The Center for DataDriven Reform in Education

1
The Center for Data-Driven Reform in Education
2

• Well now, lets see who made adequate yearly
  progress this year

3
C8

• I want that mountain gone by 2014. Heres your
  shovel.

4
Before AfterRenita, Reinvented
                                                  
        
5
The Gremlin
6
On Beyond Diagnosis

• Assumptions
• Districts and schools know strengths and
  weaknesses
• State assessments
• Benchmark assessments
• Subscales
• Subgroups

7

• Our new equipment lets us find out exactly
  which of 127 viruses is causing your cold. Of
  course, we cant cure any of them.

8
How Big is Your Problem?

• Problem
• Fractions

• Potential Solution
• Workshop on fractions

9
How Big is Your Problem?

• Problem
• Readingsome subgroups

• Potential Solution
• After school, summer school, remedial program

10
How Big is Your Problem?

• Problem
• Math-across the board

• Potential Solution
• New math program

11
How Big is Your Problem?

• Problem
• Reading and Math-across the board

• Potential Solution
• Comprehensive school reform

12
How Big is Your Problem?

• Problem
• Completely Fubar

• Potential Solution
• Reconstitution

13
Theres nothing normal about his distribution!
14
Advantages of Outside Reform Organizations

• Track record
• Research base (hopefully)
• Accountability for results

15
Adapting Outside Reforms

• Adapt to state standards, assessments
• Adapt to local needs and culture

16
Finding and Evaluating Evidence-Based Solutions

• Locate Impartial Reviews
• What Works Clearinghouse
• Show Me the Evidence
• Reviews in RER
• Other Reviews

17
The Gremlins Question

• Hold on a minutehas this program been
• Evaluated in comparison to a control group
• In studies of at least one semester
• In multiple schools
• Found to improve achievement more than the
  control group, and
• Published in a peer-reviewed journal?

18
Comparisons to Control Groups

• Random assignment
• Matched, with controls for pretests
• Study reports data for all schools involved

19
How long?

• At least 1 semester
• Longer is bettermultiple years if possible

20
How many schools? How many studies?

• At least 40, across studies
• At least 3 studies

21
Gains Compared to Controls

• Statistically significant
• Educationally significant
• Design is more important than effect size
• Control for pretests

22
Publication

• Top peer-reviewed journals
• Other peer-reviewed journals
• Technical reports that meet high standards

23
(No Transcript)