Some aspects affecting the outcome of the studies on

As is perhaps evident from the above review, there is no consensus on the temporal processing abilities of dyslexics. This is due to many factors which may have a decisive influence on the outcome of the experiments. In the following section a brief survey of the evidence reviewed above is presented with some speculations as to the reasons for the outcomes of some of the studies. All this information facilitates the formulation of the hypothesis for the two experiments of this chapter.

The nature of the stimuli, in the extent to which they were comprised of speech or nonspeech, did not appear to influence the results of the studies to any noticeable extent. Scarcely more than half of the studies (53%) using nonspeech stimuli showed that dyslexics were inferior to controls in the perception of the durational features of the stimuli. It should be noted that the majority of the studies reviewed here utilised nonverbal stimuli. In contrast, the studies employing speech stimuli were scarce. These studies indicated that in just over half of the cases (54%) the dyslexics were worse than the controls in perceiving durational differences in speech stimuli. Therefore, it is by no means clear if dyslexics differ from controls in perceiving durational characteristics of speech stimuli or if their deficits are evident in the processing of more basic auditory stimuli or if both of the domains are affected. What seems quite clear is that all except one (Hurford et al. 1990 being the exception) of those studies which found some temporal processing deficits in dyslexics showed that only a part of the dyslexic population displayed signs of the deficit.

One of the important factors in assessing experimental evidence is to pay attention to the means by which the data has been collected together, especially regarding the nature of the stimuli. Discrimination tasks (SD and ABX) with stimuli of variable duration yielded more results (five out of eight¹⁾ according to which there was no difference in the performance of the dyslexics when compared to the controls. In addition, one of the remaining three experiments (Watson 1992), with positive results as to the existence of a deficiency in discriminating durational differences, was replicated later with opposite results (Watson and Miller 1993). Thus, there does not seem to be clear evidence eitl1er for or against ttte possible i:nferiorit)⁷ of dyslexics irt discrimination tasks with temporal parameters.

As for the discrimination tasks in which the durational variations were located in the interstimulus intervals of stimuli (all of them with nonspeech stimuli), the results appear to be in no agreement either. Five experiments out of nine showed that dyslexics are inferior in discriminating stimuli which are presented with a short interstimulus interval. Two of these discrimination tasks consisted of discrimination tasks using relatively long rhythmic patterns, and thus, they do not provide direct evidence on the ability to discriminate

1 The results of Brandt and Rosen are not included here since they did not specifically report statistical fi_gures on the data.

durational differences. Thus, although the evidence is somewhat inclined to show deficiencies in dyslexics in this respect this aspect should be studied further to confirm these results.

In the scarce identification tasks the results are also divergent, sixty percent of them indicating a deficiency in dyslexics. In the only two identification tasks (Reed 1989, Hurford & al. 1990) in which the ISis were varied, the results indicate a difference in performance between dyslexic and control subjects: all the stimuli were less consistently identified by dyslexics when compared to controls when the ISis were relatively short. However, in Hurford et al' s study (1990) there was not a statistically significant difference between the older dyslexics and their age matched controls. Since the experimental evidence is divergent in this respect more information is needed in elucidating these abilities of dyslexics. The current study deals specifically with the identification of speech stimuli, and, thus, by testing the ability to categorise speech stimuli according to durational changes more evidence on the identification skills of dyslexics is provided here.

The results of the TOJ tasks with variable durations of ISis are also divergent, six of the ten tasks producing positive results indicating significantly poorer performance in dyslexic subjects in comparison to that of controls. Among the evidence are Tallal's two experiments which yielded contradictory results using exactly the same method in both experiments but with different subject groups. Also, the nature of the stimuli affected the performance in Reed's studies indicating that the briefness of the perceptual cues may cause difficulties for dyslexics in perceiving the order of presentation of the stimuli. It seems that the nature of the stimuli affected the performance of the dyslexics in this task and there seems to be a subgroup of dyslexics which are, in particular, poorer in the task with durational changes of ISis.

It is important also to pay attention to the demands of the different types of task for the subjects. The TOJ tasks do not provide evidence on mere identification or discrimination of the stimuli since also the sequential order of the stimuli needs to be processed. Thus, some level of memory load is also present in the performance of a TOJ task. The auditory discrimination of stimuli on the other hand means that listeners have to detect a difference between two or more stimuli. In the case of speech stimuli, the criterion according to which the decision is made can be based on either the purely physical (i.e., durational) side of the stimuli or to the phonological side (i.e., durational differences which goinside with phoneme categories of the particular language in question), or both of these. Auditory identification means that listeners have to identify one stimulus or more, i.e. they have to label them. If the listeners are given two choices (as was the case in virtually all of the identification tasks reviewed) then they are forced to identify the stimuli as one of the provided responses: The forced choice identification tasks also imply that the listener can discriminate the difference between the two choices either acoustically or phonologically or both. The two experiments of this chapter comprise categorisation tasks: the experiment with adult subjects is a forced choice categorisation task whereas that of infant subjects can be considered a categorisation task in which the choices of responses are not provided directly. The question of the nature of the tasks will be returned to later on in the Methods part of this chapter.

As can be seen the comparison of the results is by no means easy and it is made even more difficult and even futile by the fact that the selection criteria for the dyslexic groups varied significantly between studies (e.g., Pallay did not test severe dyslexics, Tallal and Stark's dyslexics were deviant particularly in their comprehension abilities, Steffens et al. studied compensated dyslexics, etc.). Also the age of the subjects varied a great deal between the studies, which has implications for the interpretation of the results. In addition, there appears to have been a good deal of variability in the performance of the individual dyslexics in various tasks. Therefore, the selection criteria should be rigorous and similar between studies and only then could the results of different studies be comparable between similar tasks using similar stimuli.

One factor that surfaced from the investigation of these dyslexia studies, however, is that it appears that if there exist dyslexics who differ in the auditory processing of the temporal features of stimuli, it seems that they have specific problems dealing with stimuli that are short in duration or that are presented with short ISis (McCroskey & Kidder 1980, Tallal 1980, Pallay 1986, Reed 1989, Hurford & Sanders 1990, Steffens et al. 1992, Farmer & Klein 1993, Hari & Kiesila 1996): When the ISis get sufficiently long or the stimuli cues are longer in duration the perception abilities of the dyslexics and controls became similar. It is important to notice, however, that several of the studies (Reed 1989, Hari & Kiesila 1996) indicated that the performances of the two subject groups differed also in the longest durations of ISis, dyslexics being inferior to the controls. According to the majority of the results of previous studies it could be hypothesised that in the current experiments the dyslexics would differ from the controls in the categorising of the stimuli in cases in which the duration of a stop closure belongs to the shorter end of the durational continuum. In other words, previous experimental evidence indicates that dyslexics in comparison to nondyslexics may need longer durational differences between stimuli in order to categorise them differently. This hypothesis will be tested in the two perception experiments of this chapter.