ANITA R. SCHMIDT, MPH & PHUNG K. PHAM, MS, MA Friday, March 22, 2019
Disclosure We have no financial or other conflicts of interest to disclose.
OBJECTIVES RECOGNIZE RECOGNIZE the difference between mixed and multi- • methods research DIFFERENTIAT DIFFERENTIATE among the 3 foundational designs, • including sampling and data collection strategies DESIGN DESIGN a study to answer a research question • DESCRIBE DESCRIBE data mixing strategies for reporting •
WHY IS THIS IMPORTANT? Image from: www.pixabay.com
Intuitive way of doing research “Physicians consider quantitative lab results along with a patient’s qualitative life history and symptoms when making a diagnosis and treatment plan.” - Creswell & Plano Clark 2018
OBJECTIVES RECOGNIZE RECOGNIZE the difference between mixed and multi- • methods research
MULTI ≠ MIXED There is a misconception that if you have quantitative data alongside qualitative data, then your study is mixed methods. That is NOT necessarily true.
MULTI ≠ MIXED You have to mix the two in order for your study to qualify as mixed methods research. Otherwise, what you actually have is a multi-methods study.
OBJECTIVES RECOGNIZE RECOGNIZE the difference between mixed and multi- • methods research DIFFERENTIAT DIFFERENTIATE among the 3 foundational designs, • including sampling and data collection strategies
3 FOUNDATIONAL DESIGNS • CONVERGENT CONVERGENT • EXPLANATORY EXPLANATORY Sequential • EXPLORATORY EXPLORATORY Sequential Source: Creswell, Plano Clark 2018 (3rd Edition Book)
CONVERGENT Design
CONVERGENT INTENTIONS ● Obtain nuanced understanding ● Determine convergence (or divergence)
CONVERGENT KEY FEATURES KEY FEATURES ● Separate collections of quantitative & qualitative data ● No precedence for sequence of strands ○ Does NOT matter if you collect quantitative data first, or qualitative data first, or both at around the same time ■ MOST important is independence of the two datasets
CONVERGENT QUANTITATIVE Data Collection & Analysis RESULTS RESULTS INTERPRETATION Merged & Compared QUALITATIVE Data Collection & Analysis Source: Creswell, Plano Clark 2018 (3rd Edition Book)
CONVERGENT INTENTIONS RESEARCH QUESTION ● Obtain nuanced What is the nature of understanding swarming and what are the perceptions about its ● Determine convergence implementation among (or divergence) Emergency Department clinical providers and patient families? (Perniciaro et al)
EXPLANATORY Sequential Design
EXPLANATORY Sequential INTENTIONS ● Use qualitative strand to explain initial quantitative results ● Examine possible mechanisms
EXPLANATORY Sequential KEY FEATURES KEY FEATURES ● Data collection sequence is important ● 2 phases • Quantitative strand first • Qualitative strand second
EXPLANATORY Sequential RESULT RESULTS S QUANTITATIVE QUALITATIVE connected to INTERPRETATION Data Collection Data Collection & & Analysis & Analysis explained by Source: Creswell, Plano Clark 2018 (3rd Edition Book)
EXPLANATORY Sequential INTENTIONS RESEARCH QUESTION ● Use qualitative strand What are the attitudes of to explain initial community pharmacists quantitative results toward nontechnical skills, and how do ● Examine possible community pharmacists mechanisms apply nontechnical skills? (Irwin & Wiedmann 2015)
EXPLORATORY Sequential Design
EXPLORATORY Sequential INTENTIONS ● Generalize qualitative results to larger population ● Test or apply quantitative measure, etc. with attention to context
EXPLORATORY Sequential KEY FEATURES KEY FEATURES ● Data collection sequence is important ● 3 phases • Qualitative strand first • Development or selection of a quantitative measure, etc. based on initial qualitative results • Quantitative strand last
EXPLORATORY Sequential DEVELOP OR SELECT QUALITATIVE QUANTITATIVE INTERPRETATION Measure, Data Collection Data Collection Instrument / & Analysis & Analysis Intervention / Etc. RESULT RESULTS S tested or connected to applied by Source: Creswell, Plano Clark 2018 (3rd Edition Book)
EXPLORATORY Sequential INTENTIONS RESEARCH QUESTION ● Generalize qualitative When working in the results to larger Emergency Department, how do clinical providers’ population emotional state change ● Test or apply (or remain constant) with quantitative measure, various social, behavioral, etc. with attention to and affective dynamics context during their shifts? (Assaf et al)
BREAKOUT #1
SAMPLING & DATA COLLECTION Strategies
Research Study Methods Question Design Sampling ? Data Collection
Convergent Sample Composition DIFFERENT INDIVIDUALS ● Useful when comparing data from individuals at different organizational levels SAME INDIVIDUALS ● Parallel sample ○ Acceptable to sample individuals from same population if unfeasible to have same participants in both strands ● Identical individuals ○ For research questions requiring matched quantitative & qualitative data from same individuals
Explanatory Sequential Sample Composition BEST PRACTICE ● Have individuals in the qualitative strand be a subset of the sample from the quantitative strand
Exploratory Sequential Sample Composition Participants in qualitative strand may be primed to think differently about phenomenon of interest ● Have different individuals in the two strands
SAMPLING Sample size: • Guided by conventions of specific quantitative and qualitative methods Exception for Convergent Design: If useful to merge each individual’s quantitative & qualitative data, use identical participants in both strands
Research Study Methods Question Design Sampling ? Data Collection
DATA COLLECTION 1. When will the data be collected? • Order of operations informed by study design 2. How will the data be collected? • Quantitative & qualitative methods determined by research question and the specific data you hope to elicit
DATA COLLECTION Common Quantitative Strategies: • Surveys • Systematic Observations
DATA COLLECTION Common Qualitative Strategies: • Individual Interviews • Focus Groups
MIXED METHODS Design Examples
CONVERGENT Example: Perniciaro et al What is the nature of swarming and what are the perceptions about its implementation among Emergency Department clinical providers and patient families? SYSTEMATIC OBSERVATION & FAMILY SURVEY DESCRIPTIVES CONVERGENCE & & THEMES DIVERGENCE PROVIDER FOCUS GROUPS
EXPLANATORY Sequential Example: Irvin & Weidmann 2015 What are the attitudes of community pharmacists toward nontechnical skills, and how do community pharmacists apply nontechnical skills? 3 Questionnaires RESULTS RESULT S (Teamwork Individual connected to INTERPRETATION Perceptions, & Interviews Collaboration, explained by Safety Attitudes)
EXPLORATORY Sequential Example: Assaf et al When working in the Emergency Department, how do clinical providers’ emotional state change (or remain constant) with various social, behavioral, and affective dynamics during their shifts? DEVELOP Emergency NOMINAL GROUP EXPERIENCE Interpretation Department TECHNIQUE SAMPLING coming soon! Shift Experience (NGT) METHOD (ESM) Survey THEMES applied by connected to
OBJECTIVES RECOGNIZE RECOGNIZE the difference between mixed and multi- • methods research DIFFERENTIAT DIFFERENTIATE among the 3 foundational designs, • including sampling and data collection strategies DESIGN DESIGN a study to answer a research question •
BREAKOUT #2
SHARE Your Designs
OBJECTIVES RECOGNIZE RECOGNIZE the difference between mixed and multi- • methods research DIFFERENTIAT DIFFERENTIATE among the 3 foundational designs, • including sampling and data collection strategies DESIGN DESIGN a study to answer a research question • DESCRIBE DESCRIBE data mixing strategies for reporting •
DATA MIXING Strategies for Reporting
DATA MIXING Technique #1: COMPARISON & CONTRAST ● Useful in convergent design ● Takes place during interpretation stage , after both data sets have been separately analyzed ● List results from each strand to identify c onvergence & divergence Source: O’Cathain, Murphy, Nicholl 2010 (BMJ)
DATA MIXING Technique #2: FOLLOWING A THREAD ● Useful in explanatory sequential & exploratory sequential designs ● Takes place first during analysis stage after data for the initial strand has been collected ○ Identify which variables or themes to further investigate in the subsequent strand of the study ● Takes place again at interpretation stage to help you make final conclusions Source: O’Cathain, Murphy, Nicholl 2010 (BMJ)
Q&A
Please feel free to contact us! arschmidt@chla.usc.edu ppham@chla.usc.edu Division of Emergency & Transport Medicine Children’s Hospital Los Angeles
Recommend
More recommend