No abstract

Poor reading skills of developmental dyslexics persist into adulthood with standard remediation protocols having little effect. Nevertheless, reading improves if readers are induced to read faster. Here we show that this improvement can be enhanced by training. Training follows a multi-session procedure adapted to silent sentence reading, with individually set, increasingly more demanding, time constraints (letter-by-letter masking). In both typical and dyslexic adult readers, reading times are shortened and comprehension improves. After training, the dyslexic readers' performance is similar to that of typical readers; moreover, their connected text reading times and comprehension scores significantly improve in standard reading tests and are retained at 6 months post training. Identical training without time constraints proves ineffective. Our results suggest that fluent reading depends in part on rapid information processing, which then might affect perception, cognitive processing and possibly eye movements. These processes remain malleable in adulthood, even in individuals with developmental dyslexia.

I carried out four experiments to determine the effects on decoding mistakes and comprehension of test passages at fastest and slowest reading rates. The subjects of the first three studies were 161 Israeli first graders, and those of the last experiment were 61 American first graders reading English. Analysis of reading rates obtained during a self-paced condition provided the base rates for each subject. I hypothesized that because of the constraints of short-term memory, requiring subjects to maintain their own maximal oral reading rates would result in improvements in both reading accuracy and comprehension. When presented with the text at their maximal normaJ reading rates, subjects averaged fewer reading errors and higher comprehension scores than in the self-paced conditions-By contrast, when the text was presented at the slowest reading rates, subjects' decoding accuracy improved, but their comprehension decreased significantly. In one of the experiments, the text contained deliberate letter-substitution errors. Increased reading rate once more reduced the overall errors and increased comprehension. In addition, the deliberate mistakes were more frequently corrected to normal words than in the self-paced condition. I thank Charles Perfetti for his help with Experiment 4, the study conducted in America.

The author tested the hypothesis that gains in reading performance occurring among dyslexic children during individually paced accelerated reading are partially attributable to changes in short-term memory (STM) processing. Dyslexic children's performance on a series of standard STM-sensitive tasks administered during both self-and fast-paced reading conditions confirmed this hypothesis. Trie findings provided strong support for a causal role of STM functioning in text processing but indicated that reading acceleration enhanced STM only when dyslexic children were able to rely on contextual cues.

Based on the finding that increasing stimulus presentation rate can increase short-term memory (STM) span, reading rate was accelerated in a series of experiments that demonstrated substantial gains in reading accuracy and comprehension. The present investigation tested the hypothesis that comprehension gains in fast-paced reading are attributable primarily to changes in STM functioning. A series of standard STM-sensitive tasks were administered in both self-paced and fast-paced reading conditions. The magnitude and relative specificity of effects evident in fastpaced reading confirmed the hypothesis. Several alternative hypotheses were rejected. These findings provide strong support for a causal role for STM function in text processing.Short-term memory (STM) is assumed to play an important role in reading (Baddeley, 1986;Daneman & Carpenter, 1980;Perfetti & Lesgold, 1977). STM refers to temporary storage of information being processed. This working storage system is characterized by limited capacity (Miller, 1956;Peterson & Peterson, 1959) and rapid decay (Collins & Loftus, 1975;Hitch, 1978), with stimuli retained for no longer than approximately 30 s (Anderson, 1980;Lindsay & Norman, 1977). These characteristics of the memory system pose serious limitations for information-processing tasks, such as reading, which require that the words and phrases being decoded be retained temporarily in memory to be arranged and encoded into meaningful sentences or propositions (Perfetti & Lesgold, 1977).Research on STM has indicated that subjects use a variety of strategies to minimize the functional limitations of shortterm memory. These strategies include organization of units into chunks, rehearsal, and parsing (