Vivian I. Schneider

Implementing a long-term computerized remedial reading program with synthetic speech feedback: Hardware, software, and real-world issues

This paper discusses hardware choices, software developments, implementation issues, and preliminary results from an ongoing long-term remedial reading study. Reading-disabled children read books on microcomputers linked to speech synthesizers, obtaining speech feedback on difficult words at whole-word, syllable, or subsyllable levels of segmentation. Word-recognition ability and attitude about reading improved for children using the system. In addition, segmented feedback especially benefited phonological word-decoding skills for most of the children.

Detecting Phonemes and Letters in Text: Interactions Between Different Types and Levels of Processes

In six experiments, subjects detected phonemes or letters in text presented auditorily or visually. Experiments 1 and 2 provided support for the hypothesis that a mismatch between the phoneme and letter representations of a target leads to detection errors. In addition, visual word unitization processes were implicated. Experiments 3 and 4 provided support for the hypothesis that the Gestalt goodness of pattern affected detection errors when subjects searched for letters. Experiments 5 and 6 demonstrated that the effects of unitization on the detection of letters in common words were decreased by altering the familiar configuration of the test words. The combined results of all six experiments lead to the conclusion that both visual and phonetic processes influence letter detection, that these processes communicate through a type of cross-checking, and that there are at least two levels of visual (and perhaps of phonetic) processing involved in the letter detection task.

The long-term retention of knowledge and skills

Abstract : We received three classes of guidelines we found to optimize long-term retention. The first class concerned ways to optimize the conditions of training. We discussed three general guidelines in this class. The first concerned the contextual interference found, for example, with random sequences of tasks as opposed to fixed or predictable sequences. The second concerned training parts of a task versus the whole task. The third concerned the distinction between generating and reading. The second class of guidelines concerned ways to optimize the strategies used. We found that in tasks that require deliberate retrieval from memory, training that promotes efficient encoding strategies maximizes long-term retention. The third class of guidelines concerned ways to attain direct access, or automatic retrieval, from memory. We found in several domains that achieving automaticity requires extensive practice. Further, even when retrieval appears automatic after extensive practice, mediators may still continue to exert their influence.

The role of phonetic processes in letter detection: A reevaluation

Abstract In seven experiments we investigated the finding that subjects often fail to detect the letter f in the word of . This effect was depressed in Experiment 1 when asterisks were inserted between all letters in a passage, suggesting that unitization processes contributed to the effect. Experiment 2 compared detection of the letters o and f and found a significant but much smaller effect for o , indicating that unitization can only provide a partial explanation for the effect with f . In Experiment 3 the effect was maintained for the letter f but not the letter o when subjects searched an auditory version of the passage, which implies that phonetic processes are involved in the effect for f . Experiments 4, 5, and 6 demonstrated that the effect is still observed when subjects are told to search for v as well as f , even with an auditory presentation of the passage and even when instructed to detect the phonemes rather than the letters. These findings are inconsistent with the proposal by Read (1983 , Memory and Cognition , 11, 390–399) that subjects miss the f in of because they are searching only for the phoneme /f/. Experiment 7 showed that the effect can be reduced if subjects are led to interpret the letter sequence of as a misspelled version of the word off and can be eliminated entirely if subjects are also required to write down every letter of the passage. The findings of Experiment 7 provide further evidence that unitization and phonetic processes together contribute to the effect. The results as a whole are taken to imply that both letter and phoneme levels need to be included in models of word identification as well as a cross-checking or communication between the two levels.

Effects of Instruction Modality and Readback on Accuracy in Following Navigation Commands

In 3 experiments, the authors simulated air traffic controllers giving pilots navigation instructions of various lengths. Participants either heard or read the instructions; repeated either all, a reduced form, or none of the instructions; and then followed them by clicking on the specified locations in a space represented by grids on a computer screen. Execution performance for visual presentation was worse than it was for auditory presentation on the longer messages. Repetition of the instructions generally lowered execution performance for longer messages, which required more output, especially with the visual modality, which required phonological receding from visual input to spoken output. An advantage for reduced over full repetition for visual but not for auditory presentation was attributed to an enhanced visual scanning process.

Searching for target letters in memory: Individual preferences and instructions for text representation

Participants searched for target letters in a short passage held in memory. In Experiment 1, participants were divided into two groups on the basis of a retrospective report concerning the type of representation used to store the passage in memory, and in Experiment 2, participants were instructed concerning the form of memory representation to use. Only participants using a visual representation missed more targets in the wordthe than in other words. Participants instructed to form a visual representation also made fewer content-word or phrase substitutions when learning the passage than did participants instructed to form an auditory representation. These findings show that choice of memory representation is flexible and that the representation used influences what can be retrieved from memory.

Effects of difficulty, specificity, and variability on training to follow navigation instructions

To study the relative merits of three training principles – difficulty of training, specificity of training, and variability of training – subjects were trained to follow navigation instructions to move in a grid on a computer screen. Subjects repeated and then followed the instructions by mouse clicking on the grid. They were trained, given a short distractor task, and then tested. There were three groups, each receiving different message lengths during training: easy (short lengths), hard (long lengths), and mixed (all lengths), with all subjects given all lengths at test. At test, the mixed group was best on most lengths, the easy group was better than the hard group on short lengths, and the hard group was better than the easy group on long lengths. The results support the advantages of both specificity and variability of training but do not support the hypothesis that difficult training of the form used here would lead to overall best performance at test.

Does spatial information impact immediate verbatim recall of verbal navigation instructions

The present study addresses the issue of whether spatial information impacts immediate verbatim recall of verbal navigation instructions. Subjects heard messages instructing them to move within a two-dimensional depiction of a three-dimensional space consisting of four stacked grids displayed on a computer screen. They repeated the instructions orally and then followed them manually by clicking with a mouse on the grids. Two groups with identical instructions were compared; they differed only in whether the starting position was indicated before or after the instructions were given and repeated, with no differences in the manual movements to be made. Accuracy on both the oral repetition and manual movement responses was significantly higher when the starting position was indicated before the instructions. The results are consistent with the proposal that there is only a single amodal mental representation, rather than distinct verbal and nonverbal representations, of navigation instructions. The advantage for the before condition was found even for the oral repetition responses, implying that the creation of the amodal representation occurred immediately, while the instructions were being held in working memory. In practical terms, the findings imply that being able to form a mental representation of the movement path while being given verbal navigation instructions should substantially facilitate memory for the instructions and execution of them.

How Much is Remembered as a Function of Presentation Modality

ABSTRACT According to a widespread claim used for teaching recommendations, students remember 10% of what they read, 20% of what they hear, 30% of what they see, and 50% of what they see and hear. Clearly, the percentages cannot be correct, and there is no empirical evidence for the ordering. To investigate the ordering, in a navigation paradigm, subjects were given messages instructing them to move in a grid of four stacked matrices by clicking a computer mouse. Three modalities were compared presented either once, see (visual arrows), hear (auditory words), read (visual words); twice in succession, see-see, hear-hear, read-read; or in two different successive modalities, see-hear, hear-see, see-read, read-see, hear-read, read-hear. Better performance was found for messages presented twice than once, but messages in the two modalities were not always better than twice in one modality. For the twice-presented messages, performance varied as a function of the second modality, with see best and read worst. However, the ordering for the first modality was not reliable and was inconsistent with the widespread claim. Thus, the widespread claim is clearly wrong, not only in its percentages, but also because of its lack of generality.

Training specificity and transfer in time and distance estimation

Learning is often specific to the conditions of training, making it important to identify which aspects of the testing environment are crucial to be matched in the training environment. In the present study, we examined training specificity in time and distance estimation tasks that differed only in the focus of processing (FOP). External spatial cues were provided for the distance estimation task and for the time estimation task in one condition, but not in another. The presence of a concurrent alphabet secondary task was manipulated during training and testing in all estimation conditions in Experiment 1. For distance as well as for time estimation in both conditions, training of the primary estimation task was found to be specific to the presence of the secondary task. In Experiments 2 and 3, we examined transfer between one estimation task and another, with no secondary task in either case. When all conditions were equal aside from the FOP instructions, including the presence of external spatial cues, Experiment 2 showed “transfer” between tasks, suggesting that training might not be specific to the FOP. When the external spatial cues were removed from the time estimation task, Experiment 3 showed no transfer between time and distance estimations, suggesting that external task cues influenced the procedures used in the estimation tasks.