Mixed Methods in Program Evaluation Presented by Tom

“Mixed Methods in Program Evaluation” Presented by Tom Chapel Thomas J. Chapel, MA, MBA Chief Evaluation Officer Centers for Disease Control and Prevention Tchapel@cdc. gov 404 -639 -2116

Agenda 1. The why and how of mixed methods: • • Rationale Options Challenges Criteria for making choices 2. Apply points to some simple examples

CDC’s Evaluation Framework STEPS Engage stakeholders Ensure use and share lessons learned Standards Utility Feasibility Propriety Accuracy Justify conclusions Gather credible evidence Describe the program Focus the evaluation design The Standards apply especially when we’re trying to make data collection choices.

CDC’s Evaluation Standards Utility Feasibility Propriety Accuracy The Standards provide a quick and easy way to identify the 2 or 3 best data collection choices for this evaluation.

CDC’s Evaluation Framework STEPS Not “Collect data” Not “Analyze data” Rather… Engage stakeholders Ensure use and share lessons learned Standards Utility Feasibility Propriety Accuracy Justify conclusions “Gather credible evidence” Gather credible evidence Describe the program Focus the evaluation design

Steps 1 -3 Help You Focus Design And Data Collection Choices After the first 3 steps of the Evaluation Framework, we know which evidence will work for these stakeholders in this situation. Qualitative data? Quantitative data? Randomized control trials? Performance measures?

CDC’s Evaluation Standards Utility Feasibility Propriety Accuracy The Evaluation Standards help us narrow down our data collection choices to the handful of methods that will work for this evaluation at this time.

Mixed Methods Data collection methods that will work for this evaluation at this time sometimes means surveys or focus groups. But sometimes there is no one best way. The best choice would be a combination of methods or “mixed methods”.

Six (Most) Common Ways to Collect Data ● Surveys ● Interviews ● Focus Groups ● Document Review ● Observation ● Secondary Data

How Standards Inform the Choice of Methods Consider the context : Standards Utility Feasibility Propriety Accuracy • How soon do I need the results? • What resources can I use? • Are there any ethical issues to consider?

How Standards Inform the Choice of Methods Also consider the content : Standards Utility Feasibility Propriety Accuracy • Sensitivity of the issue

How Standards Inform the Choice of Methods Also consider the content : Standards Utility Feasibility Propriety Accuracy • “The Hawthorne Effect” Will the act of being observed cause someone to distort their response?

How Standards Inform the Choice of Methods Also consider the content : Standards Utility Feasibility Propriety Accuracy • Validity • Reliability

Mixed Methods Address Concerns Key Concept: Regardless of the method, when there are validity and reliability concerns, often using more than one method-- i. e. , mixed methods-will help.

Mixed Methods: Definition “The combination of at least one qualitative and at least one quantitative component in a single research project or program. ” (Bergman 2008)

Use Complementary Methods Mixed methods is: A combination of methods that has complementary strengths and nonoverlapping weaknesses. The purpose is to supplement or complement the validity and reliability of the information.

Strengths of Quantitative Methods Strengths of quantitative methods: • • • Require less time than qualitative methods Cost less Permit researcher control Quantitative data is considered to be “scientific” Easier to explain validity and reliability Easily amenable to statistical analyses

Strengths of Qualitative Methods Choose qualitative methods when you are trying to: • Explore or describe a phenomenon

Strengths of Qualitative Methods Choose qualitative methods when you are trying to: • Look for induction (i. e. , “surprise”)

Strengths of Qualitative Methods Choose qualitative methods when you are trying to: • Identify patterns

Strengths of Qualitative Methods Qualitative data can help you understand not just “what” but “WHY”.

When to Use Mixed Methods 1. Corroboration • better understanding; more credibility • “triangulation” – measuring the same thing from several different viewpoints 2. Clarification • trying to understand why we got this result

When to Use Mixed Methods are most commonly used for: 3. Explanation – similar to clarification • want to know the “why” or “what” behind the situation 4. Exploration – similar to explanation • charting new territory • trying to observe patterns • examine different situations and varying results to induce patterns

Number of Project Facets Reported via Each Data Collection Method Source: Gregory Guest, Ph. D

Number of Project Facets Reported via Each Data Collection Method This is an example of using a qualitative method (site visits) to corroborate a quantitative method (surveys). The result was increased validity of the data. Source: Gregory Guest, Ph. D

Which to Choose? How do you choose which methods to use? Which method comes first, the quantitative or the qualitative? You have a lot of flexibility in these decisions.

Parallel or Concurrent Mixed Methods QUANTITATIVE QUALITATIVE For “parallel” or “concurrent” mixed methods, quantitative and qualitative data collection happen at the same time.

Sequential Mixed Methods QUANTITATIVE QUALITATIVE OR QUALITATIVE QUANTITATIVE For “sequential” mixed methods, either quantitative or qualitative data collection can happen first.

Example of Sequential Mixed Methods to Corroborate Data QUALITATIVE In this case, the qualitative method (site visits) was used to corroborate the quantitative (survey) method and the results were different. QUANTITATIVE

Mixed Methods Is Your Choice You are never required to use mixed methods. However, you may choose to use mixed methods when: • you have some indication that a single method may give you incorrect data. • a single method may give you an incorrect perception of reality.

Mixing Methods During Data Analysis Qualitative data (focus groups, observations, secondary data, etc. ) can be converted to “numbers” via quantitative techniques like content analysis. This is also a mixed method design approach.

Mixing Methods During Data Analysis Qualitative data can be very complex. Examining qualitative data with quantitative techniques helps to identify or validate patterns or themes.

Deciding When To Use Mixed Methods and How Key Concept: Using mixed methods is a deliberate design decision. You use it when you don’t trust the data from any single method. The reason for your uncertainty determines the methods you choose to mix and the order in which you use them.

Example 1 Concurrent Design Problem or Purpose: Validity • Do people give similar responses on surveys as well as in focus groups? Example • Survey (quantitative) and focus groups (qualitative) are conducted concurrently with similar participants.

Example 2 Explanatory Sequential Design Problem or Purpose: Explain unexpected results • Use a qualitative method to explain “blindside” results from a quantitative method. Example • Survey (quantitative) followed by focus groups (qualitative) to explain or to better understand what’s going on.

Example 3 Exploratory Sequential Design Problem or Purpose: Verify suspected patterns Example • Explore potential patterns with a qualitative method and then verify the patterns with a quantitative follow-up. • Focus groups (qualitative) first, identify potential patterns, then do a survey (quantitative) to validate any patterns.

Design Options Summary You mix quantitative and qualitative methods in a different order depending on the presenting problem: Ø Validate results

Design Options Summary You mix quantitative and qualitative methods in a different order depending on the presenting problem: Ø Validate results Ø Explain the unexpected

Design Options Summary You mix quantitative and qualitative methods in a different order depending on the presenting problem: Ø Validate results Ø Explain the unexpected Ø Explore new themes

Selected Resources (Page 1 of 2) Caracelli, V. and J. Greene (eds. ). 1997. Advances in Mixed. Method Evaluation: The Challenges and Benefits of Integrating Diverse Paradigms. San Francisco, CA: Jossey-Bass. Creswell, J. and V. Plano Clark. 2010. Designing and Conducting Mixed Methods Research, 2 nd edition. Thousand Oaks, CA. Sage Publications. Morse, J. and L. Niehaus. 2009. Mixed Method Design: Principles and Procedures. Walnut Creek, CA. Left Coast Press.

Selected Resources (Page 2 of 2) Johnson, R. Burke, and L. Christensen. Evaluation Methods. 2008. www. southalabama. edu/coe/bset/johnson/ Plano Clark, V. and J. Creswell. 2008. The Mixed Methods Reader. Thousand Oaks, CA: Sage Publications. Teddlie, C. and Tashakkori, A. 2009. Foundations of Mixed Methods Research: Integrating Quantitative and Qualitative Approaches in the Social and Behavioral Sciences. Thousand Oaks, CA. Sage Publications.

Recommended Resource Creswell, J. and V. Plano Clark. 2010. Designing and Conducting Mixed Methods Research, 2 nd edition. Thousand Oaks, CA. Sage Publications.

The Community Tool Box http: //ctb. ku. edu Chapter 37, Section 5. Collecting and Analyzing Data

End “Mixed Methods” Return to Webinar 4: Gathering Data, Developing Conclusions, and Putting Your Findings to Use Return to Evaluation Webinars home page