Title: Evidence Based Chronic Disease Prevention Module 8: Evaluating The Program Or Policy
1Evidence Based ChronicDisease Prevention Module
8 EvaluatingThe Program Or Policy
- Presented by Karen Peters, DrPH
2Objectives
- Understand the basic components of program
evaluation. - Describe the differences and unique contributions
of quantitative and qualitative evaluation. - Understand the concepts of measurement validity
and reliability.
3Develop statement of the issue
Determine what is known in the literature
Dissemination
Quantify the issue
Evaluate the program or policy
Develop program or policy options
Develop an action plan
4Overview
- The whats and whys of evaluation
- Study designs and measurement issues
- Types of evaluation
- Quantitative
- Qualitative
- Organizational issues in evaluation
- Infeasible to provide in-depth discussion
mechanics - e.g., sampling/instrument development
- Can show basic components in an evaluation, what
to look for, where to turn for help
5What is Evaluation?
- A process that attempts to determine as
systematically and objectively as possible the
relevance, effectiveness, and impact of
activities in light of their objectives. - From Last JM. A Dictionary of Epidemiology.
Third Edition. New York Oxford Press 1995. - Complex and diverse field
6Some questions to consider?
- What are the most significant challenges you face
in program evaluation? - In program planning when should you begin
planning an evaluation?
7Why Evaluate?
- Public health agencies need to be accountable
- Basis for choices when resources are limited
- Helps to determine costs and benefits
- Source of information for improving programs and
policies - Increasingly mandated by funders/policy makers
- Convincing funders is not always easy
- Leads to research questions that can be tested in
other studies/programs
8New Directions
- Social programs have become more broad, complex
and interactive - Seek to bring about changes in community
capacity, social support, decision-making,
control over resources and individual behavior - Time to supplement traditional strategies with
new approaches reflecting complexity of
community-based initiatives
9New Directions
- More flexible evaluation approaches can play role
in building community capacity and self
determination - Need to re-direct program evaluation toward
community-based public health values - Traditional evaluation conducted by experts to
determine if program objectives met,
strengths/weaknesses/replicability and
contribution to scientific knowledge
10New Directions
- Some evaluators believe communities lack skills
to design, engage in and interpret evaluations - However, experts may lack insight/ flexibility
needed to capture essence of community projects
or to answer questions raised by communities and
CBOs
11New Directions
- Community-based evaluation perspective involves
more participatory and inclusive process that
incorporates the values, knowledge, expertise and
interests of the community and uses evaluation as
a tool for community capacity building - Community involved as full/equal partner allows
for development of more relevant program
success measures and produces data that are
useful in community settings
12New Directions
- Evaluation is 1 part of broader planning process
but can help in - reflecting on progress
- document where going, where coming from
- share what worked and what did not with other
communities - demonstrate need for targeted resources to
address community issues - illustrate impact of community based initiatives
to decision/policy makers - provide information on developing meaningful
community based indicators
13Research Phases Model(Greenwald and Cullen, 1985)
- Phase 1 Hypothesis Development
- Ex Link between sedentary behavior - obesity
- Phase 2 Methods Development
- Pilot test of intervention to increase physical
activity and validation of measures - Phase 3 Controlled Intx Trial (Efficacy)
- Small scale randomized trial of physical activity
- Phase 4 Defined Population (Effectiveness)
- Larger scale trial of PA in populations of
interest - Phase 5 Demonstration (Dissemination)
- Evaluation of results of PA program
14Evaluation Models
- For Practitioners
- Models identify key factors to consider when
developing/selecting health behavior programs - Factors to focus on when reading the literature
- For Researchers
- Models identify important dimensions to be
included in program evaluations
15Evaluation Models
- 2 Major Frameworks (there are many others)
- For Process and Formative Evaluation
- CDC Framework for Public Health Programs
- For Impact and Outcome Evaluation
- RE-AIM Framework
16Framework for Program Evaluation
- 2 year process by CDC
- designed as a framework for ongoing, practical
program evaluation - can be integrated with routine program program
operations - involves input from program staff, community
members, other stakeholders, not just evaluation
experts - Involves 6 basic steps and 4 broad evaluation
standards
17Broad Evaluation Framework (ala CDC)
- Engage stakeholders
- Describe the program
- Focus the evaluation design
- Gather and analyze evidence
- Justify conclusions
- Ensure use and share lessons learned
- CDC Framework for Program Evaluation in Public
Health, 9/17/99 and Center for Advancement of
Community Based Public Health, June 2000
18CDC Framework for Evaluation
- 4 Evaluation standards
- Guidelines that can help assess whether an
evaluation is well designed - Utility Is the evaluation useful?
- Does the evaluation answer questions that are
relevant to the stakeholders - Feasibility Is the evaluation practical?
- Is the evaluation realistic and cost-effective
19CDC Framework for Evaluation
- Propriety Is the evaluation ethical?
- Does the evaluation consider the rights and
interests of those involved and affected - Accuracy Is the evaluation correct?
- Do the evaluation findings convey information
that is correct and technically adequate
20CDC Framework for Evaluation
- Step 1 Engage the stakeholders
- Stakeholders - those involved in implementing the
program and those served or affected by it
including decision-makers who can do something
with the results - Standards for step 1 Utility and Propriety
- Utility Have you identified individuals and
organizations affected by the evaluation? - Are those involved in the evaluation trustworthy
and competent?
21CDC Framework for Evaluation
- Propriety
- Is there an explicit, written agreement about
what is to be done, how, by whom and when? - Does the evaluation design protect the rights and
welfare of those involved? - Are the individuals who are conducting the
evaluation interacting respectfully with
stakeholders? - Have you discussed conflicts of interest openly
and honestly?
22CDC Evaluation Framework
- Step 2 Describe the program
- summarize the program being evaluated with a
statement of need which includes expectations, a
logic model, specifying resources available to
conduct program activities and how program fits
into larger organizational/community context - A good description allows the program to be
compared to similar efforts and makes it easier
to figure out what parts brought about what
effects
23CDC Evaluation Framework
- Standards for Step 2 Accuracy/Propriety
- Accuracy Have you clearly and accurately
described the program - Have you documented the program context
- Propriety Is the evaluation complete and fair
- Does it assess program strengths and weaknesses
24CDC Evaluation Framework
- Step 3 Focus the evaluation design
- Specify the evaluations overall intent
- Determine who and what the evaluation is for
(users and uses) - What questions the evaluation should examine
- What methods are best to answer the questions
25CDC Evaluation Framework
- Purpose of evaluation depends on programs stage
of development - New or developing program - feasibility of
intervention approach - Implementation - fine tuning or changes needed
- Established program - assess program effects
- Standards for Step 3
- Feasibility
- Propriety
- Accuracy
26CDC Evaluation Framework
- Feasibility
- Have you considered the political interests/needs
of groups and obtained buy-in - Does the information produced justify the costs
- Are evaluation procedures practical
- Propriety
- Does evaluation design help to identify needs
- Are costs guided by sound/ethical accountability
procedures - Accuracy
- Is there an accurate description of evaluation
purposes and procedures
27CDC Evaluation Framework
- Step 4 Gather credible evidence
- Need well rounded picture of the program
- Develop indicators that translate program
concepts into specific measures - Use multiple sources of evidence that reflect
different perspectives about the program - Techniques used to gather and handle evidence
should be compatible with cultural conditions in
each program setting
28CDC Evaluation Framework
- Standards in Step 4 Utility and Accuracy
- Utility Are you collecting information that
addresses pertinent program issues and is
responsive to stakeholder needs - Accuracy Have you adequately described your
sources of information - Do data collection procedures address internal
validity and reliability issues - Is there a system in place for identifying and
correcting errors
29CDC Evaluation Framework
- Step 5 Justify Conclusions
- Involves making claims about a program based on
evidence gathered - Stakeholder values provide basis for making
judgements about program merits/performance - Conclusions are based on analysis,synthesis and
interpretation of information to detect patterns
and result in recommendations - Reaching good conclusions requires variety of
stakeholder perspectives
30CDC Evaluation Framework
- Standards for Step 5 Accuracy/Utility
- Accuracy Has data analysis process been
effective in answering key evaluation questions - Can you explicitly justify your conclusions
- Utility Have you carefully described the
perspectives, procedures and rationale used to
interpret the findings
31CDC Evaluation Framework
- Step 6 Ensure use, share lessons learned
- Make sure stakeholders understand the evaluation
procedure and findings - All participants should have the opportunity to
provide feedback - Evaluators should provide any needed follow-up
- Use a variety of communication strategies to
disseminate evaluation results in a timely and
unbiased manner
32CDC Evaluation Framework
- Standards for Step 6 Utility, Propriety,
Accuracy - Utility - do evaluation reports describe the
program context, purpose, procedures and findings
clearly? - Propriety - Have evaluators made sure findings
(including limitations) are accessible to
everyone affected by the program - Accuracy - do evaluation reports reflect the
findings fairly and impartially?
33RE-AIM Evaluation FrameworkGlasgow, Vogt, Boles,
1999Glasgow, McKay, Piette, Reynolds, 2001
- Reach, Efficacy or Effectiveness (depending on
research phase), Adoption, Implementation,
Maintenance - RE-AIM relies on 2 comprehensive models
- PRECEDE-PROCEED (Green Kreuter 1999)
- Diffusion Theory (Rogers 1995 Nutbeam 1996)
34Dimensions of RE-AIM
- Reach
- Individual Level
- What of potentially eligible participants will
take part - How representative are they?
- Efficacy or Effectiveness
- Individual Level
- What was the impact on all who began?
- What was the impact on intermediate primary
outcomes? - What was the positive/negative (unintended)
outcomes, including quality of life?
35Dimensions of RE-AIM
- Adoption
- Setting Level
- What of settings/intervention agents will
participate? (worksites, schools, educators,
nurses) - How representative are they?
- Implementation
- Setting or Agent Level
- To what extent were the intervention components
delivered as intended (in the protocol), when
conducted in applied settings by non researchers
36Dimensions of RE-AIM
- Maintenance
- Individual and Setting Levels
- Individual Level What are the long term effects
(minimum 6-12 months following intervention)? - Setting Level To what extent are different
intervention components continued or
institutionalized?
37Common Challenges and Suggested Strategies Using
RE-AIM Framework
- Reach
- Challenge Not including a relevant, high risk,
or representative sample - Strategy Use population-based recruitment or
over sample high risk groups, reduce exclusion
criteria - Efficacy or Effectiveness
- Challenge Ambiguous outcomes
- Strategy Assess broader set of outcomes, conduct
subgroup analyses, use different assessment points
38Common Challenges and Suggested Strategies Using
RE-AIM Framework
- Adoption
- Challenge Program never adopted or endorsed -
used only in academic settings - Strategy Involve participants in all phases,
approach numerous settings early on while
revision is still possible - Implementation
- Challenge Protocols not delivered as intended
(Type III error) - Strategy Assess treatment as too complicated,
intensive, incompatible involve non-researchers
39Common Challenges and Suggested Strategies Using
RE-AIM Framework
- Maintenance
- Challenge Program or effects not maintained over
time - Strategy Include maintenance phase in protocol
and evaluation plans Leave treatment behind
after study and plan for institutionalization
40Common Study/Evaluation Designs
- Experimental/randomized
- Quasi-experimental
- Time-series
- Use of existing data
41Study/Evaluation Designs
- Quasi-experimental
- Increasing attention
- At least one intervention and one comparison
group, without randomization - Appeal of intervening thru intact social groups
- See Koepsell chapter in readings (in Module 4)
42Study/Evaluation Designs
- Use the best designs feasible
- Pre- post-data
- Comparison groups
- Complete program records
- Reliable and valid measures
- Proper analytic techniques
- Review principles/tools from Goodman, page 39 (in
Module 8)
43Challenges in CommunityWide Studies
- Varying degrees of intervention exposure
- Running programs in multiple locations
- Accounting for community-level variance
- Lack of sensitivity of the community
- Concepts of participatory research
- Equity, collective decision making
- High-quality, ethical research
- Addressing social inequalities
- Maximize learning opportunities
- See Goodman article on community capacity
44Challenges in CommunityWide Studies
- Community-level variance
- Individuals in communities, neighborhoods,
schools, worksites are correlated - ICC (intra-class correlation coefficient)
- People have related characteristics (not
independent) - In practical terms, means increased chance of
Type I error (saying there is a difference when
there really is not)
45Measurement Issues
- Components of a good evaluation
- Adequate sample size
- High validity
- High reliability
- Sample size considerations
- Number of communities
- Number of individuals per community
- Increasing number of communities versus number of
individuals per community - Can rely on simple, accessible programs like Epi
Info
46Concepts of Validity and Reliability and their
Importance
- Measurement Issues
- Evaluation threats
- Validity
- Is the instrument or design measuring exactly
what was intended? (Self report vs.... biologic
test) - Reliability
- Is the measurement being conducted consistently?
(Face to face vs...... telephone, different
interviewers)
47Measurement Issues
- Validity best available approximation to the
truth - Internal Validity
- The extent of causality (the effects are really
attributable to the program) - External Validity
- The extent of generalizability
- Importance?? (so what???)
48Measurement Issues
- Major threats to validity
- Low statistical power
- Violated assumptions in statistical tests
- Reliability of measures
- Reliability of treatment implementation
- Random confounders in the experiment
- Random heterogeneity of respondents
- adapted from Cook and Campbell, 1979
49Measurement Issues
- Reliability (repeatability)
- Consistency in measurement
- Multiple types
- Inter-observer
- Test-retest
- Internal consistency
50Measurement Issues Examples
- Validity
- Self-reported rate of having a health
professional check for hemoglobin A1C among
diabetics in an intervention program compared
with clinic records - Reliability
- Test-retest data from the BRFSS on self-report of
seeing a health care professional in the last
year for diabetes among diabetics in Illinois
51Measurement Issues Ensuring Validity and
Reliability in Evaluation
- Literature/contacting researchers may show you
accepted methods - Multiple statistical methods available to report
validity and reliability - Evaluation instruments often need community
contouring - Participatory methods may prevent use of existing
instruments/questions
52Four Purposes of Evaluation
- Needs Assessment
- Process
- Impact
- Outcome
53How do you decide?
- Use data available
- What does the data tell you?
- What other information do you want?
- What other programs are already available in
different locations? - Reality constraints Staffing, resources, time
54Needs Assessment
- Diagnostic Evaluation
- Context Evaluation
- feedback on knowledge, attitudes, risk behaviors,
health status, and perceived needs of the target
population and of the status of available health
promotion programs
55Process Evaluation
- Formative evaluation
- feedback on program implementation site
response participant response including
appropriateness of materials, methods,content
practitioner response, and personnel competency
56Example IL Arthritis Awareness
- 3 components
- brochure alone
- brochure with community outreach
- brochure, community outreach, physician education
- What would you want to know about each component?
57Types of Evaluation Process evaluation
- Field of Dreams evaluation
- If you build it, will they come?
- Shorter-term feedback on program implementation,
content, methods, participant response,
practitioner response - What is working, what is not working
- Uses quantitative or qualitative data
- Data usually involves counts, not rates or ratios
58Types of Evaluation
- Considerations for process evaluation
- Sources of data
- Program data
- Limitations of data
- Time frame
- Availability
- Example
- Number of diabetics being obtaining foot
examinations statewide through primary care
providers
59Impact Evaluation
- Summative evaluation
- long or short term feedback on knowledge,
attitudes, beliefs and behavior change (KABBBs)
of participants (skills development), programs
and policies of organizations and governments - Time 1 - decide what to include in intervention
- Time 2 - did the intervention make a difference
- Conduct survey or interview participants
60Impact Evaluation
- Uses quantitative or qualitative data
- Probably more realistic endpoints for most public
health programs and policies
61Impact Evaluation
- Considerations for impact evaluation
- Sources of data
- Surveillance or program data
- Limitations of data (validity and reliability)
- Time frame
- Availability
- Example
- Eye examination rates of diabetics in Illinois
62Types of Evaluation
- Outcome evaluation
- Long-term feedback on health status, morbidity,
mortality, disability or quality of life - Uses quantitative data
- Present throughout integrated strategic plan
- Also called summative evaluation
63Types of Evaluation
- Considerations for outcome evaluation
- Sources of data
- Routine surveillance data
- Limitations of data (validity and reliability)
- Time frame
- Availability
- IDPH IPLAN website
- Example
- Rates of end stage renal disease among diabetics
in Illinois
64Types of Evaluation Outcome evaluation
- Some move toward standardizing (outcome)
indicatorse.g., CDCs consensus indicators - What if yours is not on the list?
- Race-ethnicity-specific infant mortality rate
- Motor vehicle crash death rate
- Work-related injury death rate
- Suicide rate
- Lung cancer death rate
- Breast cancer death rate
- Cardiovascular disease death rate
- Homicide rate
65Types of Evaluation Outcome evaluation
- All-cause mortality rate
- AIDS incidence
- Measles incidence
- Tuberculosis incidence
- Syphilis incidence
- Incidence of low birth weight
- Births to adolescents
- Prenatal care
- Childhood poverty
- Proportion of persons living in counties
exceeding EPA standards for air quality
66Types of Evaluation Outcome/impact evaluation
- Common public health data sources
- Vital events registries
- Disease specific registries
- Population and house census
- Routine health services records
- Health programs delivery records
- Hospital discharge data
- Disease notification within surveillance systems
- Sample surveys within surveillance systems
- Other sample surveys
- Other data banks from programs outside the health
sector
67Data Questions for Consideration
- What type of evaluation is this?
- How would you analyze these data?
- What would be your initial conclusions?
- What other data might support or detract from
your conclusions? - What are some limitations of your risk factor
data?
68Types of Evaluation
- Outcome
- Doing the Right
Things - Impact
- (performance measures)
- Process Doing Things Right
69Evaluation Framework
- Program
- -
Instructors? - - Content?
- Process -
Methods? - - Time
allotment? - -
Materials? - Behavior
- - Knowledge gain?
- Evaluation Impact -Attitude change?
- Types -Habit change?
- -Skill development?
- Health
- - Mortality?
- Outcome -Morbidity?
- -Disability?
- (Adapted from Green et al, 1980) -Quality of
life
70Evaluation Polarities
- Formative Summative
- Qualitative Quantitative
- Process Outcome
- Impact
71Types of Evaluation
- Quantitative versus qualitative methods
- Avoid choosing one or the other
- Generally more familiarity with quantitative
methods
72Quantitative vs.... Qualitative
- Quantitative (numbers) surveillance data,
surveys, records - Qualitative (words) observations, in-depth
individual interviews, group interviews, focus
groups, diaries
73Why Qualitative?
- When there are no quantitative instruments
available - When you are not sure if the measures are
appropriate for the population you are working
with - When you are not sure if you are asking the right
questions - When you want to understand program processes not
just impact and outcomes
74Why Qualitative?
- When you want more in depth information about
program implementation - - What do clients experience,
- what services are provided to clients,
- How is the program organized?
- What do staff do? How has the program developed?
- When you want detailed, descriptive information
about the program for the purpose of improving
the program
75Observation Dimensions
- Role of observer
- full participant, partial participant,
onlooker/observe as outsider - Covert or overt
- Do others know you are observing?
- Duration of observation
- Short term vs..... long term
- Focus of observation
- narrow, single component of program vs..... whole
76Types of Questions You Can Ask
- Behavioral/experience
- Opinion/value
- Feeling
- Knowledge
- Sensory
- Demographic/background
- (ALL from past, present, future)
77In Depth Interviewing Types
- Informal conversational interview
- General interview guide
- Standardized open-ended interview
78Informal Conversational Interview
- No interview guide no predetermined set of
questions - Particularly useful when do not know what types
of questions to ask and when yo will be in the
environment for some time - Each interview builds on the previous one -
flexibility and responsiveness to previous
experiences and the environment
79Informal Conversational Interview
- Requires a long time to get systematic
information - Open to interviewer effects
- tired
- conversational skills
- capacity to build rapport
80Interview Guide
- List of questions or issues to be explored - a
framework - Can get similar information from several
individuals - General topics and probes
- General outline to discuss on predetermined
subject - Maintain conversational quality
81Standard Open-Ended Interview
- Set of carefully worded and arranged questions
- Less flexibility of issues to explore or probing
- Minimizes variation in questions asked, thus
reduces bias - Useful when have several interviewers
82Focus Group and Group Interviews
- Relatively homogeneous group
- 6-10 people
- Semi-structured
- Focus group - build on each others ideas -
advantage of group processcan see influence of
social networks on issue at hand - Group interview - several people together, not
necessarily take advantage of group process
83Wording of Questions
- Open ended - do not presuppose dimensions of
feeling, thoughts, or experiences - Clear - use appropriate language, reflect back
language you hear - Singular questions - can preface with going to
ask about strengths, weaknesses, likes, dislikes
- then ask each individually - Neutrality re subject and rapport with individual
84Wording of Questions
- Acknowledge multiple experiences - can provide
illustrations or examples some have said this,
others that, still others have different opinions
- what do you think? - Avoid leading questions
85Probes
- Can you give me an example of that?
- Can I stop you here for a minutecan you tell me
a bit more about x can re-direct conversation - Does anyone else have other experiences with this?
86Recording Data
- Permission and consent
- Notes - during and after the interview include
your feelings tired/excited how did the
interview seem to you? - Note surroundings
- Tape recording - make sure it is on, make sure it
works
87Sampling
- Purposeful sampling - notion of gaining
information from those who have it - extreme or deviant cases - outstanding successes
or notable failures - maximum variation - look for common themes across
multiple participants or programs - small sample,
great diversity - homogenous samples - individual or focus groups
88Sampling
- Snowball or chain sampling
- recommended informants - who else should I speak
with - Criterion sampling
- all cases who meet some predetermined criteria
89Analysis
- From transcription and notes
- Focused coding - with predetermined categories in
mind - Open coding - categories and themes from the data
itself - Multiple coders
- Label so you can go back to context
- Triangulation
90Reliability and Validity
- Neutrality rather than objective and subjective
- impartial, not predisposed toward certain
findings ahead of time, no predetermined results
to support - Notion of generalizability vs.... extrapolation
- generalizable - impossible because nothing is
context free - extrapolation - speculations on the likely
applicability of findings to other situations
under similar, not identical, conditions
91Other Considerations in Qualitative Evaluation
- Involve stakeholders in development of program
objectives and evaluation questions - Evaluation requires clear program objectives
- Measure program processes, impacts, and outcomes
using measures appropriate for the questions
asked - Expect frustration from those collecting data,
resistance from those feeling judged
92Organizational Issues/Summary
- Program personnel may be threatened by the
evaluation - Need to maintain objectivity
- Involvement may reduce resistance but may
threaten objectivity - Trade-off of comprehensive evaluation versus
nothing at all - Think of 10 Rule as you design and implement
programs
93Organizational Issues/Summary
- How can I do evaluation when theres so much
real work to do? - Independent (outside) evaluation may be useful
- What to look for in a good evaluation
- Remember the use of multi-disciplinary teams
94A Word on Coalitions and Partnerships
- Seems to be an interest in and emphasis on the
use and effectiveness of coalitions as part of
public health programming - Little is known currently about impact and
outcome measures for coalitions - More is known about process of coalitions
- Both qualitative and quantitative measures are
needed in this area
95Other Resources on Evaluation
- World Wide Web Virtual Library Evaluation
- http//user.berlin.de/alfio.cermi/index.html
- CDC Evaluation Resources
- http//www.cdc.gov/eval/over.htm
- CDC Framework for Program Evaluation in Public
Health - http//www.cdc.gov/mmwr/preview/mmwrhtml/rr4811a1/
htm - Center for the Adv. of Comm Based Public Health
(www.cbph.org) - University of Toronto (http//www.utoronto.ca/shp/
hcu) - Univ.of WI Ext http//www.uwex.edu/ces/pdante/eval
uat/htm - Kellogg Fndt http//www.wkkf.org/Publications/eva
lhdbk/
96Course Objectives
- Construct a concise, measurable statement of a
public health issue or statement - Determine what is known in the scientific
literature - Describe public health issues using quantitative
data sources - Generate public health program or policy options
- Create public health program or policy action
plans - Construct a logic model
- Evaluate a public health program