robert j. mislevy & min liu university of maryland geneva haertel sri international robert j....

22
Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International The Critical Role of Design Patterns in Large-Scale Assessment DR K-12 grant #0733172, “Application of Evidence-Centered Design to State Large-Scale Science Assessment.” This material is based upon work supported by the National Science Foundation under Grant No. DRL- 0733172. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

Upload: logan-davidson

Post on 18-Dec-2015

222 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

Robert J. Mislevy & Min Liu University of MarylandGeneva Haertel SRI International

Robert J. Mislevy & Min Liu University of MarylandGeneva Haertel SRI International

The Critical Role of Design Patterns in Large-Scale Assessment

The Critical Role of Design Patterns in Large-Scale Assessment

DR K-12 grant #0733172, “Application of Evidence-Centered Design to State Large-Scale Science Assessment.”

This material is based upon work supported by the National Science Foundation under Grant No. DRL- 0733172. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

DR K-12 grant #0733172, “Application of Evidence-Centered Design to State Large-Scale Science Assessment.”

This material is based upon work supported by the National Science Foundation under Grant No. DRL- 0733172. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

Page 2: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

Three design challenges Three design challenges Hard-to-assess standards in the

domain of interest e.g., inquiry science skills

Efficient and valid design and development of complex tasks e.g., scenarios, simulations

Accessibility of tasks for diverse learners varying perceptual and expressive

capabilities valid use of assistive technology,

modifications, alternative assessments

Page 3: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

Design Patterns in Architecture

Design Patterns in Software Engineering

Design Patterns in Literature

Design PatternsDesign Patterns

Page 4: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

Motivation for Assessment Design Patterns

Motivation for Assessment Design Patterns

In-between structure, to connect... Thinking about science learning & inquiry Technical elements of measurement &

delivery Narrative, not technical, contents Some Design Patterns from PADI

Model-Based Reasoning Model Formation; Evaluation; Model Revision; Use

Observational & Experimental Investigations Systems Thinking

Page 5: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

Motivation for Assessment Design Patterns

Motivation for Assessment Design Patterns

They lay out a design space for developers Choices, connections, examples Things to be aware of (e.g., research

on Universal Design for Learning) Can improve both Efficiency +

Validity Attributes reflect assessment

argument structure

Page 6: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

Assessment ArgumentsAssessment Arguments

What complex of

knowledge, skills, or other

attributes should be

assessed?

What behaviors or

performances should reveal

those constructs?

What tasks or situations

should elicit those

behaviors?

Messick, S. (1994). The interplay of evidence and consequences in the validation of performance assessments. Educational Researcher, 23(2), 13-23.

StudentModel

EvidenceModel

TaskModel

Mislevy, R.J., & Haertel, G. (2006). Implications for evidence-centered design for educational assessment. Educational Measurement: Issues and Practice, 25, 6-20.

Page 7: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

The Structure of Assessment Design Patterns

The Structure of Assessment Design Patterns

ATTRIBUTE DESCRIPTION

Rationale How/why this DP addresses evidence about focal KSAs.

Focal Knowledge,

Skills, Abilities

The primary knowledge / skills / abilities (KSAs) targeted by this design pattern.

Additional KSAs

Other knowledge/skills/abilities that may be required by tasks.

Characteristic features

of tasks

Aspects of assessment situations that are needed to evoke evidence about the focal KSAs.

Variable features of

tasks

Aspects of assessment situations that can be varied to shift difficulty or focus.

Potential work

products

What students actually say, do, or make, to produce evidence.

Potential observation

s

Aspects of work products we might identify and evaluate, as evidence about students’ KSAs.

Potential rubrics

Ways of evaluating work products to produce values of observations.

StudentModel

EvidenceModel

TaskModel

Page 8: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

ATTRIBUTE DESCRIPTION

Rationale How/why this DP addresses evidence about focal KSAs.

Focal Knowledge,

Skills, Abilities

The primary knowledge / skills / abilities (KSAs) targeted by this design pattern.

Additional KSAs

Other knowledge/skills/abilities that may be required by tasks.

Characteristic features

of tasks

Aspects of assessment situations that are needed to evoke evidence about the focal KSAs.

Variable features of

tasks

Aspects of assessment situations that can be varied to shift difficulty or focus.

Potential work

products

What students actually say, do, or make, to produce evidence.

Potential observation

s

Aspects of work products we might identify and evaluate, as evidence about students’ KSAs.

Potential rubrics

Ways of evaluating work products to produce values of observations.

The Rationale provides background into the nature of the Focal KSAs, and the kinds of things that people do in what kinds of situations that evidence it. E.g., overview, research links, examples.

StudentModel

EvidenceModel

TaskModel

How Design Patterns Support Thinking about the Assessment

Argument

How Design Patterns Support Thinking about the Assessment

Argument

Page 9: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

StudentModel

EvidenceModel

TaskModel

ATTRIBUTE DESCRIPTION

Rationale How/why this DP addresses evidence about focal KSAs.

Focal Knowledge,

Skills, Abilities

The primary knowledge / skills / abilities (KSAs) targeted by this design pattern.

Additional KSAs

Other knowledge/skills/abilities that may be required by tasks.

Characteristic features

of tasks

Aspects of assessment situations that are needed to evoke evidence about the focal KSAs.

Variable features of

tasks

Aspects of assessment situations that can be varied to shift difficulty or focus.

Potential work

products

What students actually say, do, or make, to produce evidence.

Potential observation

s

Aspects of work products we might identify and evaluate, as evidence about students’ KSAs.

Potential rubrics

Ways of evaluating work products to produce values of observations.

How Design Patterns Support Thinking about the Assessment

Argument

How Design Patterns Support Thinking about the Assessment

Argument

The design pattern is organized around Focal KSAs. They will be

involved in the Student Model, although there may be other KSAs

that are included in the target of inference (e.g., Model Revision—but what models, what context?).

Associated with Characteristic Features of Tasks.

Page 10: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

ATTRIBUTE DESCRIPTION

Rationale How/why this DP addresses evidence about focal KSAs.

Focal Knowledge,

Skills, Abilities

The primary knowledge / skills / abilities (KSAs) targeted by this design pattern.

Additional KSAs

Other knowledge/skills/abilities that may be required by tasks.

Characteristic features

of tasks

Aspects of assessment situations that are needed to evoke evidence about the focal KSAs.

Variable features of

tasks

Aspects of assessment situations that can be varied to shift difficulty or focus.

Potential work

products

What students actually say, do, or make, to produce evidence.

Potential observation

s

Aspects of work products we might identify and evaluate, as evidence about students’ KSAs.

Potential rubrics

Ways of evaluating work products to produce values of observations.

How Design Patterns Support Thinking about the Assessment

Argument

How Design Patterns Support Thinking about the Assessment

Argument

Additional KSAs play multiple roles. You need to think about which ones you really DO want to include as targets of inference

(validity) and which ones you really DON’T (invalidity).

StudentModel

EvidenceModel

TaskModel

Page 11: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

ATTRIBUTE DESCRIPTION

Rationale How/why this DP addresses evidence about focal KSAs.

Focal Knowledge,

Skills, Abilities

The primary knowledge / skills / abilities (KSAs) targeted by this design pattern.

Additional KSAs

Other knowledge/skills/abilities that may be required by tasks.

Characteristic features

of tasks

Aspects of assessment situations that are needed to evoke evidence about the focal KSAs.

Variable features of

tasks

Aspects of assessment situations that can be varied to shift difficulty or focus.

Potential work

products

What students actually say, do, or make, to produce evidence.

Potential observation

s

Aspects of work products we might identify and evaluate, as evidence about students’ KSAs.

Potential rubrics

Ways of evaluating work products to produce values of observations.

How Design Patterns Support Thinking about the Assessment

Argument

How Design Patterns Support Thinking about the Assessment

Argument

The Additional KSAs you DO want to include as targets of inference are part of the claim. E.g., knowing Mendel’s laws as well as being able to formulate a model in an investigation.Connected with Variable

Features of Tasks.

StudentModel

EvidenceModel

TaskModel

Page 12: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

ATTRIBUTE DESCRIPTION

Rationale How/why this DP addresses evidence about focal KSAs.

Focal Knowledge,

Skills, Abilities

The primary knowledge / skills / abilities (KSAs) targeted by this design pattern.

Additional KSAs

Other knowledge/skills/abilities that may be required by tasks.

Characteristic features

of tasks

Aspects of assessment situations that are needed to evoke evidence about the focal KSAs.

Variable features of

tasks

Aspects of assessment situations that can be varied to shift difficulty or focus.

Potential work

products

What students actually say, do, or make, to produce evidence.

Potential observation

s

Aspects of work products we might identify and evaluate, as evidence about students’ KSAs.

Potential rubrics

Ways of evaluating work products to produce values of observations.

How Design Patterns Support Thinking about the Assessment

Argument

How Design Patterns Support Thinking about the Assessment

Argument

The Additional KSAs you DON’T want to include as targets of inference introduce construct-irrelevant reasons for poor performance. (Especially important for assessing special populations – UDL & accommodations.)

Connected with Variable Features of Tasks & Work Products.

StudentModel

EvidenceModel

TaskModel

Page 13: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

ATTRIBUTE DESCRIPTION

Rationale How/why this DP addresses evidence about focal KSAs.

Focal Knowledge,

Skills, Abilities

The primary knowledge / skills / abilities (KSAs) targeted by this design pattern.

Additional KSAs

Other knowledge/skills/abilities that may be required by tasks.

Characteristic features

of tasks

Aspects of assessment situations that are needed to evoke evidence about the focal KSAs.

Variable features of

tasks

Aspects of assessment situations that can be varied to shift difficulty or focus.

Potential work

products

What students actually say, do, or make, to produce evidence.

Potential observation

s

Aspects of work products we might identify and evaluate, as evidence about students’ KSAs.

Potential rubrics

Ways of evaluating work products to produce values of observations.

How Design Patterns Support Thinking about the Assessment

Argument

How Design Patterns Support Thinking about the Assessment

Argument

The Characteristic Features of Tasks help you think about critical features of the tasks

situation you need to get evidence about the Focal KSAs.

StudentModel

EvidenceModel

TaskModel

Page 14: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

ATTRIBUTE DESCRIPTION

Rationale How/why this DP addresses evidence about focal KSAs.

Focal Knowledge,

Skills, Abilities

The primary knowledge / skills / abilities (KSAs) targeted by this design pattern.

Additional KSAs

Other knowledge/skills/abilities that may be required by tasks.

Characteristic features

of tasks

Aspects of assessment situations that are needed to evoke evidence about the focal KSAs.

Variable features of

tasks

Aspects of assessment situations that can be varied to shift difficulty or focus.

Potential work

products

What students actually say, do, or make, to produce evidence.

Potential observation

s

Aspects of work products we might identify and evaluate, as evidence about students’ KSAs.

Potential rubrics

Ways of evaluating work products to produce values of observations.

How Design Patterns Support Thinking about the Assessment

Argument

How Design Patterns Support Thinking about the Assessment

Argument

StudentModel

EvidenceModel

TaskModel

Variable Features of Tasks also help you think about data concerning the situation – but now to influence difficulty …

or to bring in or reduce demand for Additional KSAs to avoid alternative explanations.

Page 15: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

ATTRIBUTE DESCRIPTION

Rationale How/why this DP addresses evidence about focal KSAs.

Focal Knowledge,

Skills, Abilities

The primary knowledge / skills / abilities (KSAs) targeted by this design pattern.

Additional KSAs

Other knowledge/skills/abilities that may be required by tasks.

Characteristic features

of tasks

Aspects of assessment situations that are needed to evoke evidence about the focal KSAs.

Variable features of

tasks

Aspects of assessment situations that can be varied to shift difficulty or focus.

Potential work

products

What students actually say, do, or make, to produce evidence.

Potential observation

s

Aspects of work products we might identify and evaluate, as evidence about students’ KSAs.

Potential rubrics

Ways of evaluating work products to produce values of observations.

How Design Patterns Support Thinking about the Assessment

Argument

How Design Patterns Support Thinking about the Assessment

Argument

Some Variable Features of Tasks help you match features of tasks and background / knowledge / characteristics of students: Interests, familiarity, previous instruction.

StudentModel

EvidenceModel

TaskModel

Page 16: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

ATTRIBUTE DESCRIPTION

Rationale How/why this DP addresses evidence about focal KSAs.

Focal Knowledge,

Skills, Abilities

The primary knowledge / skills / abilities (KSAs) targeted by this design pattern.

Additional KSAs

Other knowledge/skills/abilities that may be required by tasks.

Characteristic features

of tasks

Aspects of assessment situations that are needed to evoke evidence about the focal KSAs.

Variable features of

tasks

Aspects of assessment situations that can be varied to shift difficulty or focus.

Potential work

products

What students actually say, do, or make, to produce evidence.

Potential observation

s

Aspects of work products we might identify and evaluate, as evidence about students’ KSAs.

Potential rubrics

Ways of evaluating work products to produce values of observations.

How Design Patterns Support Thinking about the Assessment

Argument

How Design Patterns Support Thinking about the Assessment

Argument

Potential Work Products help you think about what you want to capture from a performance –product, process, constructed model, written explanation, etc.

Can also call attention to demand for Additional KSAs, & avoid alternative explanations for poor performance

StudentModel

EvidenceModel

TaskModel

Page 17: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

ATTRIBUTE DESCRIPTION

Rationale How/why this DP addresses evidence about focal KSAs.

Focal Knowledge,

Skills, Abilities

The primary knowledge / skills / abilities (KSAs) targeted by this design pattern.

Additional KSAs

Other knowledge/skills/abilities that may be required by tasks.

Characteristic features

of tasks

Aspects of assessment situations that are needed to evoke evidence about the focal KSAs.

Variable features of

tasks

Aspects of assessment situations that can be varied to shift difficulty or focus.

Potential work

products

What students actually say, do, or make, to produce evidence.

Potential observation

s

Aspects of work products we might identify and evaluate, as evidence about students’ KSAs.

Potential rubrics

Ways of evaluating work products to produce values of observations.

How Design Patterns Support Thinking about the Assessment

Argument

How Design Patterns Support Thinking about the Assessment

Argument

Potential Observations are possibilities for the qualities of Work Products – i.e., the data concerning the performance.

StudentModel

EvidenceModel

TaskModel

Page 18: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

ATTRIBUTE DESCRIPTION

Rationale How/why this DP addresses evidence about focal KSAs.

Focal Knowledge,

Skills, Abilities

The primary knowledge / skills / abilities (KSAs) targeted by this design pattern.

Additional KSAs

Other knowledge/skills/abilities that may be required by tasks.

Characteristic features

of tasks

Aspects of assessment situations that are needed to evoke evidence about the focal KSAs.

Variable features of

tasks

Aspects of assessment situations that can be varied to shift difficulty or focus.

Potential work

products

What students actually say, do, or make, to produce evidence.

Potential observation

s

Aspects of work products we might identify and evaluate, as evidence about students’ KSAs.

Potential rubrics

Ways of evaluating work products to produce values of observations.

How Design Patterns Support Thinking about the Assessment

Argument

How Design Patterns Support Thinking about the Assessment

Argument

And Potential Rubrics are algorithms/rubrics/rules for evaluating Work Products to interpret evidence from the

student’s performance.

StudentModel

EvidenceModel

TaskModel

Page 19: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

Current Catalog of Design PatternsCurrent Catalog of Design Patterns

ECD/PADI related projects have produced over

100 Design Patterns

Domains include science inquiry, science content,

mathematics, economics, model-based reasoning

Span grades 3-16+

Organized around themes, models, and

processes, not surface features or formats of

tasks Support the design of scenario-based, multiple choice,

and performance tasks

Page 20: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

Current Catalog of Design PatternsCurrent Catalog of Design Patterns

Subject Areas

Education

Standards

Unifying Themes/ Inquiry

Big Ideas within

Disciplines

Learning Progression

s

Language Proficiency

Total

Science 17 57 4 2 0 80

Mathematics 30 2 3 0 0 35

Economics 0 0 3 0 0 3

Language Arts

30 0 0 0 1 31

Management/ Business

0 7 3 0 0 10

Second Language Acquisition

0 0 0 0 3 3

Grand Total 77 66 13 2 4 162

Page 21: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

For more information…For more information…

PADI: Principled Assessment Design for Inquiry

http://padi.sri.com Links to NSF & IES follow-on projects Lots of Tech Reports, interactive online

examples Bob Mislevy home page

http://www.education.umd.edu/EDMS/mislevy/ Links to papers on ECD Cisco applications

Page 22: Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI International Robert J. Mislevy & Min Liu University of Maryland Geneva Haertel SRI

Now for the Good Stuff …

Now for the Good Stuff …

Examples of design patterns with content Different projects

Different grain sizes

Different users

How they are being used to tackle pervasive challenges of large-scale assessment.

How they evolved to suit needs of users Same essential structure, but

Representations, language, emphases, and affordances tuned to users and needs