THE NEMETH CODE TUTORIAL: A FINAL REPORT

1
THE NEMETH CODE TUTORIAL A FINAL REPORT

• Gaylen Kapperman
• Jodi Sticken
• Visual Disabilities Program
• Department of Teaching and Learning
• Northern Illinois University
• DeKalb, IL 60115

2
The Nemeth Code Tutorial for the BrailleNote

• Interactive software for the BrailleNote, an
  electronic notetaking device manufactured by
  Humanware Group
• Designed to assist blind students in learning to
  read and write the code of braille mathematics
• Software was field tested by blind students and
  their teachers over a 2-year period

3
Field Test, Year One

• Assessment of each students ability to read and
  write Nemeth code was conducted at the beginning
  of the school year (pre-test)
• Students in the treatment group received
  instruction in the use of the tutorial
  instruction was delivered by their special
  education teachers
• Students in the control group were not exposed to
  the software
• Assessment of each students ability to read and
  write Nemeth code was conducted at the conclusion
  of the school year (post-test)

4
Year Two

• Assessment of each students ability to read and
  write Nemeth code was conducted at the beginning
  of the school year (pre-test)
• Students in the control group received
  instruction in the use of the tutorial
  instruction was delivered by their special
  education teachers
• Students in the treatment group were allowed to
  continue their work with the tutorial if they
  chose to do so
• Assessment of each students ability to read and
  write Nemeth code was conducted at the conclusion
  of the school year (post-test).

5

• The data analysis shown in the following slides
  represents the outcome of the first years
  activities
• The results of the data analysis indicate that
  the tutorial is a very effective tool in helping
  blind students learn to read and write the code
  of braille mathematics

6
Sample

• Treatment and control group students were matched
  based on
• Grade level
• Math grade level
• Reading grade level
• Discrepancy between actual grade level and math
  grade level
• One student from control group was excluded from
  analyses to facilitate matching

7
Sample (ctd.)

• Treatment group (received Nemeth code training)
  n 28
• Mean age 13.3
• Mean grade level 6.9
• Mean math grade level 6.7
• Mean reading grade level 6.4
• 54 male, 46 female
• 71 Caucasian, 11 African-American, 14
  Hispanic, 4 Other ethnicity
• Control group (did not receive Nemeth code
  training) n 28
• Mean age 12.7
• Mean grade level 6.4
• Mean math grade level 6.1
• Mean reading grade level 6.2
• 50 male, 50 female
• 79 Caucasian, 7 African-American, 11 Hispanic,
  1 Other ethnicity

8
Instrumentation

• Math Reading Test
• 75 items
• Three raters scored each item
• Mean consistency of raters across items 96
  (pre-test), 92 (post-test)
• Total score number of correct items
• Math Writing Test
• 75 items
• Three raters scored each item
• Mean consistency of raters across items 95
  (pre-test), 92 (post-test)
• Total score number of correct items

9
Instrumentation (ctd.)

• Split-half reliability

10
Descriptive Statistics
11
Descriptive Statistics (ctd.)
12
Distribution of Growth Scores (Treatment Group)
13
Distribution of Growth Scores (Control Group)
14
Math Reading Test Scores
15
Math Writing Test Scores
16
Math Reading Growth

• Difference in growth between treatment and
  control groups was statistically significant
  (t(27) 2.58, p .016), with a moderate effect
  size (d 0.49)

17
Math Writing Growth

• Difference in growth between treatment and
  control groups was statistically significant
  (t(27) 5.37, p lt .001), with a large effect
  size (d 1.01)

18
Non-parametric Tests

• Because distribution of growth scores showed some
  skewness, we additionally carried out
  non-parametric tests for differences in growth
• Results again showed significant differences in
  growth for Math Reading (p .014) and Math
  Writing (p lt .001)

19
Variability

• Treatment group showed more variability in growth
  scores than control group

20
Possible Reasons for Variability in Growth

• Gender
• Ethnicity
• Age
• Time spent on tutorial

21
Gender Differences in Growth
Difference between males and females was
non-significant for both Math Reading Growth
(t(26) 0.06, p .95) and Math Writing Growth
(t(26) 0.58, p .57).
22
Ethnic Differences in Growth
Difference between minority students and
non-minority students was non-significant for
both Math Reading Growth (t(26) 0.22, p .83)
and Math Writing Growth (t(26) 0.82, p .42).
23
Relationship of Student Age to Math Reading Score
Correlation was statistically non-significant (r
.19, p .33)
24
Relationship of Student Age to Math Writing Score
Correlation was statistically non-significant (r
.09, p .67)
25
Relationship of Total Time Spent on Tutorial to
Math Reading Score
Correlation was statistically significant (r
.52, p lt .01)
26
Relationship of Total Time Spent on Tutorial to
Math Reading Score
Correlation was statistically significant (r
.46, p .01)
27
Credits

• The following individuals contributed
  significantly to this project, and we appreciate
  their efforts
• Professor Thomas Smith data analysis
• Julie Hart, project assistant organization of
  the countless minutiae, and maintaining lines of
  communication between field testers and
  researchers
• Professors Kim Zebehazy and Stacy Kelly
  evaluation of pre- and post-tests
• Field test students and teachers their
  participation was the heart and soul of this
  project

28
The Nemeth Code Tutorial Project was supported by
funding from the U.S. Department of Education
Office of Special Education and Rehabilitative
ServicesGrant No. H327A050093