological factors that are responsible for the individual differences and enormous variation in a range of clinical speech and language outcome measures that are routinely obtained from deaf children who have received cochlear implants as a treatment for profound hearing loss. Introduction Many of the traditional methods for measuring working memory span and the capacity of immediate memory use recall tasks that require a subject to repeat back a sequence of test items using an overt articulatory-verbal motor response (Dempster, 1981). Because deaf children with cochlear implants may also have delays and/or disorders in speech motor control and phonological development, it is possible that any differences in performance between deaf children and age-matched normal-hearing children using memory span tasks could be due to the nature of the response requirements used during retrieval and output. Differences in articulation and speech motor control could magnify other differences in encoding, storage, rehearsal or retrieval processes. To eliminate the use of an overt articulatory-verbal response, we developed a new experimental methodology to measure immediate memory span in deaf children with cochlear implants based on Milton-Bradley's Simon, a popular memory game. Figure 1 shows a display of the apparatus which was modified so it could be controlled by a PC. In carrying out the procedure, a child is asked to simply "reproduce" a stimulus pattern by manually pressing a sequence of colored panels on the four-alternative response box. In addition to eliminating the need for a verbal response, the Simon methodology permitted us to manipulate the stimulus presentation conditions in several systematic ways while holding the response format constant. This particular property of the experimental procedure was important because it provided us with a novel way of measuring how auditory and visual stimulus dimensions are analyzed and processed alone and in combination and how these stimulus manipulations affected measures of memory span. The Simon memory game apparatus and methodology also offered us an opportunity to SEQUENCE LEARNING AND AUDIOLOGICAL OUTCOMES 321 study learning processes, specifically, sequence learning and the relations between memory and learning using the same identical experimental procedures and response demands. Figure 1. The memory game response box based on the popular Milton Bradley game “Simon.” In our initial studies with the Simon apparatus, three different stimulus presentation formats were employed (Pisoni & Cleary, 2004). In the first condition, the target sequences to be reproduced consisted only of spoken color names (A). In the second condition, sequences of colored lights (L) were presented in the visual modality. In the third presentation condition, the spoken color names were presented simultaneously with matching colored lights (A+L). Forty-five hearing-impaired children with cochlear implants were tested using the Simon memory game apparatus. Thirty-one of these children were able to complete all six conditions included in the testing session. They also were able to identify the recorded color-name stimuli used in this task when these items were presented alone in isolation. Thirty-one normal-hearing children who were matched in terms of age and gender with the group of children with cochlear implants were also tested. Finally, 48 normal-hearing adults were recruited to serve as an additional comparison group (see Pisoni & Cleary, 2004). Of the six conditions tested, three measured the children’s immediate memory skills and three measured the children’s sequence learning skills. In the immediate memory task, the temporal sequences systematically increased in length as the subject progressed through successive trials in the experiment. Within each condition, the child started with a list length of one item. If two lists in a row at a given length were correctly reproduced, the next list was increased by one item in length. If a list was incorrectly reproduced, the next trial used a list that was one item shorter in length. This adaptive tracking procedure is similar to methods used in psychophysical testing (Levitt, 1970). Sequences used for the Simon memory game task were generated pseudo-randomly by a computer program, with the stipulation that no single item would be repeated consecutively in a given list. We computed a weighted memory span score for each child by finding the proportion of lists correctly reproduced at each list length and averaging these proportions across all list lengths. A summary of the results from the Simon immediate memory task for the three groups of subjects is shown in Figure 2. The normal-hearing adults are shown in the left panel, the normal-hearing agedmatched children are shown in the middle panel and the children with cochlear implants are shown in the right panel. Within each panel, the scores for auditory-only presentation (A) are shown on the left, scores for lights-only presentation (L) are shown in the middle and scores for the combined auditory and lights presentation condition (A+L) are shown on the right.