Closing the Achievement GAP in Mathematics

1
Closing the Achievement GAP in Mathematics

• Dr. Jan Keating
• President, Silicon Valley K-12 Education
  Foundation
• Headmaster, EPGY Online High School _at_ Stanford
  University

2
U.S. Students Are Not Being Prepared for Success
in Higher Levels of Mathematics

• Stanford Universitys EPGY Differentiated
  Kindergarten Through Pre-Algebra Mathematics
  Courses May be the Solution

3
The Problem - Background

• An educated population is vital for economic
  success skills in mathematics are particularly
  important
• Mathematics instruction has shown little
  improvement in the past 30 years
• The US consistently performs poorly on
  international measures of mathematics
• While the first five years makes all the
  difference for the possibility of future success,
  most efforts focus on students in middle school
  or high school

4
No Improvements In Thirty Years
Scores and data from NAEP 2004 Trends In
Academic Progress
Substantial score gaps among groups
5
Internationally the US lags behindTrends in
International Mathematics and Science Study (2004)
USA
6
US Minorities Do Even Worse Program for
International Student Assessment (2000)
7
Why is this problem hard?
8
American Educators Face Difficult Challenges To
Teach Math
Larger class sizes
English is a second language for many students
School funding has decreased
Family issues, decreased involvement of parents
Teacher training and quality
Student mobility
Diverse population of students
Teaching unions can reduce flexibility
K-8 Math teachers tend to not have degrees in
math
Student poverty

Its difficult to catch up once a student falls
behind
9
Teaching to the Mean
Number of Students
Higher
Lower
Performance Level
10
Teaching to the Mean
Teachers Focus
Number of Students
Higher
Lower
Performance Level
11
Teaching to the Mean
Teachers Focus
Lost? Underperform
Number of Students
Bored? Underperform
Higher
Lower
Performance Level
12
A Conceptual Look At Student Population in Two
Dimensions
How Quickly Students Learn
Slow
Fast
Ahead
How Much Progress Have Students Made Compared to
Grade
Behind
13
Some Difficulties for Educators
How Quickly Students Learn
Slow
Fast
Ahead
How to differentiate these groups
How Much Progress Have Students Made Compared to
Grade
Behind
14
Some Difficulties for Educators
How Quickly Students Learn
Slow
Fast
Ahead
Math classes consist of members from all these
groups
How Much Progress Have Students Made Compared to
Grade
Behind
15
How could we fix this problem?

• This problem can be addressed in two ways
• 1. One-on-one tutoring would work, but would be
  prohibitively expensive and not practical
• 2. Use technology to replicate the benefits of
  one-on-one tutoring
• Stanfords approach

16
When New Material is Presented
Mini-lessons provide general instruction at the
conceptual level.
Exercises help the student master the concept
Mastery!
Mini-lessons explain key concepts and
demonstrate how to do exercises Make sure
students understand concepts before practicing
them
17
When Students are Lost Tutorial Hints Mimic
Live Tutors to get students back on the right
track
Mini-lessons provide general instruction at the
conceptual level.
Exercises help the student master the concept
Mastery!
Tutorial hints within exercises provide
individual attention - Immediate feedback and
direction while topics are fresh in students
minds - Seize the teaching moment
18
Efficient Use of Student Time
Mini-lessons provide general instruction at the
conceptual level.
Exercises help the student master the concept
Mastery!
Sophisticated motion and mastery optimizes
student learning the right lesson at the right
time and in the right amount
19
We know this works with gifted students at home,
but could it work with all students, at school,
and during the school day?

• To answer this question, Stanford conducted a
  randomized control trial study.
• The answer is yes, and data is compelling.

20
Effectiveness Study

• 8 Title I elementary schools
• Roughly 1,700 students in grades K - 5
• Matched pair / randomized control trial design
  with randomization at the student level
• Students received 60-100 minutes of EPGY software
  instruction per week
• Control group received a similar amount of
  additional mathematics, either teacher led,
  worksheet, or other programs
• Students began program in November 2006
• Measure was CST results May 2007

21
What Did the Data Show?
22
Background California Standards Test

• In California students are ranked using the
  following scale
• Advanced 5
• Proficient 4
• Basic 3
• Below Basic 2
• Far Below Basic 1
• The goal of No Child Left Behind is to move all
  students into categories 4 and 5.

23
How the Data Were Analyzed

• The first basis for comparison was change in
  performance as measured by the CST
• Did their ranking go up, go down, or remain the
  same
• Such improvements are the NCLB money measure
• EPGY student performance on the 2007 CST was
  compared with that of the control group.
• In evaluating impact, we ordered students by the
  number of correct first attempts on exercises
• This is a measure of serious work done by
  students
• We look successfully at the top 25, 33, 50,
  and top 75 of students using this measure of
  work completed

24
plt10-11
25
plt 10-8
26
plt10-9
27
plt10-7
28
plt10-7
29
It Works!

• With very high certainty (p lt 0.0000001) EPGY was
  significant in improving student NCLB rankings
• Substantially more EPGY students moved up in
  their ranking than did their control counterparts
• It did not matter where they started, all that
  mattered was that they worked carefully and in a
  sustained fashion on EPGY throughout the year
• And the more they worked, the better they did

30
Comparison of Mean Scaled Scores

• Students receive a mean scaled score (0 to 600)
  for their performance on the mathematics CST.
• EPGY students in the top 25 of exercises
  completed significantly outperformed their
  control counterparts.

31
Regression Analysis

• Our regression models show a consistent positive
  relationship between 2007 Math CST scores and
  students EPGY work for all schools, every
  district and every school in the Effectiveness
  Study.
• Independent of prior student achievement, the
  more exercises students did correctly on their
  first try, the higher CST scores they got.

32
Lowell Elementary School

• Sustained use of the program in elementary grades
  can have dramatic impact on student achievement
• Jack OConnell, CA State Superintendent
  recognized Lowell as a model for school
  improvement

33
What did the students think about EPGY K-7
mathematics?
34
Student Survey

• Approximately 1600 Students responded
• Grades 1- 5
• 11 schools
• 9 Title I Schools
• 2 Program Improvement Schools
• Response was overwhelmingly positive

35
Choose the statement that best describes how you
feel about learning math
Students like learning math more than before
using the EPGY Program.
36
Please choose the statement that best describes
your math skills.
Students feel they are doing better in math this
year using the EPGY Program.
37
How do you prefer to learn math?
Students prefer learning mathematics working on
the EPGY program.
38
Do you feel the program helps you enough when you
have questions?
Students feel the EPGY Program provides enough
help when they have questions.
39
How easy or difficult has EPGY been for you?
EPGY provides the optimal pace or motion through
the curriculum.
40
What did the teachers think of EPGY?

• 55 teachers took the teacher survey
• 60 had more than 10 years of experience
• Teaching grades 1-5
• 9 Title I Schools, 2 Higher Socio Economic
  Schools, 2 PI Schools

41
Have you worked online with math programs before?
42
Have you modified your Math instruction as a
result of participating in the EPGY study?
43
What changes do you expect to see in your
students math work during the 2 year study
period?
44
Summing it up

• Stanford Universitys EPGY offers an approach to
  mathematics instruction that has been proven to
  be
• Effective in improving performance of K-5
  students on end of year exams
• Enjoyable for students to use
• They like math more after using it
• They prefer this approach to traditional
  classroom

45
Long Term Sustainability

• Schools will need to provide all students with
  access to computers
• Schools will need to provide additional personnel
  to maintain the software/hardware infrastructure
  required to support this approach to teaching
  mathematics
• Schools will need to provide teachers with
  professional development time
• Schools will need to pay a dollar yearly
  software licensing fee per student

46
Dr. Jan Keatingjkeating_at_stanford.eduhttp//epgy
.stanford.edu/schools/contact.html