Relationships between temporal acuity, intelligence, phonological processing, and reading (Studies I-IV)

The associations between temporal acuity, phonological process-ing, readprocess-ing, and intelligence were investigated with principal components analyses (separately for unimodal and crossmodal tasks). In the study on children, the IQs could not be included in the analyses due to the small number of participants. As for

adults, the IQ was included but its exclusion did not change the structure of the solutions.

For the data from children (Study I), only a very preliminary principal components analysis was conducted, since the data on the reading-related variables from the controls were not availa-ble, and the number of dyslexic children was so small. However, as presented in Tables 3A and B, the results seem to suggest that crossmodal temporal acuity and phonological processing are re-lated, and independently of this, that phonological processing and reading are related. In unimodal comparisons (Table 3A), the first component explained the variation in temporal acuity (except the tactile component) as contrasted with phonological processing (phonological discrimination, naming speed, pho-neme analysis). The second component explained the variation in phonological processing (nonword spelling, phoneme analy-sis, phonological synthesis) and reading. The last component contrasted the tactile temporal acuity with nonword repetition (and more modestly spelling). In crossmodal comparisons (Ta-ble 3B) the relationships were clearer. The first component ex-plained the variation in all the temporal acuity tasks and in pho-nological processing (phopho-nological synthesis, nonword spelling, but in naming speed inversely). On the second component load-ed phonological processing (nonword spelling, phoneme analy-sis, naming speed) and the actual reading, as in the unimodal comparisons. The third component contrasted other tasks of pho-nological processing (phoneme analysis, phopho-nological discrimi-nation, phonological synthesis) with nonword repetition. There were few extreme outliers in the reading-related tasks, one in reading speed and another one in nonword spelling. When their values were replaced with the remaining groups’ mean, the prin-cipal components solution structures paralleled qualitatively those presented above.

A principal components analyses computed over the larger data set of developmentally dyslexic older adults (Study IV) re-sulted in four-component solutions with acceptable explanation power (see Tables 3A and 3B). These data were used, instead of those presented in the original Studies II and III, because of the

Table 3A. Principal components solution (biquartimax normalized ro-tated principal components loadings) over intelligence, reading-relat-ed, and logarithmically transformed unimodal temporal acuity tasks, stated separately for fluently reading older adults (Study IV), dyslexic children (Study I) and dyslexic older adults (Study IV).

D-child D-adult F-adult

NonWordRep 0.13 -0.20 -0.85 NonWordSpell 0.02 0.85 -0.31

(-) TactTOJ 0.71 -0.18 0.21 0.40 0.36 0.66 0.16 0.20

F: fluent reader, D: developmentally dyslexic reader, PhonDiscr: phonological discrimination, PhonSynth: phonological synthesis, LetRot: letter rotation, LexDec: lexical decision, RAS: naming speed, WordSeg: word segmentation speed, Rspeed:

reading speed, ComprNFic/Fic: reading comprehension non-fiction/fiction, FW: forward, BW: backward, PhonAn: phoneme analysis, NonWordRep/Spell: nonword repetition/spelling, Tact: tactile, Aud: auditory, Vis: visual

(-) Sign change to indicate better performance with a larger value.

(-C) Sign change to indicate better performance with a larger value only in the case of children.

greater number of participants and the overlap between the groups. This unimodal data is from the original Study IV and the following crossmodal data is from an unpublished crossmodal study of our research group (Laasonen, Lahti-Nuuttila, & Virsu, 2002). In unimodal comparisons (Table 3A), the first component explained the variation in the temporal acuity tasks and to a modest extent in the phonological processing tasks (nonword span, phonological discrimination, phonological synthesis). The second component explained the combined variation of phono-logical processing (phonophono-logical synthesis, naming speed) and reading (lexical decision, word segmentation, reading speed, reading comprehension of non-fiction). Letter rotation, too, load-ed best on this component. On the third component loadload-ed mem-ory tasks (span backward, WMS-R) and intelligence quotients (VIQ, PIQ) and to a modest extent the reading comprehension tasks (fiction, non-fiction). The fourth component explained best the variation in verbal short-term memory (span forward, mod-estly nonword span) as contrasted to reading comprehension (fic-tion). In crossmodal comparisons (Table 3B), the first component was again a combination of the temporal acuity tasks and pho-nological processing (nonword span, phopho-nological discrimina-tion, phonological synthesis). Letter rotation loaded best on this component as well. The second component combined, as in the case of unimodal comparison, memory (span backwards, WMS-R), intelligence (VIQ, PIQ), and reading comprehension tasks (fic-tion, non-fiction). The crossmodal TOJ tasks loaded modestly on this component as well. The third component combined, as with the unimodal comparisons, the tasks of phonological processing (naming speed, phonological synthesis) and reading (lexical de-cision, reading speed, word segmentation, reading comprehen-sion of non-fiction). The fourth component was again related to short-term memory (span forward, nonword span), but in this case to letter rotation and reading comprehension (non-fiction) as well.

In fluent adult readers’ unimodal comparisons (Table 3A), the first component explained the variation in intelligence (VIQ, PIQ), reading (reading speed, word segmentation, lexical decision),

Table 3B. Principal components solution (biquartimax normalized ro-tated principal components loadings) over intelligence, reading-relat-ed, and logarithmically transformed crossmodal temporal acuity tasks, stated separately for fluently reading older adults (Study IV), dyslexic children (Study I), and dyslexic older adults (Study IV).

D-child D-adult F-adult

NonWordRep -0.04 0.02 -0.78 NonWordSpell 0.43 0.73 -0.09

(-) AudTactTOJ 0.54 0.50 0.10 -0.30 0.03 0.71 0.26 0.44

F: fluent reader, D: developmentally dyslexic reader, PhonDiscr: phonological discrimination, PhonSynth: phonological synthesis, LetRot: letter rotation, LexDec: lexical decision, RAS: naming speed,

WordSeg: word segmentation speed, Rspeed: reading speed, ComprNFic/Fic: reading comprehension non-fiction/fiction, FW: forward, BW: backward, PhonAn: phoneme analysis, NonWordRep/Spell: nonword repetition/spelling, AudTact: audiotactile, VisTact: visuotactile, AudVis: audiovisual

(-) Sign change to indicate better performance with a larger value.

(-C) Sign change only to indicate better performance with a larger value in the case of children.

phonological processing (naming speed, phonological discrimi-nation, phonological synthesis), and reading comprehension (non-fiction, fiction). The second component explained the vari-ation in temporal acuity tasks and, to a modest extent, in phono-logical processing (nonword span, phonophono-logical synthesis). The third component was a combination of verbal intelligence (VIQ) and memory (spans backward and forward, WMS-R, nonword span). The fourth component combined the auditory TOJ with the reaction time tasks (letter rotation, lexical decision). Cross-modal comparisons (Table 3B) for the fluent readers resulted, as with the unimodal comparison, in the first component combin-ing phonological processcombin-ing (namcombin-ing speed), readcombin-ing (word seg-mentation, reading speed, lexical decision), reading comprehen-sion (non-fiction, fiction), intelligence (VIQ, PIQ), and WMS-R.

The second component combined the temporal acuity tasks and the reaction time tasks (letter rotation, lexical decision). The third component, as with the unimodal comparisons, explained the variation in verbal memory (span backwards, forwards, nonword span, WMS-R) and intelligence (VIQ). The fourth component combined the audiovisual TPA (and to a lesser extent the audio-tactile TOJ), phonological processing (phonological synthesis, phonological discrimination, nonword span, naming speed) and reading speed.