The thesis opens with an exposition of different uses of the term dyslexia. In that context its conceptual relevance is discussed. The empirical studies investigate a) different aspects of validity of cognitive and achievement instruments often used in diagnostic assessment of dyslexia, b) different cognitive profiles for adults with dyslexia, and c) the relationship between cognitive and achievement measures.
Study I demonstrated the factor structure of the Swedish WAIS-R to be in close agreement with results of comparable analyses on normal and clinical groups in many countries world-wide, giving strong support for the construct validity of the Swedish WAIS-R for a group of adults diagnosed with dyslexia. The results from the three-factor solution were interpreted in terms of theoretical constructs, notably those posited by Carroll (1993) and Horn (1989).
The cognitive profiles of a sample of Swedish adults diagnosed with dyslexia—when using the three factors, the ACID profile, and the four-category Bannatyne profile—all resemble closely the profiles observed for a wide array of U.S. samples of children and adults diagnosed with dyslexia or learning disabilities. Gender differences on Digit Symbol, favoring females, were substantial in magnitude for the present sample, consistent with a bulk of research on gender differences for samples of normal individuals and of those identified with learning problems.
In Study II, factor analysis of nine Swedish achievement tests often used for dyslexia assessment yielded five meaningful factors in a group of adults diagnosed with dyslexia. Factors appeared to measure decoding, visual speed, reading comprehension, reading fluency, and phonological ability. The relationship between the five achievement factors and WAIS-R variables was subsequently examined. The Visual Speed and Reading Comprehension factors each correlated significantly with four of six global WAIS-R scores (two of three IQs and two of three factor scores). Visual Speed tended to correlate significantly with nonverbal scores and Reading Comprehension with verbal scores. Although the Phonological Ability factor did not correlate significantly with any verbal or nonverbal global score, it did correlate significantly with the Freedom from Distractibility factor and with all its component subtests—Digit Span, Arithmetic, and Digit Symbol. Decoding (technical reading skill) showed no significant relationship to any WAIS-R variable studied. More surprisingly, the Reading Fluency factor failed to show significant relationships to the WAIS-R.
Study III examined the validity of The Word Chain Test, a frequently used instrument in Swedish screening and diagnostic assessments of dyslexia. Different sources of validity evidence were evaluated. In summary, the results failed to support validity both for the WRI-index and the Wordchain subtest, suggesting that the instrument seems to be of questionable value in screening or diagnostic assessment of dyslexia.