Examining Hypermedia as a Means to Improve Access to the General Education Curriculum for Students with Reading Difficulties

1
Examining Hypermedia as a Means to Improve Access
to the General Education Curriculum for Students
with Reading Difficulties

• Dr. Matthew T. Marino
• Washington State University
• 2007

2
Overview

• Context for this investigation
• Students with LD in reading and poor readers
• Adolescents and expository texts
• Barriers to the accommodations process
• Improving reading comprehension
• Technology in education
• Can hypermedia improve access to the general
  education curriculum?
• Current Research

3
Students with Learning Disabilities (LD)

• Students with LD comprise the largest subgroup of
  students served under IDEA (U.S. Department of
  Education, 2002)
• The number of students age 12 to 17 identified
  with LD has increased 44 in the past 10 years
  (Lyon et al, 2001)
• Only 62 of students with LD graduate with a
  diploma (U.S. Department of Education, 2002)
• Educational costs for students with LD are 1.6
  times that of regular education students (Special
  Education Expenditure Project SEEP, 2003)
• Students served under IDEA must be provided with
  access to the general education curriculum to the
  greatest extent possible (IDEA, 1997)

Review of Literature
4
LD Identification Process Implications

• IQ-achievement discrepancy identification is
  changing
• LD in reading defined
• IQ does not strongly correlate with reading
  achievement or rate of reading development
  (Share, McGee, Silva, 1989 Vellutino, Scanlon,
  Lyon, 2000)
• IQ does not predict a students ability to read
  or profit from remediation (Siegel, 1988
  Vellutino et al., 2000)

Review of Literature
5
LD in Reading and Poor Readers

• Poor readers defined (Vellutino et al., 2000)
• Poor readers show many of the same
  characteristics as students with LD in reading
  (Fletcher et al., 1994 Stanovich and Siegel,
  1994)
• Students with LD in reading and poor readers
  possess virtually indistinguishable reading
  growth curves in grades 1 through 12 (Vellutino
  et al., 2000)
• For the purposes of this study, students with LD
  in reading will be combined with poor readers and
  referred to as students with reading difficulties
  (RD)
• Students with RD in this study scored below the
  25 on standardized measures of reading
  achievement during the academic year prior to
  this study.

Review of Literature
6
Statement of the Problem

• More than 12.4 million students experience
  significant difficulties learning to read
  (National Center on Educational Statistics,
  2003).
• Students with RD are failing to make adequate
  yearly progress in the general education
  curriculum (Lyon et al., 2001 Mastropieri,
  Scruggs, Graetz, 2003 U.S. Department of
  Education, 2003).
• Traditional methods of instruction (e.g.,
  lectures based on readings in expository texts)
  are not effective for students with RD (Lapp,
  Flood, Ranck-Buhr, 1995 Maccinin, Gagnon,
  Hughes, 2002).
• The number of students with RD in inclusive
  classrooms is increasing (Lyon et al. 2001).

Review of Literature
7
Adolescents and Expository Texts

• Expository text defined
• Students with RD
• Often lack prior knowledge (Gersten, Fuchs,
  Williams, Baker, 2001)
• Are unaware of the text structures they are
  reading (Meyer, Brandt, Bluth, 1980)
• Retrieve information randomly (Wilson Rupley,
  1997)
• Have difficulty determining essential information
  (Engert Thomas, 1987)
• Do not utilize text cues (Gersten et al, 2001)
• Fail to recognize when they are not comprehending
  new information (Gersten et al., 2001)

Review of Literature
8
This Leads To

• Low levels of reading comprehension (Gersten et
  at., 2001)
• An inability to formulate questions and
  hypotheses (Wilson Rupley, 1997)
• Failure to make abstract connections (Engert
  Thomas, 1987)
• Frustration, lower motivation, expected failure
  (McKinney, Osborne, Schulte, 1993)

What do we do for these students?
Review of Literature
9
Barriers to the Accommodations Process

• Teachers at the secondary level feel pressured to
  progress through the curriculum (Mastropieri,
  Scruggs, Graetz, 2003)
• Teachers ability to provide meaningful
  accommodations is hampered by large class sizes
  and a lack of resources (Lancaster, Schumaker,
  Deshler, 2002)
• Many general education teachers do not have the
  time or expertise to provide meaningful
  accommodations (Mastropieri et al., 2003)

How can we improve this process?
Review of Literature
10
Improving Reading Comprehension

• Modify instructional materials to match the
  students reading ability (Mastropieri Scruggs,
  2004)
• Present information using multiple modalities
  (MacArthor, Ferretti, Okolo, Cavalier, 2001)
• Offer repeated reading and practice opportunities
  (Jenkins, Stein, Wysocki, 1984 Bryant, 2003)
• Provide students with opportunities to gain
  additional background knowledge quickly
  (MacArthor et al., 2001)

How can we do this?
Review of Literature
11
Consider Technology

• Current policy calls for the increased use of
  technology to support student learning (U.S.
  Department of Education, 2003)
• Technology use in regular education classrooms is
  rapidly increasing (U.S. Department of Education,
  2003 Vannatta OBannon, 2002 )
• Technology may improve access to the general
  education curriculum (Pucket, 2004 Behrmann
  Jerome, 2002 Edyburn, 2002 Fisher Frey, 2001)
• Technology should be used in concert with other
  instructional methods (e.g., classroom discourse)
  (Yerrick, 2000 De La Paz MacArthor, 2003
  Fuchs, Fuchs, Kazdan, 1999)

Review of Literature
12
Can Technology Improve Access to the General
Education Curriculum?

• Hypermedia defined
• Hypermedia may eliminate the overuse of
  expository texts (Lancaster et al., 2002)
• Hypermedia provides
• Information on demand (McKenna, Reinking, Labbo,
  Keiffer, 1999)
• Tools that support cognitive processes (Lajoie,
  1993)
• Built-in accommodations by allowing teachers to
  modify task difficulty and select appropriate
  readability levels (Edyburn, 2000 Behrmann
  Jerome, 2002 Pucket, 2004)

Review of Literature
13
Hypermedia Research

• Hypermedia allows teachers to monitor student
  progress and resource use (McKenna et al., 1999)
• Students and teachers report positive outcomes
  from using hypermedia (Lancaster et al.,, 2002
  Garthwait, 2004 Lewis, 2000)
• There are a limited number of studies examining
  the efficacy of hypermedia as a means of
  improving comprehension for students with RD
  (Maccini, Gagnon, Hughes, 2002)
• Research with students in regular education
  classrooms is promising, but inconclusive
  (Oliver, 1999 Land, 2000 Liu, 2004)

Review of Literature
14
Need for Current Research

• Preliminary research examining the effects of
  technology-based textual modifications (e.g.,
  readability levels within hypermedia programs) is
  inconclusive (MacArthur et al., 2001).
• Research is needed to determine the types of
  comprehension instruction that are most
  beneficial for low ability readers in secondary
  content area courses (Report of the National
  Reading Panel, 2000).
• Additional research is needed to determine which
  cognitive tools are most beneficial to students
  with RD in middle school science classes (Liu,
  2004).

Review of Literature
15
Theoretical framework

• Knowledge Construction Through Conceptual Change
  (KCTCC) expands on the theory of Schema Change
  (Winn, 2004).
• Perpetual cycle of learning where prior knowledge
  directs how individuals seek, identify, and
  interpret information (Neisser, 1976).
• The way in which individuals construct knowledge
  can not be predicted or supported using teacher
  directed instructional designs (Duffy Jonassen,
  1992).
• KCTCC is most effective in problem-based learning
  environments that incorporate technology (Winn,
  2004).

Review of Literature
16
Universal Design for Learning (UDL)

• UDL supports KCTCC by
• Assuming students enter learning experiences with
  varying degrees of prior knowledge (Hitchcock,
  Meyer, Rose, Jackson, 2002).
• Using technology to incorporate video clips,
  graphic organizers, illustrations, and text
  modifications into the curriculum (Mastropieri,
  Scruggs, Bakken, Whedon, 1996).
• Allowing learners to access information quickly
  using multiple modalities (MacArthur, Ferretti,
  Okolo, Cavalier, 2001).

Review of Literature
17
Why Middle School Science?

• Science is one of the most difficult subjects for
  students with RD to learn due to its complex
  vocabulary and theoretical nature (Mastropieri et
  al., 2003).
• Students with RD in middle school science classes
  typically read at the 4th or 5th grade level
  (Mastropieri et al., 2003).
• If textual modifications and cognitive tools can
  improve access to the middle school science
  curriculum, their application in other content
  areas may yield similar results.

Review of Literature
18
Research Purpose

• Statement of the research problem
• There is a need to determine whether hypermedia
  can improve access to the general education
  curriculum for students with RD and low ability
  readers (Lyon and Moats, 1997 MacArthur et al.,
  2001 National Reading Panel, 2000).
• For the purposes of this study
• Low ability readers are defined as students
  scoring lt 50 on the Degrees of Reading Power
  (DRP) subtest of the Connecticut Mastery Test.
  The DRP is a norm-referenced measure of reading
  achievement (Touchstone Applied Science
  Associates, 2004).
• Students with reading difficulties (RD) are
  defined as scoring lt 25 on the DRP.

Research Questions
19
Research Questions

• Are there posttest differences on the science
  posttest and solutions forms measure between
  students with RD (lt25th percentile) and those who
  are not RD (25th - 50th percentile)?
• Are there posttest differences on the science
  posttest and solution forms measure between
  students who participated in the 4th grade
  readability condition and those who participated
  in the 8th grade readability condition?
• Is there an interaction between reading ability
  (i.e., students lt25th percentile and students in
  the 26th - 50th percentile) and treatment
  condition (4th grade readability and 8th grade
  readability)? If so, what is the nature of the
  interaction?
• Is there a relationship between students reading
  ability, use of cognitive tools, and their
  comprehension of scientific concepts and
  processes as measured on the posttest and
  solutions form measure? If so, what is the nature
  of the relationship?

Research Questions
20
Research Design

Methods Design
21
Threats to Validity

Methods Design
22
Setting

• Four middle schools in New England volunteered to
  participate in the study
• Schools were selected based on 1) administrative
  consent, 2) 100 teacher agreement to
  participate, and 3) available technology
  resources

Setting Participants
23
Student Demographic Data

• 50 male, 50 female
• 87 White, 5 African American, 4 Asian, 4
  Hispanic
• 89 have a computer at home they can use
• 54 have a personal computer
• Response to the statement
• I am good at using computers
• 32 strongly agree
• 62 agree
• 5 disagree
• 1 strongly disagree

Setting Participants
24
Participants

• Students (N 1153) were grouped based on 2004
  DRP scores
• Group 1 - Students with RD ( lt 25 on DRP)
• Group 2 - Poor readers (26 - 50 on DRP)
• Group 3 - Proficient readers ( gt 50 on DRP)
• Students from groups 1 (n 113) and 2 (n 189)
  were randomly assigned at the student level
  within each class to either the 4th grade
  readability or 8th grade readability condition.
• Students in group 3 (n 851) received text from
  the program at the 8th grade level

Setting Participants
25
The Intervention Alien Rescue

Instrumentation
26
Performance Measures

• Pre/Posttest
• 25 item paper pencil multiple choice test
• Reliability of .85 established in previous study
  (Pedersen Williams, 2004)
• 19 items assess knowledge and comprehension
  (e.g., A world would have a magnetic field
  if________ )
• 6 items assess students ability to apply
  knowledge (e.g., Scientist want to measure Mars
  atmosphere. What instrument would they use ______
  ?)

• Six solution forms
• One paper and pencil solution form for each alien
  species
• Two column form that requires students to
  analyze, synthesize, and evaluate data from the
  Alien Rescue program
• Scaffolds the learning process through prompts
  (e.g., magnetic field)
• Contains open-ended and narrative response items
• Piloted with 300 students (Marino, 2005)
• Established inter-rater reliability of .90

Instrumentation
27
Procedures

Methods
28
Preliminary Analyses

• Students who were absent more than 3 days (20)
  of the intervention were excluded from analyses.
• Independent samples t-test results, t (302)
  1.22, p .22, d .01 indicated that there were
  no significant differences (DRP group 1 vs. 2) on
  the pretest measure.
• Principal component factor analysis of posttest
  measures (posttest solutions forms) indicates
  68 of variability was explained by posttest.
• Solutions forms grouped into one subscale (total
  score) with one factor solution reliability of
  .94.

Analysis
29
Analysis Research questions 1 - 3

• Compute descriptive statistics
• Separate two-way ANOVAs
• DRP group (DRP group 1 vs. DRP group 2)
  and treatment condition (4th grade text vs.
  8th grade text) as between-subjects
  independent variables. Posttest and combined
  solutions forms scores (total score) as
  dependent variables.

Analysis Results
30
Results Research Questions 1 - 3

• RQ1 Differences by DRP group
• Results of the ANOVA for the posttest were not
  significant
• F (1, 298) 3.552, p .060, d 0.01
• Results of the ANOVA for total score were
  significant
• F (1, 281) 3.974, p .047, d 0.01
• RQ2 Differences by treatment (4th vs. 8th grade
  text)
• Results of the ANOVA for the posttest were not
  significant
• F (1, 298) 0.369, p .554, d .001.
• Results of the ANOVA for total score were not
  significant
• F (1, 281) 0.03, p .872, d lt .001.
• RQ3 Interaction between treatment and DRP group
• Results of the ANOVA for posttest were not
  significant
• F (1, 298) 2.56, p .111, d .008
• Results of the ANOVA for total score were not
  significant
• F (1, 281) 1.28, p .259, d .004

Results
31
Student Learning

• Paired samples t-test (pre/posttest) with DRP
  groups collapsed was significant (t 25.719,
• p lt .001, d 1.5)
• Posttest mean was 14.22 (56) on a 25-point scale
• On Solutions Forms (216-point scale)
• DRP 1 mean score 142.05 (66)
• DRP 2 mean score 151.77 (70)

Did they use the built-in cognitive tools?
Results
32
Analysis Research Question 4

• Group cognitive tools into four categories
  (Lajoie, 1993 Liu, 2004 Liu Bera, 2005)
• Tools that share cognitive overload (e.g.,
  databases)
• Tools that support cognitive process (e.g., field
  journal)
• Tools that support out-of-reach activities (e.g.,
  astroengineering)
• Tools that support hypothesis testing (e.g.,
  telemetry)
• Obtain correlations between students use of
  tools and performance on posttest and total score
• Separate one-way ANOVAs with tool use category as
  the dependent variable and DRP group as
  independent variable.
• Simultaneous multiple linear regression analysis
  to determine how tool use (by category) predicts
  performance on posttest and total score measures

Analysis Results
33
Results Research Question 4

• The strongest correlation was between tools that
  share cognitive overload and the posttest (r
  .224).
• ANOVA (tool use DV, DRP IV) was significant for
  each of the four tool categories.
• Cog. overload F (2, 954) 12.60, p lt .001,
  d .03
• Cog. process F (2, 954) 4.86, p
  .008, d .01
• Out-of-reach F (2, 954) 3.07, p
  .047, d .006
• Hypothesis F (2, 954) 5.57, p
  .004, d .01
• Scheffes post hoc analysis indicates significant
  differences between students by DRP groups

Analysis Results
34
Results of Multiple Regression - RQ4

• Multiple regression indicates main effects for
  tools that share cognitive overload and tools
  that support out-of-reach activities on posttest
  scores.
• A significant interaction between DRP group and
  tools that support cognitive overload was present
  on the posttest regression analysis.
• A significant interaction between DRP group and
  hypothesis testing was present on the solution
  form regression analysis.

What does this tell us?
Analysis Results
35
Interpreting the Multiple Regression Analysis

• On the Posttest Regression Analysis
• For every unit increase in the SQRT of cognitive
  overload
• We predict a .792 unit increase in posttest score
  for students in DRP groups 1 2.
• We predict a .33 unit increase for students in
  DRP group 3.
• For every unit increase in the SQRT of tools that
  support out-of-reach activities students scores
  on the posttest decrease by .674 units.
• On the Solutions Form Regression Analysis
• Students in DRP 3 gained 7.886 units for each
  unit gain using tools that support hypothesis
  testing.
• Students in DRP groups 1 2 gained 1.264 units
  for each unit gain using tools that support
  hypothesis testing.

Analysis Results
36
Discussion

• Students in DRP groups 1 and 2 (lt50) performed
  in similar ways during the intervention.
• There was a lack of treatment effect due to
  readability level of text.
• Students used cognitive tools differently
  depending on their reading ability level.

Discussion
37
Implications

• Students need systematic, explicit, scaffolded
  instruction to utilize cognitive tools (Gersten
  Baker, 1998)
• Opportunities for students to reflect on and have
  immediate feedback regarding tool use (Swanson
  Hoskyn, 1998)
• Inclusion of tools that support cognitive
  overload and out-of-reach activities (Land, 2000
  Liu, 2004 Williams Peterson, 2004)
• Technology and content area training for special
  education teachers (Sharpe Hawes, 2003)
• Increased collaboration between special education
  and regular education teachers (Moore Keefe,
  2001)
• Improved instructional strategies for general
  education teachers (Washburn-Moses, 2005)

Implications
38
Limitations

• Paper and pencil assessments
• Levels and types of classroom discourse
• Tool use data highly skewed
• Student differences with problem-based learning
• Sample (SES, representation of minority
  populations)
• Teacher effects

Limitations
39
Questions?

40
Links

• Descriptives by DRP group
• Descriptives by Treatment Condition
• Two-Way ANOVA Posttest
• Two-Way ANOVA Total Score
• Tool Category Correlations
• Posttest Multiple Regression
• Total Score Multiple Regression
• Scheffes Post Hoc for Tool Use
• Threats to Validity
• Cell sizes for low ability readers
• Teacher Effects
• Research Questions

41
Descriptive Statistics by Reading Ability

Maximum Posttest Score 25 Maximum Total Score
216
RQ 1 -3
Results
42
Descriptive Statistics by Treatment Condition

Maximum Posttest Score 25 Maximum Total Score
216
RQ 1 - 3
Results
43
Posttest Two-Way ANOVA

RQ 1 -3
Results
44
Total Score Two-Way ANOVA

RQ 1 - 3
Results
45
Tool Category Correlations

RQ 4
Results
46
Scheffes Post Hoc Analysis for Tool Use

RQ 4
Results
47
Multiple Regression Posttest

MR 4
Results
48
Multiple Regression Total Score

MR 4
Results
49
Teacher Effects

• Results of one-way ANOVA with posttest and
  solution form scores as dependent variables and
  Teacher as the independent variable were
  significant.
• Posttest
• F (15, 1190) 16.131,
• p lt .001, d 0.17
• Solutions Forms
• F (15, 1153) 112.334,
• p lt .001, d .59

Results
50
Cell sizes for low ability readers

Group 1 Group 2
Text presented at 4th grade level
Text presented at 8th grade level
Power analysis indicates optimum group sizes of
68 (f .20, d .57) to achieve a power of .80.
Methods