Issues in Vertical Scaling and ValueAdded Models in Oregon

1
Issues in Vertical Scaling and Value-Added Models
in Oregon

• Its always something.
• CCSSO Growth Conference
• November 15-16, 2004

2
Assessment System In Place

• Tests operational in 1991
• Vertical scale (grades 3-11)
• Rasch item calibration
• Aligned to state content standards
• Performance standards set in 1995
• Unique statewide student ID since 2002
• Political commitment to improvement

3
Current Issues

• Its a New World
• Performance differences at grade levels
• Will our vertical equating hold up?
• How do we include performance assessments?
• Issues not well understood

4
Its a New World

• Key decisions made long ago
• Stakes are higher ()
• Resources are reduced
• Stakes are higher (Media)
• Purpose for assessment has expanded
• From program evaluation
• To student certification
• To school accountability

5
Performance Differences Over Time

• Oregon schools have produced greater growth in
  meeting at the elementary level.
• There is pressure to adjust student performance
  standards to equalize

6
(No Transcript)
7
(No Transcript)
8
Issue in Vertical Scaling and Standard Setting
9
Vertical Equating

• New grades
• Adaptive testing within each grade level
• Multiple attempts within grade level
• Subgroup individual growth may vary

10
Performance Assessments

• Not scaled
• Scaling approaches hard to explain to public
• Should we use a common scale with multiple choice
  tests?

11
Issues with Issues

• Many people like the general concept of tracking
  individual student growth
• Instructional support
• School accountability
• But,
• The math is complex
• How do we include all students? (Mobility)
• Growth targets to close achievement gaps

12
Conclusion

• Issues expand to exceed the time and resources
  available
• Therefore we must be on the right track