Measuring What Matters

1
Measuring What Matters

• Elena Silva, Education Sector
• esilva_at_educationsector.org
• 202-552-2844
• Rennie Center
• April 13, 2009

2
Learning and Policy Paradox

• Increasing demand for broader learning
• outcomes is set against an educational
• accountability system that is focused on a
• narrow band of basic knowledge and skills.

3
2 Big Challenges

• 1. Measuring whats most important.
• 2. Measuring for accountability and improvement.

4
What Should We Be Measuring?

• Public consensus on student outcomes content
  and skills.
• Decades of research on how people learn content
  and skills.
• Still a debate on 21st century skills What are
  they? Do they matter? Are they outside the bounds
  of measurement?

5
21st Century Skills

• There are dozens of frameworks for 21st century
  skills

• Students must know how to analyze and evaluate
  ideas,
• communicate and CREATE new ideas. These higher
• order skills are not new, they are newly
  important.

6
So Can We Measure These Skills?

• Yes. The challenge is not as technical
• as it may seem.
• We can isolate and measure most
  constructsCollege Board, ETS.
• We can measure higher order skills and content at
  the same time.

7
Assessment Models to Watch

• Collegiate Learning Assessment (CLA) and College
  Work and Readiness Assessment (CWRA)
• River City (Dede/Harvard multi-user virtual
  environments, middle school science hypotheses)
• CRESST POWERSOURCE (pilot in 80 middle schools)

8
What To Do (and not to do)

• Generate common understanding about skills and
  content, and frame standards, curriculum and
  assessment, as well as teacher training, around
  the nexus of skills and content. Do not tack on
  21st century skills to existing frameworks.
• Invest in the research and development of
  assessment tools and an assessment system that
  has a broader purpose (accountability and
  improvement). Do not add another layer of tests.
• Look to emerging technologies as tools for new
  forms of assessment. But do not assume new is
  better (comparing known problems of old with
  assumed benefits of new).

9
(No Transcript)