Mixed Methods Research

Hui Bian Office for Faculty Excellence

Mixed Methods Research

My contact information Office: 1001 Joyner library (room 1006) bianh@ecu.edu Phone: 252-328-5428 Website: http://core.ecu.edu/StatisticsResearch
2

Mixed Methods Research

Reference
Creswell, J. W., & Plano Clark, V. L. (2011). Designing and conducting mixed methods research (2nd ed.). Thousand Oaks, CA: Sage Publications, Inc.

Mixed Methods Research

Definition
as a method, it focuses on collecting, analyzing, and mixing both quantitative and qualitative data in a single study or series of studies. Its central premise is that the use of quantitative and qualitative approaches, in combination, provides a better understanding of research problems than either approach alone.

Mixed Methods Research

Characteristics of mixed methods research
Collect and analyze both quantitative and qualitative data. Mix two forms of data in different ways. Give priority to one or both forms of data. Can be in a single study or in multiple phases of a study.

Mixed Methods Research

Strength and weakness of quantitative and qualitative methods.
Quantitative Generalization Strength and weakness Large sample Small sample details, in depth Qualitative

Mixed Methods Research

Why use mixed methods One data resource may not be enough; Initial results need to be further explained; A second method is needed to enhance a primary method; The project has multi-phases.
7

Mixed Methods Research

How to choose an appropriate mixed methods design? Level of interaction between two strands: independent or interactive. Relative priority: equal/unequal priority Timing: concurrent, sequential, or combination of those two Where or how to mix the strands: point of interface and mixing strategies
8

Mixed Methods Research

Point of interface: is a point where the two
strands are mixed: possible point of interfaces
Data collection: quan or qual results build to the subsequent collection of qual or quan data. Data analysis: transform one type of data into other type of data and analyze combined data. Interpretation: comparing or combining results from both methods.

Mixed Methods Research

Six major designs
Convergent parallel design Explanatory sequential design Exploratory sequential design Embedded design Transformative design Multiphase design
Basic designs

10

Mixed Methods Research

Major designs (1). Convergent parallel design: purpose of this design
to best understand or develop more complete understanding of the research problem by obtaining different but complementary data. Validation purpose
11

Mixed Methods Research

Convergent parallel design (diagram)
Quantitative data collection and analysis
Compare or relate Interpretation

Qualitative data collection and analysis

Point of interface
12

Convergent Parallel Design

Three published papers
Sherrilene Classen, Ellen DS Lopez, Sandra Winter, Kezia D Awadzi, Nita Ferree, et al. Population-based health promotion perspective for older driver safety: Conceptual framework to intervention plan. Clinical Interventions in Aging 2007, 2:677-693 03 January 2007 http://www.dovepress.com/population-basedhealth-promotion-perspective-for-older-driversafety--peer-reviewed-article-CIA
13

Convergent Parallel Design

Three published papers
David F. Feldon and Yasmin B. Kafai. Mixed methods for mixed reality: understanding users avatar activities in virtual worlds. Educational Technology Research and Development 2008 56:575-593 http://www.springerlink.com/content/g66m160n 75444mx7/fulltext.pdf

14

Convergent Parallel Design

Three published papers
Marsha N. Wittink, Frances K. Barg, and Joseph J. Gallo. Unwritten Rules of Talking to Doctors About Depression: Integrating Qualitative and Quantitative Methods. Ann Fam Med 2006 4:302309; doi:10.1370/afm.558 . http://www.annfammed.org/content/4/4/302.full .pdf+html

15

Convergent Parallel Design

Key points
Collect and analyze two independent strands of quantitative and qualitative data at roughly the same time/ in a single phase. Prioritize the methods equally. Keep the data analysis independent. Mix the results during the overall interpretation. Try to look for convergence, divergence, contradictions, or relationships of two sources of data.
16

Convergent Parallel Design

Procedure (Flowchart)
Collect both types of data concurrently Analyze two data sets separately Merge the results Interpret combined results

17

Mixed Methods Research

Convergent parallel design: flowchart

Point of interface

18

Convergent Parallel Design

Design Research questions: create parallel questions for the qual and quan studies.
Parallel questions mean the same concepts need to be addressed in both data collections.
19

Convergent Parallel Design

Samples: different or same group of people in quantitative and qualitative studies?
Can be same group of people Or different group of people

20

Convergent Parallel Design

Design
Samples: If the purpose is to combine information on a topic from different views, use different group of participants. If the purpose is to compare, corroborate, or relate two sets of findings, use same group of participants.
21

Convergent Parallel Design

Design
Sample sizes: equal or unequal Equal sample size Large equal sample size may sacrifice some of richness of the qualitative data. Small equal sample size may sacrifice the use of rigorous statistical tests.
22

Convergent Parallel Design

Design Data will be collected from one source or different sources: survey/interview or only use survey. Order of two types of data collections: survey first then focus group or one-on-one interview.

23

Convergent Parallel Design

Merged data analysis strategies Side-by-side comparison (in a results or discussion section or a summary table).
Present quantitative or qualitative results Followed by qualitative or quantitative results Followed by comments describe how qual/quan confirm or disconfirm quan/qual results.
24

Convergent Parallel Design

Side-by-side comparison

25

Convergent Parallel Design

Merged data analysis strategies
Joint display: using table or figure to show both quan and qual results

26

Convergent Parallel Design

Joint display

27

Convergent Parallel Design

Merged data analysis strategies
Data transformation merged analysis: transform one type of data (qual) into the other type of data (quan). Create a new variable based on presence of a theme Create a new variable based on number of times a theme appears.

28

Convergent Parallel Design

Interpreting merged results Look for similarity and convergence How to handle discrepancy? State the limitations of the study Revisit two types of data Could collect additional data
29

Convergent Parallel Design

Challenges
Needs both quantitative and qualitative expertise Consequences of having different samples and different sample size when merging two data sets. How to merge two types of data. How to deal with the situation in which quantitative and qualitative results contradict each other.

30

Mixed Methods Research

Convergent parallel design variants
Parallel-databases variants: two sets of data merge at the final step. Data-transformation variant Data-validation variant: such as open-ended questions on a questionnaire is used to validate or confirm the results from closeended questions.
31

Mixed Methods Research

Major designs (2). Explanatory sequential design: purpose of this design is to use qualitative approach to explain quantitative results (significant, nonsignificant, outliers or surprising results) or to guide to form groups based on quantitative results
32

Mixed Methods Research

Explanatory sequential design (diagram)

Quantitative data collection and analysis

Follow up with

Qualitative data collection and analysis

Interpretation

Point of interface

33

Explanatory Sequential Design

Published paper
Nataliya V. Ivankova and Sheldon L. Stick (2007). Students' persistence in a distributed doctoral program in educational leadership in higher education: A mixed methods study. Research in Higher Education, 48(1):93-135 http://www.jstor.org/stable/25704494

34

Explanatory Sequential Design

Published paper
Niobe Way, Helena Y. Stauber, Michael J. Nakkula and Perry London (1994). Depression and substance use in two divergent high school cultures: A quantitative and qualitative analysis. Journal of Youth and Adolescence, 23(3): 331-357 http://www.springerlink.com/content/l367l0l77r213712/fullt ext.pdf

35

Explanatory Sequential Design

Mixed methods question In what ways do the qualitative data help explain the quantitative results?

36

Explanatory Sequential Design

Key points
Typically it is a two-phase design. Collect quantitative and qualitative data at different time. Qualitative study depends on quantitative results. Usually quantitative data collection is the priority.
37

Explanatory Sequential Design

Procedure
First, collect and analyze quantitative data. Identify specific quantitative results that need additional explanation. Design qualitative study based on what learn from quantitative results.
38

Explanatory Sequential Design

Procedure
Collect and analyze qualitative data. Interpret combined results.

39

Mixed Methods Research

Explanatory sequential design: procedure

40

Explanatory Sequential Design

Design
Samples: different or same group of people in both studies?
The participants in the qualitative study should be those who participated in the quantitative study.

Sample sizes: equal or unequal

Qualitative study uses smaller sample.
41

Explanatory Sequential Design

Design
Decide what quantitative results to follow up.
Unclear Unexpected Significant/non-significant results Outliers or extreme cases

42

Explanatory Sequential Design

Design How to select participants for qualitative study
Individuals who volunteer to participate in interviews (weaker connection between two phases). Systematic approach: based on quantitative results and select participants best able to fit in qualitative study (IRB issue).
43

Explanatory Sequential Design

Design IRB issues: suggestions
Separate IRB for each phase. One IRB, state the follow up phase as tentative. From the start, inform participants the possibility of second data collection.
44

Explanatory Sequential Design

Select qualitative sample
Participants who are representative of different groups. Participants with extreme scores. Participants differed in their scores on significant predictors.

45

Explanatory Sequential Design

Interpreting connected results
Conclusion is about whether the follow up qualitative data provide a better understanding of the research problem than simply the quantitative results.

46

Explanatory Sequential Design

Explanatory sequential design variants
Follow-up explanation variant Participation-selection variant: it needs quantitative results to help select best participants. It places priority on the second, qualitative phase.

47

Explanatory Sequential Design

Challenges
Time consuming IRB issue Decisions about which quantitative results need further explanation. Decisions about who to sample and what criteria used for sample selection for qualitative study.
48

Mixed Methods Research

Major designs (3). Exploratory sequential design: also referred to as instrument development design. The purpose of this design is to generalize qualitative findings to a larger sample.

49

Exploratory Sequential Design

Reference for instrument design
DeVellis, R. F. (2003). Scale development: theory and application (2nd ed.). Thousand Oaks, CA: Sage Publications, Inc. Downing, S. M. & Haladyna, T. M. (2006). Handbook of test development. Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Fishman, J. A. & Galguera, T. (2003). Introduction to test construction in the social and behavioral sciences: a practical guide. Lanham, MD: Rowman & Littlefield Publishers, Inc. Pett, M. A., Lackey, N. R.,& Sullivan, J. J. (2003). Making sense of factor analysis: the use of factor analysis for instrument development in health care research. Thousand Oaks, CA: Sage Publications, Inc.

50

Exploratory Sequential Design

Published paper
Myers, Karen Kroman; Oetzel, John G. (2003). Communication Quarterly, 51(4), 438-457.
http://ehis.ebscohost.com/ehost/pdfviewer/pdfviewer?vid= 3&hid=2&sid=2339ee9b-08f8-45b1-babfb7e2c0d193ef%40sessionmgr12

51

Exploratory Sequential Design

Design diagram

Qualitative data collection and analysis

Builds to

Quantitative data collection and analysis

Interpretation

Point of interface

52

Exploratory Sequential Design

Purpose of this design:
The qualitative phase is used to help develop or inform the quantitative study.
Instrument design (explore) Grounded theory (generalize qualitative results)
53

Exploratory Sequential Design

Reasons for using this design
Instruments are not available The variables are not known There is no theory or model as a guide

54

Exploratory Sequential Design

Key points
Typically it is a two-phase design.
Three phases for instrument development (instrument development phase, a phase testing, and apply the instrument).

Collect quantitative and qualitative data at different time. Qualitative results can help and inform the second quantitative method.
55

Exploratory Sequential Design

Mixed design research question In what ways do the quantitative results generalize the qualitative findings?

56

Exploratory Sequential Design

Procedure
First, collect and analyze qualitative data. Develop quantitative study based on what you learn from qualitative results. Collect and analyze quantitative data.

57

Mixed Methods Research

Exploratory sequential design: flowchart

58

Exploratory Sequential Design

Design
Samples: different or same group of people in both studies?
The participants in the quantitative study are NOT same individuals who provided qualitative data.

Sample sizes: equal or unequal

Quantitative study uses larger sample.
59

Exploratory Sequential Design

Design IRB issues for emerging follow-up phase:
Separate IRB for each phase. One IRB, state the follow up phase as tentative.

60

Exploratory Sequential Design

Design Decide what qualitative results to use.
Useful quotes Codes > variables Themes > constructs

61

Exploratory Sequential Design

Design How to develop a good instrument: scale development. Steps for instrument development

62

Exploratory Sequential Design

Step 1

Determine what you want to measure Generating an item pool Determine the format for items Expert review of initial item pool Add social desirability items Pilot testing and item analysis Administer instrument to a larger sample Evaluate the items Revise instrument
DeVellis (2003); Fishman & Galguera (2003); Pett, Lackey, & Sullivan (2003)
63

Step 2

Step 3

Step 4

Step 5

Step 6

Step 7

Step 8

Step 9

Exploratory Sequential Design

Exploratory sequential design variants
Theory-development variant: test emergent theory Instrument development variant: initial qualitative phase plays a secondary role.

64

Mixed Methods Research

Major designs (4). Embedded design: purpose of this design is to answer different questions that requires different types of data.

65

Mixed Methods Research

Embedded designs
Quantitative or qualitative design
Quantitative or Qualitative data collection and analysis

Interpretation
Qualitative or quantitative data collection and analysis

66

Mixed Methods Research

Embedded design
A quantitative or qualitative data collection is within a quantitative or qualitative procedure. A single data set is not enough. Two types of data answer different research questions. The collection and analysis of the second data set may occur before, during, and/or after the first data collection.
67

Mixed Methods Research

Examples of embedded design: Qualitative data in quantitative study:
Develop an instrument in an intervention trial. Try to understand the impact of the intervention on participants. Test long term effects of an intervention after a trial.

68

Mixed Methods Research

Embedded design: procedure

69

Mixed Methods Research

Embedded design variants
Embedded-experiment variant: qualitative data within an experiment trial. Embedded instrument development and validation variant. Mixed methods case studies Embed both quantitative and Mixed methods narrative research qualitative data within traditional Mixed methods ethnography qualitative designs.
70

Mixed Methods Research

Published paper
Victor, C. R., Ross, F., & Axford, J. (2004). Capturing lay perspectives in a randomized control trial of a health promotion intervention for people with osteoarthritis of the knee. Journal of Evaluation in Clinical Practice, 10(1), 63-70.

71

Mixed Methods Research

Major designs
(5). Transformative design: the purpose of this design is to address issues of social justice and call for change for underrepresented or marginalized populations. This design more relates to the content than to the methodology. Is beyond first four basic mixed methods designs mentioned before.
72

Mixed Methods Research

Transformative designs
Transformative Framework
Quantitative data collection and analysis Follow up with Qualitative data collection and analysis

Interpretation

73

Mixed Methods Research

Transformative Framework Is a framework for advancing the needs of underrepresented or marginalized populations. Such as: Feminist theory, racial or ethnic theory, sexual orientation theory, and disability theory.
74

Mixed Methods Research

Transformative design
All decisions about interaction, priority, timing, and mixing are made within the context of the transformative framework. Researchers can implement any of four basic mixed methods designs within the transformative framework.

75

Mixed Methods Research

Transformative design

76

Mixed Methods Research

Challenges

Little guidance in the literature to assist researchers with implementing mixed methods in a transformative way. Researchers need to have expertise in theoretical foundations of the study.
77

Mixed Methods Research

Transformative design variants
Feminist lens transformative variant Disability lens transformative variant Socioeconomic class lens

78

Mixed Methods Research

Major designs (6). Multiphase: is an another example of a mixed methods design that goes beyond four basic designs. It is a combination of sequential and concurrent aspects. Most common in large funded or multiyear projects.
79

Mixed Methods Research

Multiphase design
Overall program objective
Study 1: Qualitative Informs Study 2: Quantitative Informs Study 3: Mixed Methods

80

Mixed Methods Research

Multiphase design

81

Mixed Methods Research

Challenges
Challenges associated with individual concurrent and sequential designs. Needs sufficient resources, time, and effort. May need a research team to implement research.

82

Mixed Methods Research

Multiphase design variants Large scale program development and evaluation Multilevel statewide study Single mixed methods study that combines both concurrent and sequential phases
83

Mixed Methods Research

Resources
International Congress for Qualitative Inquiry Conference Mixed methods international conference Journal of Mixed Methods Research OBSSR (Office of Behavioral and Social Sciences Research) from NIH : Scientific areas > Methodology > Mixed Methods Research

84

85