Modeling Waiting List Reduction Strategies Using …...Discrete Choice Conjoint Experiments Goals of the Talk Brief Introduction to Preference Modelling Discrete Choice Experiments

Charles E. Cunningham, Ph.D.Professor

Department of Psychiatry & Behavioural Neurosciences

Jack Laidlaw Chair in Patient-Centred Health Care

Michael G. DeGroote School of Medicine

Faculty of Health Sciences

McMaster University

Modeling Implementation and Knowledge Translation Strategies Using

Discrete Choice Conjoint Experiments

Goals of the Talk

Brief Introduction to Preference Modelling

Discrete Choice Experiments

Highlights from Studies of KTE Preferences

Addiction Professionals

Children’s Mental Health Professionals

Educators

Implications?

Why Study Implementation?

“Multiple studies across differing treatments found that training (workshop, manual, and supervision) was not enough to produce proficient change in therapist adherence, competence, and skill. In turn, Therapists were not necessarily able to engender client change.”

Source: Biedas and Kendall (2010). Training Therapists in Evidence-Based Practice: A Critical Review of Studies From a Systems-Contextual Perspective. Clinical Psychology Science and Practice 17, 1-30.

For treatments to be effective and sustained in practice settings, treatment developers should consider design features that increase the appeal to the therapists who are ultimately responsible for using them.

Source: Journal of Consulting and Clinical Psychology, 2013, 81, 6, 999-1009,

Modeling the Implementation of School-based Mental Health Strategies

CIHR Team in Access to Children’s

Mental Health Services

Principal Investigator Melanie Barwick (University of Toronto)

Co-Investigators (Selected) Kathy Short (Hamilton-Wentworth District School Board

Research Staff Heather Rimas Stephanie Mielko Jenna Ratcliffe Cathy Campbell (Team Secretary) Yvonne Chen (McMaster HRM Program)

Funding Canadian Institutes of Health Research Jack Laidlaw Chair in Patient-Centred Health Care

Discrete Choice Conjoint Survey

Sample Size = 1010 Educataors

Return Rate = 82%

17 4-Level KT Attributes

Partial Profile Design

17 Choice Tasks Per Informant

999 Near Orthogonal Versions of the Survey

Source: Cunningham, Barwick et al. School Mental Health (2014)

Assume you are considering changing your teaching practice to improve the behavioral and emotional outcomes of students:

Click below the practice change strategy you would prefer:

Next

Strategy 1 Strategy 2 Strategy 3

Learning requires 2 days

My colleagues endorse this 67%

Other schools say this strategy works

Learning requires 1 day


This strategy is promising but

unproven

Learning requires 3 days


Research says this strategy works

Note: 999 versions – 1 randomly assigned to each participant

Partial Profile Designs

Reduce Effect of Dominant Attributes

Reduce Response Error

Reduce Total Design Error

Produce Similar Utility Coefficients

Improve Predictive Validity of Simulations

(Reduce MAE)

Source: Chrzan, K. (2010). Using partial profile choice experiments to handle large numbers of attributes. International Journal of Market Research, 52, 827-840.

Data Analysis

Hierarchical Bayes or Conditional Logit

Importance Scores (Relative Influence on Choices)

Zero Centered Utility Coefficients (Strength of Preference)

Latent Class Analysis

Multi-level Latent Class Analysis

Randomized First Choice Simulations

Source: Hauber, B., Gonzalez, J., Groothuis-Oudshoorn, C., Prior, T., Marshall, D.,Cunningham, C., . . .

Bridges, J. (2016). Statistical methods for the analysis of discrete-choice experiments: A report of the ISPOR

conjoint analysis good research practices task force. Value in Health, 19(4), 300-315.

Why Use Discrete Choice Experiments?

Approximate Complexity of Real World Choices

Each Level Experimentally Manipulated

Each Attribute in the Context of Others

Complexity Activates Decision Making Heuristics

Users Inform Implementation Planning Tradeoffs

Reduce Influence of Social Desirability Biases

Segments with Different Preferences? Latent Class Segmentation Analysis

Demand Sensitive (23%)

Change Ready (77%)

Source: Cunningham, Barwick et al., School Mental Health (2014)

Covariates in Latent Class Analysis

Can be Linked to Latent Class Membership

Can be Included in the Formation of Latent Classes

Improve Model Fit

Enhance the Interpretation of Segment Membership

But - Increase Number of Parameters Estimated

What Attitudes Distinguish Change Ready vs Demand Sensitive Segments?

Change Ready Educators:

Anticipate More Benefits to Practice Change

Find the Social Context to be More Influential

Report Higher Change Self Efficacy

More Intent on Participating

Relative Importance of KT Design Attributes

0 1 2 3 4 5 6 7 8 9 10 11 12

Presentor Qualitites

Colleague Support

Compatibility

Administrative Support

Union Endorsement

Coaching

Supporting Evidence

Provincial Curriculum Links

Workshop Size

Skills vs Knowledge

Follow-up Support

Observability Trialability

Training Time Demand

Selection Process

Universal vs Targeted

Internet Options

Relative Importance

Imple

menta

tion A

ttribute

Attributes of the Implementation

Process

Sensitivity to Variation in an Attribute’s Levels

Importance= Variation In Each Attribute’s Utility/Total Utility Variation

Attributes of the Social Context Influencing the Decision to Adopt Mental Health Practice Changes

Relative Importance of KT Attributes: Colleague Support for Change

0 1 2 3 4 5 6 7 8 9 10 11 12


Colleague Support

Compatibility


Union Endorsement

Coaching

Supporting Evidence


Workshop Size

Skills vs Knowledge

Follow-up Support



Selection Process


Internet Options

Relative Importance

Imp

lem

enta

tio

n A

ttrib

ute

Demand Sensitive Change Ready

Support of Colleagues for Change Exerts an Important

Influence


Colleague Support

-2

-1.5

-1

-0.5

0

0.5

1

1.5

0% Support

this Option

33% Support

this Option

67% Support

this Option

100% Support

this Option

Utility

Valu

e

Change Ready Demand Sensitive

*

Both Segments Prefer Programs

Supported by 100% of Colleagues


Relative Importance of KT Design Attributes: Administrative Support

0 1 2 3 4 5 6 7 8 9 10 11 12


Colleague Support

Compatibility


Union Endorsement

Coaching

Supporting Evidence


Workshop Size

Skills vs Knowledge

Follow-up Support



Selection Process


Internet Options

Relative Importance

Imple

menta

tion A

ttribute



Important

Source: Cunningham, Barwick et al., School Mental Health (2014)

Support by Administrators

-100

-80

-60

-40

-20

0

20

40

60

80

Does Not

Support

33% 67% 100%

Utility

Valu

e

KT Choices of Both Segments Sensitive to Administrative Support


Relative Importance of KT Design Attributes: Union Endorsement

0 1 2 3 4 5 6 7 8 9 10 11 12


Colleague Support

Compatibility


Union Endorsement

Coaching

Supporting Evidence


Workshop Size

Skills vs Knowledge

Follow-up Support



Selection Process


Internet Options

Relative Importance

Imple

menta

tion A

ttribute


Union Endorsement Exerted a Strong Influence on Choices


Union Endorsement

-2

-1.5

-1

-0.5

0

0.5

1

1.5

Does Not

Support

33% Support 67% Support 100% Support

Utility

Valu

e


*

Decisions of Both Segments Sensitive to Union Support


Attributes of Mental Health Practices that Influence Implementation Decisions

Relative Importance of KT Attributes: Compatibility with Practice

0 1 2 3 4 5 6 7 8 9 10 11 12


Colleague Support

Compatibility


Union Endorsement

Coaching

Supporting Evidence


Workshop Size

Skills vs Knowledge

Follow-up Support



Selection Process


Internet Options

Relative Importance

Imp

lem

enta

tio

n A

ttrib

ute


Compatibility with Teaching Practice Highly

Important


Compatibility with Teaching Practice

-2

-1.5

-1

-0.5

0

0.5

1

1.5

0% 33% 67% 100%

Utility

Valu

e


Educators More Likely to Choose Mental Health Practices

Consistent with Teaching Practice

Source: Cunningham, Barwick, et al., School Mental Health (2014)

Relative Importance of KT Design Attributes: Supporting Evidence

0 1 2 3 4 5 6 7 8 9 10 11 12


Colleague Support

Compatibility


Union Endorsement

Coaching

Supporting Evidence


Workshop Size

Skills vs Knowledge

Follow-up Support



Selection Process


Internet Options

Relative Importance

Imple

menta

tion A

ttribute


Supporting Evidence

Moderately Strong Importance


Supporting Evidence

-1.5

-1

-0.5

0

0.5

1

Promising butUntested

Other Schools Saythis Works

Research Says thisWorks

Research & OtherSchools Say this

Works

Utility

Valu

e


Both Segments Prefer Options Based on Research + Experience of Other Schools

Demand Sensitive Educators Value Experience of Other Schools vs Research


Relative Importance of KT Attributes:Provincial Curriculum Links

0 1 2 3 4 5 6 7 8 9 10 11 12


Colleague Support

Compatibility


Union Endorsement

Coaching

Supporting Evidence


Workshop Size

Skills vs Knowledge

Follow-up Support



Selection Process


Internet Options

Relative Importance

Imp

lem

en

tatio

n A

ttrib

ute


Provincial Curriculum Links Moderately Important to

Both Segments


Links to Provincial Curriculum

-1.5

-1

-0.5

0

0.5

1

0% 33% 67% 100%

Utility



Attributes of the Implementation Process that Influence Decisions

Relative Importance of KT Design Attributes: Qualities of Presenters

0 1 2 3 4 5 6 7 8 9 10 11 12


Colleague Support

Compatibility


Union Endorsement

Coaching

Supporting Evidence


Workshop Size

Skills vs Knowledge

Follow-up Support



Selection Process


Internet Options

Relative Importance

Imple

menta

tion A

ttribute


Qualities of Presenters Very

Important Influence


Qualities of the Presentor

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

Not EngagingNor Expert

Engaging NonExpert

Non EngagingExpert

Engaging Expert

Utility

Valu

e


Demand Sensitive Educators Value Engaging More than Expertise

Educators Prefer an Engaging Expert


Relative Importance of KT Attributes:Coaching to Support Implementation

0 1 2 3 4 5 6 7 8 9 10 11 12


Colleague Support

Compatibility


Union Endorsement

Coaching

Supporting Evidence


Workshop Size

Skills vs Knowledge

Follow-up Support



Selection Process


Internet Options

Relative Importance

Imp

lem

en

tatio

n A

ttrib

ute


Coaching to Support Implementation More

Important to the Change Ready Segment


Coaching Support to Improve Skills

-2

-1.5

-1

-0.5

0

0.5

1

No Coaching If I Want If I Need It All Participants Get

Coaching

Utility

Valu

e



Relative Importance of KT Attributes:Workshop Size

0 1 2 3 4 5 6 7 8 9 10 11 12


Colleague Support

Compatibility


Union Endorsement

Coaching

Supporting Evidence


Workshop Size

Skills vs Knowledge

Follow-up Support



Selection Process


Internet Options

Relative Importance

Imp

lem

en

tati

on

Att

rib

ute


Workshop Size Exerts a Moderately Important Influence on Choices


Workshop Group Size

-1

-0.5

0

0.5

1

I Learn This Alone Group of 10 Group of 50 Group of 100

Utility

Coeffic

ient


Both segments prefer information delivered in small groups of 10


Relative Importance of KT Attributes: Training Time Demands

0 1 2 3 4 5 6 7 8 9 10 11 12


Colleague Support

Compatibility


Union Endorsement

Coaching

Supporting Evidence


Workshop Size

Skills vs Knowledge

Follow-up Support



Selection Process


Internet Options

Relative Importance

Imp

lem

en

tatio

n A

ttrib

ute


Training Time Demands Exert More Influence on Demand Sensitive Educators


Training Time Demands

-1

-0.5

0

0.5

1

1 Day 2 Days 3 Days 4 Days

Utility

Valu

e

Number of Training Sessions


Both Segments prefer 1 day workshops


Relative Importance of KT Attributes:Follow-up Support

0 1 2 3 4 5 6 7 8 9 10 11 12


Colleague Support

Compatibility


Union Endorsement

Coaching

Supporting Evidence


Workshop Size

Skills vs Knowledge

Follow-up Support



Selection Process


Internet Options

Relative Importance

Imple

menta

tion A

ttri

bute


Follow-up Support More Important to Change

Ready Educators


Follow-up Support

-1.5

-1

-0.5

0

0.5

None 1 2 3

Utility

Valu

e

Number of One Hour Follow-up Sessions

Change Ready Demand SensitiveDemand Sensitive Educators Prefer No Follow-up Support


Relative Importance of KT Design Attributes: Selection Process

0 1 2 3 4 5 6 7 8 9 10 11 12


Colleague Support

Compatibility


Union Endorsement

Coaching

Supporting Evidence


Workshop Size

Skills vs Knowledge

Follow-up Support



Selection Process


Internet Options

Relative Importance

Imple

menta

tion A

ttri

bute


Selection Process of Relatively Low Importance


KT Selection Process

-0.5

-0.25

0

0.25

0.5

Ministry Selects Board Selects Individual Schools

Select

Individual Teachers

Select

Utility

Valu

e



Relative Importance of KT Design Attributes: Internet Options

0 1 2 3 4 5 6 7 8 9 10 11 12


Colleague Support

Compatibility


Union Endorsement

Coaching

Supporting Evidence


Workshop Size

Skills vs Knowledge

Follow-up Support



Selection Process


Internet Options

Relative Importance

Imple

menta

tion A

ttri

bute


Internet Options Exert Little Influence on Choices


Internet Training Options

-1

-0.5

0

0.5

1

No Internet Options Moderated Internet

Discussion

Internet Learning +

Moderated Discussion

Internet Learning

Activities

Utility

Valu

e



Limited Interest in Internet Options

Randomized First Choice Simulation

Standard Dissemination 1 Day Large Group (N=50) No Internet 67% Focus on Knowledge No Coaching or Follow-up

Enhanced Dissemination 3 Day Small Group (N=10) No Internet 67% Focus on Skills Coaching and 3 One Hour Follow-ups

Internet Dissemination 3 Day Individual Learning Internet Learning and Moderated Discussion 67% Focus on Skills No Coaching but 3 Internet Follow-up Sessions


0

20

40

60

80

100

Standard

Dissemination

Enhanced

Dissemination

Internet

Dissemination

1.4

98.4

0.3

53.3

44.8

1.9

Perc

ent

Change Ready

Demand Sensitive



Is this Initiative Supported by My Principal, Colleagues, and Union?

Is this Compatible with My Practice & Consistent with Provincial Curricula?

What is the Evidence that This Works?

Does this Work for Other Schools?

Is It Supported by Research?

Is the Presenter an Engaging Expert?

Is this Consistent with My Learning Preferences1 Day, Small Group (n=10), Skill Focused, Supported by Coaching,1 to 2 Follow-ups

Multi-Stage Implementation Decisions

Modeling Evidence-based Practice Dissemination in Addiction Agencies Serving Women

SELECTED PROJECT PARTNERS

Members of Research Team Joanna Henderson Alison Niccols Maureen Dobbins Wendy Sword Karen Milligan Ellen Lipman Lehana Thabane Louis Schmidt

Research Team Heather Rimas (Research Coordinator) Stephanie Mielko (Research Assistant) Yvonne Chen (HRM Graduate Student) Ainsley Smith

Research Support Canadian Institutes of Health Research Jack Laidlaw Chair in Patient-Centred Health Care


Sample Size = 1379

Return Rate = 60%



999 Versions of the Survey

Source: Cunningham, Henderson, Niccols, et al., Addiction (2012)

Latent Class Segmentation Analysis

Process Sensitive (30%)

Outcome Sensitive (52%)

Demand Sensitive (18%)


Relative Importance of Knowledge Translation Attributes

0 2 4 6 8 10 12

Benef it t o Client s

Compat ibilit y

Evidence

Ease of Implement at ion

Administ rat ive Support

Co-Worker Support

Cont ent Select ion

Present or Credent ials

Push Pull

Knowledge vs Skills

Endorsement

Modalit y

Act ive Learning

Group Size

Implement at ion Support

Time Cost

KTE A

ttri

bute

s

Importance Score

KT Choices Most Sensitive to Percentage of Clients Benefiting



0 2 4 6 8 10 12


Compat ibilit y

Evidence



Co-Worker Support

Cont ent Select ion


Push Pull

Knowledge vs Skills

Endorsement

Modalit y

Act ive Learning

Group Size


Time Cost

KTE A

ttri

bute

s

Importance Score

Compatibility of KT Content Important



0 2 4 6 8 10 12


Compat ibilit y

Evidence



Co-Worker Support

Cont ent Select ion


Push Pull

Knowledge vs Skills

Endorsement

Modalit y

Act ive Learning

Group Size


Time Cost

KTE A

ttri

bute

s

Importance Score

Participants Sensitive to the Support of Colleagues &

Administrators


Relative Importance of Knowledge Translation & Exchange Attributes to Professionals

0 2 4 6 8 10 12


Compat ibilit y

Evidence



Co-Worker Support

Cont ent Select ion


Push Pull

Knowledge vs Skills

Endorsement

Modalit y

Act ive Learning

Group Size


Time Cost

KTE A

ttri

bute

s

Importance Score

Attributes of the KT Process Exerted Less Influence on

Choices


Relative Importance of Knowledge Translation & Exchange Attributes to Professionals

0 2 4 6 8 10 12


Compat ibilit y

Evidence



Co-Worker Support

Cont ent Select ion


Push Pull

Knowledge vs Skills

Endorsement

Modalit y

Act ive Learning

Group Size


Time Cost

KTE A

ttri

bute

s

Importance Score

Content Selection Process of Moderate Importance


-60

-40

-20

0

20

40

60

Government Administrators Me Co-Workers & I

Content Selection Process

Utility

Valu

es

Outcome Process Demand

Content Selection Process

Participants Prefer Local Collaborative Content

Selection


Relative Importance of Knowledge Translation Attributes: KT Presenter

0 2 4 6 8 10 12


Compat ibilit y

Evidence



Co-Worker Support

Cont ent Select ion


Push Pull

Knowledge vs Skills

Endorsement

Modalit y

Act ive Learning

Group Size


Time Cost

KTE A

ttri

bute

s

Importance Score

Credentials of KT Presenter Important


-60

-40

-20

0

20

40

60

Administrator Former Client Researcher Clinician

Presenter

Utility

Valu

es

Outcome Process Demand

KT Presenter

Participants Prefer Presenters Who are Clinicians vs

Researchers


Simulating Response to KT Options

Standard KT

1-Day Time Committment

Large Group Format

Focus is 67% Knowledge 33% Skill

Easy to Apply

No Review Questions or Practice

No Implementation Follow-ups


Enhanced KT

4-Day Time Commitment

Small Group Format

Focus is 67% Knowledge 33% Skill

Difficult to Apply

Review Questions and Practice

3 Implementation Follow-ups


Simulating Practice Change Options

66.3

44.3

81

33.7

55.7

19.1

0

20

40

60

80

100

% S

hare

of Pre

fere

nce

Standard Enhanced

Outcome Process DemandShare of Preference is the Percentage of

Participants Predicted to Prefer the Standard or Enhanced KT Strategies


62.5

29.1

7.5

24

37.5

70.9

92.5

76

0

20

40

60

80

100

Baseline Enhanced + Evidence Enhanced +

Evidence+Colleague +


Government Selection &

Endorsement

Standard

Enhanced

Percent Predicted to Prefer Enhanced KT


Modeling Implementation of EBPs in Children’s Mental Health Agencies

CIHR Team in Access to Children’s

Mental Health Services

Principal Investigator Melanie Barwick (University of Toronto)

Selected Research Staff Heather Rimas Stephanie Mielko Yvonne Chen (McMaster HRM Program) Raluca Barac

Funding Canadian Institutes of Health Research Jack Laidlaw Chair in Patient-Centred Health Care


Sample Size = 563 Mental Health Service Providers

31 Agencies



Partial Profile Design

999 Versions of the Survey

Source: Cunningham, Barwick et al., Administration and Policy in Mental Health (2018).

Segments with Different Preferences? Latent Class Segmentation Analysis

Accelerated Adopters (12%)

Gradual Adopters (88%)


Attributes of the Social Context Influencing the Decision to Adopt Mental Health Practice Changes

Relative Importance of KT Design Attributes: Supervisor Support

0 1 2 3 4 5 6 7 8 9 10 11 12 13

Supervisor Support for EBP

% Clients Benefiting

Skill vs Knowledge

Colleagues Support EBP

Initial Training on Internet

Engaging Expert Trainer

Active vs Passive Learning

Follow-up Trainin

Effectivenes in Other Agencies

Group Size

Initial Training Time


Selection Control

% Change in Practice

Internet Options

Relative Importance

Imple

menta

tion A

ttri

bute

Gradual Adopters Accelerated Adopers

Accelerated Adopters Most Sensitive to Variation in


Source: Cunningham, Barwick et al. Administration and Policy in Mental Health (2018)

Relative Importance of KT Design Attributes: Trainer Qualitites

0 1 2 3 4 5 6 7 8 9 10 11 12 13



Skill vs Knowledge





Follow-up Trainin


Group Size



Selection Control


Internet Options

Relative Importance

Imple

menta

tion A

ttri

bute

Gradual Adopters Accelerated Adopters

Engaging Expert Trainer Most Important to Gradual

Adopters


-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

Not Engaging Not

Expert

Engaging Not

Expert

Expert Not

Engaging

Engaging Expert

Utility

Accelerated Adopters

Gradual Adopters


Trainer Qualities

Attributes of the Implementation Process that Influence Decisions

Relative Importance of KT Design Attributes: Initial Training Time

0 1 2 3 4 5 6 7 8 9 10 11 12 13



Skill vs Knowledge





Follow-up Trainin


Group Size



Selection Control


Relative Importance

Imple

menta

tion A

ttri

bute

Gradual Adopters Accelerted Adopters

Initial Training Time Demand of Relatively Low Importance


-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

1 Day 2 Days 3 Days 4 Days

Utility



Gradual Adopters



Relative Importance of KT Design Attributes: Initial Internet Training

0 1 2 3 4 5 6 7 8 9 10 11 12 13



Skill vs Knowledge





Follow-up Trainin


Group Size



Selection Control


Relative Importance

Imple

menta

tion A

ttri

bute


Initial Internet Training Appealed to Accelerated

Adopters


-1

-0.5

0

0.5

1

No Internet

Learning

33% Initial

Training On Line

66% Initial

Training Online

100% Initial

Training Online

Utility


Gradual Adopters


Initial Internet Training

Relative Importance of KT Design Attributes: Initial Training Time

0 1 2 3 4 5 6 7 8 9 10 11 12 13



Skill vs Knowledge





Follow-up Trainin


Group Size



Selection Control


Internet Options

Relative Importance

Imple

menta

tion A

ttri

bute


Initial Training Time of Relatively Low Importance


Mental Health Professionals

Both Segments Prefer

100% Support of Supervisors and Colleagues for EBP Changes

EBP Changes Benefiting All Patients

Proven in Research and Other Agencies

Trainers Who are Engaging Experts

Selected by Individual Programs within Agencies (Rather than Ministries)

Focus on 33% Knowledge 67% Skills

Observational Learning, Practice, and Feedback

Mental Health Professionals

Accelerated Adopters 12%

Greater Intent to Participate

100% of Initial Training on Line

Devote More Time (4 Days) to Initial Training

Would Make More Changes to Their Practice

Introduce only Minor Modifications to EBPs

Gradual Adoplers 88%

Lower Intent to Participate

33% of Initial Training on Line

Devote Less Time (2 Days) to Initial Training

Would Make Fewer Changes to Practice

Preferred More Follow-up Training

Introduce Moderate Modifications to EBP


Basic Dissemination

2 Days of Initial Training

2 Days of Follow-up Training

33% Skills 67% Knowledge

25% Change in Practice

Enhanced Dissemination

4 Days of Initial Training

4 Days of Follow-up Training

67% Skills 33% Knowledge

50% Change in Practice

Source: Cunningham, Barwick Administration and Policy in Mental Health (2018)

0

20

40

60

80

100

Basic

Implementation

Enhanced

Implementation

49.2 50.8

80.3

19.7

Perc

ent

Accelerted Adopters

Segment 2

Randomized First Choice Simulation: Basic vs Enhanced


0

20

40

60

80

100

Basic

Implementation

Enhanced

Implementation

23.6

76.4

37.4

62.7

Perc

ent

Segment 1 Segment 2

Randomized First Choice Simulation: Basic vs Enhanced + 67% Supervisor Support


Selected Implications

Segments with Different KT Design Preferences

Change Ready versus Demand Sensitive

Outcome Sensitive vs Process Sensitive vs Demand Sensitive

Accelerated Adopters vs Gradual Adopters

Different Pattern of Preferences Across Professional Groups

Social Context Exerts More Influence Than Evidence Base

Administrators, Supervisors, Principals

Colleagues

Unions

Most Prefer Learning in a Small Group (N=10 Social Context)

An Engaging Expert Facilitator Important

Interest in Internet Options Varies Across Professional Groups

Highest in Mental Health Professionals in Accelerated Adopters Segment

Merci!

Documents

Modeling Waiting List Reduction Strategies Using …...Discrete Choice Conjoint Experiments Goals of the Talk Brief Introduction to Preference Modelling Discrete Choice Experiments