Genetic%20and%20Environmental%20Influences%20on%20Reading%20(SWIDA,%202/11/00)

1
Genetic and Environmental Influences on Reading
Ability and Disability (Difficulty) Richard
Olson, University of Colorado, Boulder
2

• Science Announces 2007 Breakthrough of the Year
  Human Genetic Variation After years of research
  into genetic similarities among people and even
  apes, research advances on several fronts during
  2007 defined the degree to which genomes differ
  from one human to another. This breakthrough
  represents a conceptual leap that allows
  discovery of individual risk for disease and
  specific medical treatment. Science recognized
  "Human Genetic Variation" as the leading
  breakthrough of 2007

3

• In 2007, researchers were dazzled by the degree
  to which genomes differ from one human to another
  and began to understand the role of these
  variations in disease and personal traits.

4
Colorado Learning Disabilities Difficulties
Research Center (CLDRC)

• Twin Studies (Psychometric assessment) (John
  DeFries, Sally Wadsworth, Erik Willcutt)
• ADHD and executive function (Bruce Pennington,
  Erik Willcutt)
• DNA Linkage analysis and physical mapping
  (Shelley Smith)
• Response to Computer Assisted Reading Instruction
  (Barbara Wise, Brian Byrne, Ron Cole, Sarel van
  Vuuren)
• Reading and language processes (Jan Keenan,
  Richard Olson)

5
How do we know if a personal trait is
influenced by genes?

• Because we have identified the responsible gene
  or genes from molecular genetic analyses?
• Because the trait runs in families?
• Because there are two kinds of twins.

6
Why Study Twins?

• Monozygotic (MZ) twins share 100 of all their
  genes
• Dizygotic (DZ) twins share 50 of their
  segregating genes (genes that make us different),
  on average
• MZ and DZ pairs reared together have similar
  shared environment in the home and school
  (assumption of analyses)
• (But, MZ twins may create more similar
  environments, such as print exposure, because of
  their genes, a genotype-environment correlation

7
Twin Data Can Estimate Population Averages for
Three Causes of Individual Differences

• Heritability (genes) genetic influences.(does
  not tell us which genes or individuals)
• Shared (Common) Environment environmental
  influences that affect similarly both twins in a
  pair.(prenatal, home, school)
• Nonshared Environment environmental influences
  that are specific to individual twins.(accident,
  teacher, measurement error)

8
Our behavior-genetic analyses of twin data assume
a normal distribution (the bell curve)
number of people
Low Reading High Reading
Population Mean
9

• Twins TOWRE Sight Word Efficiency Standard
  Scores
• at the End of First Grade

10
Such normal distributions allow us to estimate
genetic and environmental influences by comparing
correlations for MZ and DZ twins

• So lets look at some correlations

11
P1
P2

• Expected correlation for randomly paired
  individuals

12
T1
T2

• Expected DZ correlation if all influence due to
  genes

13
T1
T2
Expected MZ correlation if all influence due to
genes
14
T1
T2
Actual MZ correlation for end of first grade word
reading efficiency
15
International Longitudinal Twin Study of Early
Reading, Language, and Attention
Development Brian Byrne, Richard Olson, Stefan
Samuelsson
16
Twins seen in homes or pre-schools at age 4,
prior to kindergarten
17
Twins typically tested at home during summer
break after kindergarten
18
And again during summer break after first grade
19
With a final assessment after second grade
20
Colorado Sample Characteristics for MZ and
Same-Sex DZ Twins

• Representative sample of 489 twin pairs
  ascertained from birth records, all English first
  language.
• Most from Colorado Front-Range (Colorado
  Springs-Denver-Boulder-Fort Collins).
• First tested at preschool age 4-5, followed up at
  end of kindergarten, 1st, 2nd, while children are
  learning to read, and 4th grade while children
  are reading to learn

21
Genetic and Environmental Influences on Preschool
Print Knowledge in Colorado (Samuelsson et al.,
2007)
(95 confidence intervals in parentheses)
22
Kindergarten TOWRE Word Reading Efficiency in
Colorado (Samuelsson et al., 2007)
23
Genetic and Environmental Influences on
Individual Differences in Preschool and
Kindergarten Reading

• Shared family environment influence is the major
  influence on preschool print knowledge in all
  countries
• Genetic influence on reading increases in
  kindergarten, but varies depending on amount and
  consistency of kindergarten instruction which is
  high in Australia (heritability 89), medium in
  Colorado (heritability 61), and low in
  Scandinavia (heritability 38)
• The country differences in kindergarten
  heritability are related to when reading is
  taught.

24
End-of-First Grade Word Reading Efficiency in
Colorado (Byrne et al., 2007)
25
Conclusions about genetic and environmental
influences on the early development of word
reading efficiency in Colorado

• Shared environment influences are dominant for
  print knowledge at preschool age
• Genetic influence begins its dominance for word
  reading efficiency by the end of kindergarten,
  but shared environment influence remains
  significant
• Genetic influence becomes strongly dominant by
  the end of first grade, as children vary in their
  responses to formal instruction in their schools
  (RTI!)
• We see a similar developmental pattern for
  spelling from kindergarten through first grade

26
End of Kindergarten Spelling in Colorado (Byrne
et al., 2007)
Shared Environment .40 (.18, .57)
27
End of First Grade Spelling in Colorado (Byrne
et al., 2007)
28
End of First Grade Woodcock-Johnson III Reading
Comprehension in Colorado (Byrne et al., 2006)
29
1st grade Woodcock Reading Comprehension in
Australia, Colorado, and Ohio (Petrill)
30
G x E interaction with parents average years of
educ. for individual differences in TOWRE word
rec. at the end of kindergarten in the Colorado
sample
genetic
Shared Env.
Non Shared Env.
Low Years of Ed. High
31
A2, C2 and E2 Estimates as a Function of Parental
Ed. In First Grade
32
Are there different genes influencing different
reading skills?

• Genetic correlations are above .9 for word
  reading, spelling, and reading comprehension at
  the end of first grade (Byrne et al., 2006)
• However, there is partial independence for
  genetic influences on word recognition and
  reading comprehension skills in older children.
  (Keenan et al., (2006) Is starting to figure it
  out)

33
Implications for Public Policy on Education

• By the end of first grade, genes account for
  between 65 and 80 percent of individual
  differences in word reading, spelling, and
  reading comprehension in Australia, Colorado, and
  Scandinavia
• But No Child Left Behind makes no mention of
  genetic influence, and instead blames teachers
  and schools for childrens reading failure
• which may be true in extreme cases, but what is
  their average influence on individual differences
  in reading at the end of 1st grade in Colorado?

34
Failing Teacher Influences on Individual
Differences in Reading?

• Jonathan Alter (Newsweek, Feb. 12, 2007)anyone
  with an ounce of brains knows what must be done.
  Its time to move from identifying failing
  schools to identifying failing teachers.
• But if teachers are the main influence on
  individual differences in early reading
  development, on average, twins sharing the same
  teacher should be much more similar than twins
  who have different teachers

35
Colorado Correlations for MZ Pairs in Same (N74)
Versus Different (N108) 1st Grade Classes
No significant differences by same different
class
36
Colorado Correlations for DZ Pairs in Same (N83)
Versus Different (N132) 1st Grade Classes
No significant differences by same different
class
37
U.K. TOWRE Correlations for MZ and DZ Pairs in
Same Versus Different 1st Grade Classes
Pairs 1038, 578 939,
528 (Nicole Harlaar, Personal Communication,
9/10/2006)
38
Teacher effects on individual differences in
reading and spelling at the end of 1st grade

• Very similar correlations for MZ twins sharing or
  not sharing the same 1st grade teacher suggest
  minimal average teacher effects on individual
  differences (of course teachers do teach the kids
  to read)
• So anyone with an ounce of brains(Jonathan
  Alter)
• However, both twins in a pair go to the same
  school, and maybe that obscures a real teacher
  effect because bad schools have bad teachers
  (suggeston from Ed Steinberg, Special Ed.
  Director, Colorado Department of Education).
• So lets look at mean school performance effects
  on individual differences in our twins reading
  scores.

39
A similar conclusion about teacher effects from

• Mehta, P.D., Foorman, B.R., Branum-Martin, L.,
  Taylor, P.W. (2005). Literacy as a unidimensional
  multilevel construct Validation, sources of
  influence, and implications in a longitudinal
  study in grades 1-4. Scientific Studies of
  Reading, 9, (2), 85-116.
• Now for school effects on individual differences

40
The twins mean school performance is assessed by
the third-grade Colorado Student Assessment
Profile (CSAP), McGraw-Hill

• 32 multiple choice and 8 constructed response
  items assessing
• Use of word recognition for comprehension
• Identification of main idea
• Drawing of inferences
• SummarizationCronbach Alpha .89

41
Standard CSAP score means for the schools
attended by our twins
Mean 573 Std.Dev. 26 (Note that the
between-school standard dev. of 26 is much
less than the average within-school standard
dev. of 70)
42
School CSAP correlations with 2nd grade reading
and language in green, (second order r in blue,
partialed by parent education)
43
Parent education correlations with 2nd grade
reading and language, 2nd order r partialed by
school CSAP PA
44
School effects on reading at the end of second
grade

• Our twins reading and spelling correlations with
  the average 3rd grade school score on the state
  test are low (.12 - .14) but statistically
  significant
• But they are not significant and close to 0 after
  controlling for their parents years of
  education
• Parent education correlations with their twins
  reading are not moderated by mean school score
• There may be small average school effects in our
  sample, but we cant separate them from the
  effects associated with parent education

45
Similar conclusions from

• McCoach, D.B. et al. (2006). Growing readers A
  hierarchical linear model of childrens reading
  growth during the first two years of school.
  Journal of Educational Psychology, 98, 14-28.
  between school differences in achievement
  kindergarten and first grade in the Early
  Childhood Longitudinal Study are largely
  explained by differences in school clientele
  SES, rather than differences in instruction or
  resource allocation.

46
Conclusions continued

• But, school quality of instruction has been
  experimentally manipulated with significant
  effects on school CSAP scores (Pueblo 60, Sadoski
  Willson, 2006)

47
Percent of Pueblo 60 Third Grade Proficient
CSAP Scores 1998-2005
48
Overlapping 3rd grade CSAP Distributions for
Pueblo 60 and Comparable District
Pueblo 579 (SD57)
District? 530 (SD70)
300 400 500 600
700 800
(573 is the State average for non ESL kids)
49
Pueblo 60 Decoding Focus Students Mean Pre- and
Post-test Percentiles after 60 hours of intensive
additional instruction
50
No child left behind All children must read at
grade level
before
after
?
dyslexics
Elimination of CocaCola (Whole Language) from
childrens diets (education) results in massive
improvements in reading (Adapted from Byrne,
2005).
51
Implications for raising literacy in the
population as a whole and among children with
learning disabilities

• We can certainly raise the average level of
  literacy in the U.S. by reading more and better
  text in and out of school.
• We can also improve reading skills among children
  with dyslexia with more intensive instruction and
  greater reading practice, but

52
Two take-home lessons

• Most cases of dyslexia (and individual
  differences) that occur within normal educational
  environments are not the fault of the child,
  their parents, their teachers, or their schools
• It is unrealistic to expect all children to reach
  grade level, as the No Child Left Behind law
  insists they must, because some children may have
  strong genetic or other biological constraints
  on their reading development

53
From the No Child Left Behind reauthorization
web site HIGHLIGHTS OF BUILDING ON
RESULTS Every Child Performing at or Above
Grade Level by 2014 AccountabilityStates will
be held accountable for ensuring that all
students can read and do math at grade level by
2014. http//www.ed.gov/policy/elsec/leg/nclb/fa
ctsheets/blueprint.html
54
From Robert Linn, co-director,National Center
for Research on EvaluationUCLA

• There is a zero percent chance that we will ever
  reach a 100 percent target.But because the title
  of the law is so rhetorically brilliant,
  politicians are afraid to change this completely
  unrealistic standard. They dont want to be
  accused of leaving some children behind.(from
  Washington Post, 3/14/07)

55
From the U.S. Secretary of Education, Margaret
Spellings

• In the past, we underestimated what students
  with disabilities could learnWe now know that
  the vast majority of these children can achieve
  grade-level standards.

• http//www.ed.gov/policy/elsec/leg/nclb/factsheets
  /blueprint.html

56
From U.S. Deputy Education Secretary Raymond Simon

• We need to stay the coursethe mission is
  doable, and we dont need to back off that right
  now.(Amit R. Paley,Washington Post,
  3/14/2007)

57
From Sen. Lamar Alexander, former U.S. Education
Secretary

• Are we going to rewrite the Declaration of
  Independence and say only 85 of men are created
  equal? (Amit R. Paley,Washington Post,
  3/14/2007)

58
From Daniel Koretz, Harvard Professor of Education

• Most people are afraid that once you acknowledge
  this variation differences among children of the
  same racial or socioeconomic background, then
  you have to tolerate major inequities between
  black and white students. Thats not
  necessarily true, but thats why the political
  world does not really address the issue.
• (by Amit R. Paley,Washington Post,
  3/14/2007)
• See also, Steve Pinkers Blank Slate, Ch. 8, fear
  of inequality