Reading difficulties from birth to adolescence: Early predictors and links to motivation and well-being Minna Torppa University of Jyväskylä, Finland Den sjunde nordiska kongressen om dyslexipedagogik Stockholm 14. – 16.8.2014 OUTLINE I Decoding skills and dyslexia II Reading comprehension development … and how they are linked with: • Family risk • Cognitive skills • Motivation • Well-being Data Jyväskylä Longitudinal Study of Dyslexia (JLD) • A family risk study for dyslexia • Four age cohorts followed from birth to age 20 • About 100 children with high familial risk for dyslexia and 100 control children. •On grades 1,2,3,7, and 9 also classmates were assessed (n ≈ 1500-2500). •Started: 1993, currently funded until 2019 AIMS OF JLD (1) Identify predictors of dyslexia (2) Specify the developmental paths leading to reading difficulties (3) Examine the contribution of environmental factors associated with dyslexia (4) Examine the developmental problems co-occurring with dyslexia (5) Develop methodology for early assessment and intervention I Screening Born at the hospitals of Central Finland during 01.04.9331.07.96 Short questionnaire administered at the maternity clinics N=8427 parents N= 9368 infants Comprehensive questionnaire III Screening AT -RISK GROUP Assessment of parents’ reading and spelling skills CONTROL GROUP N=410 parents N=3146 parents N=117 infants N=105 infants III VII VIII IX grade grade grade grade grade N = N = N = N = N = 108 108 85 101 88 18 2 2½ 3½ 4½ 5 5½ 6½ I II month month years years years years years years years grade N = N = N = N = N = N = N = N = N = N = Neonatal 6 14 month N = N = 107 112 II Screening 108 108 107 107 107 107 107 107 107 107 N = N = N = N = N = N = N = N = N = N = N = N = N = N = N = N = N = 96 94 94 95 96 94 95 93 93 93 93 92 92 92 66 81 76 CLASSMATES N= N= N= 1549 1756 2641 N= 1452 N= 1705 I Decoding skills and dyslexia • Reading development and dyslexia in grades 2, 3, and 8 • Links to family risk and cognitive skills • Links to psycho-social functioning Family risk and dyslexia Grade 2 dyslexia - 38 at-risk group children (35.8 %) - 9 control group children (9.8 %) About fourfold risk if familial incidence of dyslexia Early cognitive predictors of Grade 2 dyslexia (mean age 8.9 years) Logistic regression analyses at ages 3.5 years, 4.5 years,and 5.5 years with predictors: Family risk, Phonological awareness, Short-term memory, RAN, Vocabulary, Pseudoword repetition and Letter naming Significant predictors of dyslexia were: 3.5 years: Familial risk, RAN, and Letter knowledge 4.5 years: Familial risk, Phonological awareness, and Letter knowledge 5.5 years: Familial risk, RAN, and Letter knowledge Puolakanaho, A., Ahonen, T., Aro, M., Eklund, K., Leppänen, P. H. T., Poikkeus, A.-M., Tolvanen, A., Torppa, A., & Lyytinen, H. (2007). Very early phonological and language skills: estimating individual risk of reading disability. Journal of Child Psychology and Psychiatry, 48(9), 923-931. Development beyond Grade 2? Torppa, Eklund, van Bergen, & Lyytinen Grade 8 dyslexia - 40 at-risk group children (39.6 %) - 9 control group children (11.1 %) Stability of diagnosis? Background The proportion of late-emerging cases has been reported to be approximately 40% of all RD cases (Catts et al., 2012; Leach et al., 2003; Lipka et al., 2006). Surprisingly little research has been conducted on the characteristics of resolving and late-emerging RD, and all in English. Participants n = 182, followed from birth to Grade 8 Children at risk (n = 101) had a parent and one or more other close family members with dyslexia. 86 girls and 96 boys Dyslexia Criteria in Grades 2 and 8 Dyslexia criteria were based on the following fluency tasks: 1) Word list reading, 2) Text reading, and 3) Pseudoword text reading. A cut-off criterion for deficient performance was defined for each measure, using the 10th percentile in the distribution of the children without family risk. Dyslexia = scores below the criterion in at least two out of three measures of reading speed. Stability of diagnosis? 40% had dyslexia in Grades 2 and 8 (persistent-dyslexia), 27% only in Grade 2 (resolving), and 33% only in Grade 8 (late-emerging). Group differences? Parent’s Educationa Mother Father No-dyslexia Late-emerging Resolving Persistent n = 127 n = 18 n = 15 n = 22 M SD M SD M SD M 4.52 SD 1.38 4.59 1.70 3.92 1.24 4.12 1.50 3.72 1.42 4.06 1.71 3.64 .81 3.75 1.39 100.74 10.93 100.83 6.04 97.47 9.10 95.00 10.37 Children IQ Family risk 44.9% 83.3% 73.3% 81.8% Girls 49.6% 22.2% 80.0% 45.5% Cognitive Measures Vocabulary . At age 3.5 and 5.5 years the Boston Naming Test (Kaplan, Goodglass, & Weintraub, 1983). In Grade 2 the WISC-III (Wechsler, 1991). Memory. Verbal short-term memory: 6.5 years, Grades 3 and 8 with a forward digit span test. Phonological awareness At age 4.5, 5.5, and 6.5 years phonological awareness was measured with a composite mean of z-scores from four tasks: First phoneme identification, First phoneme production, Segment identification, and Synthesis Rapid naming. The total naming time of 30 (5.5 years) or 50 objects (6.5 years, and in Grades 2, 3 and 8 ) (Denckla & Rudel, 1974) Letter knowledge. All 29 lowercase letters (23 typically used and 6 for the rare loan words) in the Finnish alphabet were presented at ages 4.5 years, 5.5 years, and 6.5 years. IQ (Grade 2). Four performance-quotient subtests (Picture Completion, Block Design, Object Assembly, and Coding) and five verbal-performance subtests (Similarities, Vocabulary, Comprehension, Series of numbers, and Arithmetic) of the WISC-III-R were used. Cognitive skills • Resolving group improved also in vocabulary, RAN and PA • Resolving group only had clear early problems in vocabulary • Late-emerging had clear problems only in RAN Parental Measures Parental assessment Text-reading fluency. Parents were asked to read aloud two passages (218 and 128 words, respectively) as fluently and accurately as possible. A measure of reading fluency was the average reading time for the two texts Phoneme deletion. Parents were asked to pronounce a given word without the second phoneme. The task included 16 words (e.g., kaupunki ‘city’ became kupunki) of 4 to 10 letters with 2 to 4 syllables. Deletion of the second phoneme yielded a pseudoword. Rapid Naming (RAN). On each of three tasks, participants were asked to name, as rapidly as possible, a matrix of 50 items comprising objects, digits, or a mixture of digits, objects, and letters. A mean composite score of the three standardized RAN scores was calculated for the prediction analyses . Verbal short-term memory. The Digit Span subtest of WAIS-III (Wechsler, 1991) was administered Two sets of items, one for forward, the other for reverse, were used. Scaled scores were derived from the manual. Vocabulary. In the Vocabulary subtest of WAIS-III (Wechsler, 1991), participants were required to define 35 words in his/her own words. Scaled scores were derived from the manual. Parental skills In sum Dyslexia status was not stable. Family risk high in all dyslexia groups Persistent dyslexia linked to difficulties in phonological awareness, RAN, and letter knowledge Late-emerging dyslexia linked to RAN (also parental) and being a boy Resolving dyslexia linked to slow early development also in vocabulary, phonological awareness, RAN, and letter knowledge and being a girl Implications: It is important to keep following children’s literacy development beyond the early grades. I Decoding skills and dyslexia • Links to family risk and cognitive skills • Reading development and dyslexia in grades 2, 3, and 8 • Dyslexia and links to psycho-social functioning Parhiala, Torppa, Eklund, Aro, Poikkeus & Ahonen Dyslexia & psychosocial functioning Dyslexia is linked to difficulties in the social domain, and in externalizing, and internalizing, and attentional problems (e.g., Dahle, Knivsberg & Andreassen, 2011; Morgan, Farkas & Wu, 2012). Are problems with psychosocial functioning secondary problems or co-occurring with dyslexia? Previous studies inconsistent in their interpretations and have typically started after school entry; I Screening Born at the hospitals of Central Finland during 01.04.9331.07.96 Short questionnaire administered at the maternity clinics Comprehensive questionnaire N=8427 parents N=3146 parents N= 9368 infants Neonatal II Screening # of children who have attended the last finished assessment phase at the 3rd grade III Screening Assessment of parents’ reading and spelling skills N=410 parents AT -RISK GROUP AT -RISK GROUP N=117 infants N=108 infants CONTROL GROUP CONTROL GROUP N=105 infants N=92 infants 6 14 18 2 2½ 3½ 4 5 5½ 6 month month month years years years years years years years 4 years N = N = N = N = N = N = N = N = N = 107 112 108 108 107 107 107 107 107 N = N = N = N = N = N = N = N = N = 96 94 94 95 96 94 95 93 93 6 School entry II I grade N =N 93 = 93 CLASSMATES 107 N =N 93= 93 grade VII VIII IX grade grade grade N = N = N = 70 82 21 N = N = N = 61 76 14 years years N N N N N = = = = = 107 107 107 107 9 grade III N = 92 N N N = = = 108 108 108 N =N 92 = 92 N = 92 N= N= 1515 2859 Psychosocial functioning measure Parent Rating Scale of the Behavior Assessment System for Children (BASC, Reynolds & Kamphaus, 1992): 4-point scale ranging from “Never” to “Almost Always”. 1) Adaptive skills (Adaptability, Social skills), 2) Attention problems, 3) Externalizing problems (Aggression, Hyperactivity), and 4) Internalizing problems (Anxiety, Depression, Somatization). Results Children with dyslexia: More problems in adaptation and social skills prior to school entry: 4 and 6 years. More inattention both prior to school entry and after school entry: 4, 6 and 9 years. No differences in externalizing or internalizing problem behavior Difference in change from age 6 to age 9? Change from 6 to 9 years Adaptation and social skills: Children with dyslexia improved more in adaptability and social skills than children without dyslexia Inattention: Differential change for girls and boys: inattention increased across transition to school only among girls. Both girls with and without dyslexia. Externalizing and internalizing problems: no change In sum Inattention: problems prior to school entry not secondary Increase for girls both with and without dyslexia not secondary Adaptability and social skills: problems prior to school entry not secondary positive change among children with dyslexia not secondary Externalizing & Internalizing problems No links to Grade 2 dyslexia II Reading comprehension Simple view of reading Reading comprehension in grades 1-2 and 9 Links to skills and motivation Background Simple view of reading (Gough & Tunmer, 1986); Comprehension reading = ƒ (decoding and comprehension) good Dyslexia poor Poor readers Hyperlexia poor good Decoding Torppa, M., Tolvanen, A., Poikkeus, A-M. Eklund, K., Lerkkanen, M-K., Leskinen, E., & Lyytinen, H. (2007). Reading Development Subtypes and Their Early Characteristics. Annals of Dyslexia, 57, 3-32. Decoding and reading comprehension in JLD A 2-year follow-up data with four group assessments of 1750 children from 93 classrooms. SC D 1st grade fall D 1st grade spring TC TC D D 2 nd grade fall 2 nd grade spring SC = Sentence Comprehension, TC = Text Comprehension, D = Decodi ng Identification of Subtypes: a 2-level Mixture Model C fwr1 fwr2 fwr3 fwr4 rc1 Wfwr rc2 The search for subtypes was based on individual variation rc3 Wrc Within Between fwr1 fwr2 fwr3 fwr4 rc1 Bfwr rc2 Brc C C = latent categorical variable (subtyping), fwr = fluent word recognition rc = reading comprehension. W = within-level, B = between-level. Numbers denote assessment time-point. rc3 Classroom effect was significant and ranged between 4 % and 10 % (Intraclass correlation). 5 Reading subtypes were identified Standard score 2 1 0 -1 -2 1st grade fall 1st grade spring 2nd grade fall 2nd grade spring Fluent Word Recognition Good readers (n=176) Slow decoders (n=411) Poor comprehenders (n=179) 1st grade spring 2nd grade fall 2nd grade spring Reading Comprehension Average readers (n=692) Poor readers (n=210) Group differences in: • family risk for dyslexia • early cognitive development profiles? Reading Subtypes and Family Risk for Dyslexia Table 6. Crosstabulation of the Reading Subtypes by Study Groups (Frequencies in Parentheses) Good readers (n = 167) Average readers (n = 643) Slow decoders (n = 478) Poor comprehenders (n = 241) Poor readers (n = 221) Total n Classmates JLD at-risk group JLD control group 10.3 % (160) 9.6 % (10) 20.7 % (18) 42.0 % (655) 30.8 % (32) 43.7 % (38) 24.6 % (383) 35.6 % (37) 14.9 % (13) 10.6 % (166) 6.7 % (7) 11.5 % (10) 12.5 % (195) 17.3 % (18) 9.2 % (8) 1559 104 87 Early cognitive development 1-6 years Standard score 2 1 0 -1 -2 1y 2y 3.5 y 5.5 y Expressive vocabulary Good readers Poor readers 3.5 y 4.5 y 5.5 y 6.5 y 3.5 y 4.5 y 5.5 y 6.5 y Phonological awareness Average readers Poor comprehenders Letter knowledge Slow decoders 3.5 y 5.5 y 6.5 y RAN Book reading: alone and with parent 1 z -score 0,5 0 -0,5 -1 4y 5y 6y 7y 8y 2y Reading alone 4y 5y 6y 7y 8y Shared Reading Good readers (n = 28) Average readers (n = 70) Poor comprehenders (n = 16) Poor readers (n = 26) Slow readers (n = 51) In sum: - Already in grade 2 differentiation in fluency and comprehension - Poor readers with both fluency and comprehension problems had the poorest cognitive skills. - Poor comprehension linked particularly to vocabulary - Slow reading linked particularly to RAN - Family risk for dyslexia was a risk for reading speed but not for reading comprehension Grade 9 PISA reading comprehension Minna Torppa, Kenneth Eklund, Sari Sulkunen, and Pekka Niemi background PISA is today the most well-known assessment and comparison tool of adolescent’s reading skill in different countries around the world and its influence on the development of educational systems is enormous. All 34 OECD member countries and 31 partner countries and economies participated in PISA 2012 representing more than 80% of the world economy (OECD, 2013). Despite its widely accepted stature and influence on global educational debate and policy making, empirical research on explanations of the success and failure in PISA tasks are scarce participants N=1309 9th grade Finnish-speaking students from 95 classrooms All schools were typical Finnish schools organizing teaching according to national curriculum. PISA reading measure PISA reading literacy link items which are items used repeatedly in each cycle of the survey to ensure the comparability of the measurement (OECD 2010b, 26; 2013, 45). In the booklet, there were 8 different texts which the students were asked to read and answer several questions per material. The reading materials included texts, tables, graphs, and figures. Of the questions, 12 required students to access and retrieve information, 12 to integrate and interpret information, and seven to reflect and evaluate information. Students were allowed to use 60 minutes for completing the task. Cronbach’s alpha reliability coefficient for the total score in this sample was .80. Reading fluency (1) Sentence reading fluency task: The task was to read as many statements in 2 minutes and decide if the sentence was true by answering ‘yes’ or ‘no’. (2) Error search task: The task was to proofread words written on a sheet of paper and mark as many incorrectly spelled (either a wrong letter, extra letter or missing letter) words as possible in 3 minutes. (3) Word chains: The task was to mark with a pencil as many word boundaries (i.e. places, where one word ends and another one begins) as possible in 90 seconds. Cronbach’s alpha reliability coefficient for reading fluency composite score was .78. Motivation Reading self-concept: Students were asked to evaluate their reading skills in comparison to their peers. Three items were included, one for reading accuracy, one for reading fluency, and one for reading comprehension. Cronbach’s alpha reliability coefficient was .75. Task avoidance: The children’s task-avoidant behavior was measured using the Behavioral Strategy Rating Scale (BSR; Onatsu & Nurmi, 1995; see also Zhang, Nurmi, Kiuru, Lerkkanen, & Aunola, 2011). The measure includes five statements concerning behavior when facing difficult tasks. Cronbach’s alpha .79 Persistence: The children’s task-avoidant behavior was measured using the Behavioral Stratety Rating Scale (BSR; Onatsu & Nurmi, 1995; see also Zhang, Nurmi, Kiuru, Lerkkanen, & Aunola, 2011). Four statements. Cronbach’s alpha .71 Leisure-time reading: how often they read different printed materials with a five-point Likert –scale-Three items, one for books, one for comics, and one for newspapers and magazines. Grade 9 SEM for explaining PISA (χ2 (145) = 321.02, p= .00, CFI=.97, TLI=.96, RMSEA=.03, SRMS=.04 Some general conclusion Reading as decoding and comprehension Developmental paths and varying pace of development Family riskfor dyslexia and cognitive skills (RAN, LK, PA) are strong predictors of decoding, but not so much for comprehension Also persistence and leisure time reading are linked to reading skills Psychosocial functioning and dyslexia: no evidence for dyslexia being a risk factor for psychosocial functioning in the early grades. Thank you! [email protected] JLD & Dyslexia diagnosis Exclusion criteria: Standard score of < 80 in both performance or verbal IQ (assessed with WISC-R at 8 years e.g. 2nd grade) was applied, but all participants had scores above the criterion. 8 Measures: 4 ACCURACY measures: Word and pseudo-word reading, Word and pseudo-word spelling, Text reading, Pseudo-word text reading 4 FLUENCY measures: Word and pseudo-word reading, Text reading, Pseudo-word text reading, Word list reading A child was considered to have deficient skills in each respective task if his or her score fell to the 10th percentile or below (JLD control group distribution). To be classified with Dyslexia the child’s skills were at or below the 10th percentile either a) in at least three out of the four accuracy measures or at least three out of the four fluency measures, or b) in two accuracy measures and in two fluency measures.