Mixed methods design in research....pptx

Mixed methods

1. Dr Abdirahman Moallim Ibrahim, MBChB, MPH(IS)

2. Outline Introduction/Definition Research designs Qualitative versus quantitative Mixed- methods research Data collection methods for a qualitative study Right interview guide Tips on how to design interview questions Writing the methods section Advantages and challenges Conclusion 2

3. Research Designs Qualitative research is descriptive and concerned with understanding behaviour. It invites people to tell their stories in their own words. Quantitative research concerns measurable quantities and numbers. It involves close- ended questions. Mixed methods research is an approach to inquiry and research that combines quantitative and qualitative methods into one study to provide a broader perspective . It is an emerging research approach in the social and health sciences that combines both statistics and stories. 3

4. Qualitative versus quantitative Criteria Qualitative Research Quantitative Research Purpose To understand and interpret social interactions To test a hypothesis, look for cause and effect and make predictions Group studied Smaller and not randomly selected Larger and randomly selected. Type of data collected Words, images or objects Numbers and statistics Forms of data collected Open- ended responses, interviews, participants’ observations, field notes, and reflections Data based on precise measurement using structured and validated tools Variables Study of the whole, not variables Specific variables studied 4

5. Qualitative versus quantitative (2) Types of analysis Identify patterns, features, and themes. Identify statistical relationships Objectivity and subjectivity Subjectivity is expected Objectivity is critical. Results Particular of specialized findings that is less generalizable. Generalizable findings that can be applied to other population Scientific method Exploratory or bottom- up, the researcher generates a new hypothesis and theory from the data collected. Confirmatory, top- bottom: the researcher test the hypothesis and theory with the data. Final report Narrative with contextual description and direct quotations from research participants Statistical report with correlations, comparison of means, and statistically significant findings 5

6. Quantitative Data Advantages Draws conclusions for a large number of people Efficient data analysis Demonstrates relationships Examines probable cause and effect Bias controlled People like numbers Limitation Impersonal, dry Do not hear the words/voices of the participants Limited understanding of the context of the participants Largely researcher- driven 6

7. Qualitative Data Advantages Detailed perspective of a few people Can hear the voices of the participants Understand participants’ experience within the context Built from the views of participants not researchers People like stories Limitation Limited generalizability Soft data, not as hard as numbers Few people studied Highly interpretative Reliance on participants minimizes researchers' expertise 7

8. Mixed- Methods research Six mixed methods design strategies as developed by Dr. John Creswell (2003), a leading expert in mixed methods research. Sequential Explanatory Design Sequential Exploratory Design Sequential Transformative Concurrent Triangulation Design Concurrent Nested (Embedded) Design Concurrent Transformative Design 8

9. Sequential Explanatory Design This method is a two-phase design where the quantitative data is collected first followed by qualitative data collection. The purpose is to use the qualitative results to further explain and interpret the findings from the quantitative phase. For example, a survey may be used to collect quantitative data from a larger group. Members of that group may then later be selected for interviews where they can explain and offer insights into their survey answers. 9

10. Sequential Exploratory Design This method is also a two- phase design. The qualitative data is collected first, followed by the collection and analysis of quantitative data. The purpose of this design is to develop an instrument (such as a survey), to develop a classification for testing, or to identify variables. Using the information from the qualitative, journals or diaries to develop an appropriate survey to administer to a larger sample is an example of this design. 10

11. Sequential Transformative This type of design also has two phases but allows the theoretical perspective of the researcher to guide the study and determine the order of data collection. The results from both methods are integrated together at the end of the study during the interpretation phase. 11

12. Concurrent Triangulation Design In this design, qualitative and quantitative data are collected concurrently in one phase. The data is analyzed separately and then compared and/or combined. For example, if a researcher collected survey data and interview data at the same time and compared the results. This method is used to confirm, cross-validate, or corroborate findings. It is often used to overcome a weakness in one method with the strengths of another. It can also be useful in expanding quantitative data through the collection of open- ended qualitative data. 12

13. Concurrent Nested (Embedded) Design This design includes one phase of data collection in which priority is given to one approach that guides the project, while the other approach is embedded or nested into the project and provides a supporting role. The embedded approach is often addressing a different question than the primary research question. 13

14. Concurrent Transformative Design This method involves concurrent data collection of both quantitative and qualitative data. It is guided by a theoretical perspective on the purpose or research question of the study. This perspective guides all methodological choices, and the purpose is to evaluate the perspective at different levels of analysis. • 14

15. Data collection methods for qualitative study Interviews – less structured, (in- depth or key informant interview) Focus Groups – less structured. Observation – less structured, requires the involvement of the researcher, smaller data sample, most subjective of data collection methods, researcher bias may be higher Historical or Archival Methods – may not include study participants, may only involve fact- finding by the researcher, may have less bias since more objective data collected 15

16. Right interview guide In-depth guide- In-depth interviews can be with anyone. key informant - Key informants are experts Thus, researchers only use key informant interviews when they can secure a participant with unique knowledge of a topic. Focused group discussion 16

17. Tips on how to design interview questions Make sure they are open- ended questions and are not leading questions. Create questions that are clear and easy to understand. Make questions focused but flexible. Encourage storytelling. Then probe. Design questions that align with your data analysis goals. Plan the questions so that it’s easy to move among questions according to the interview 17

18. Interview Question Tips (2) Questions should be open- ended rather than closed- ended. For example, instead of asking “Do you know about the clinic’s services?” ask “Please describe the clinic’s services.” You should ask a factual question before opinion questions. For example, ask, “What activities were conducted?” before asking, “What did you think of the activities?” Use probes as needed. These include: Would you give me an example? Can you elaborate on that idea? Would you explain that further? I am not sure I understand what you are saying. Is there anything else? 18

19. Methods section Study Design – a qualitative study or Mixed- methods design- in sequential explanatory mixed methods using focused group discussion (FDG) Study Population 40 survey participants, that is four sessions of 10 participants each were considered adequate. 19

20. Sample Size Determination Sample selection for the qualitative study “Data saturation” (the point at which the researcher no longer gets/hears new information) has been recommended for qualitative studies, Having a focused group discussion with a sub- sample of 40 survey participants, that is four sessions of 10 participants each was considered adequate. 20

21. Sampling Technique The FGD participants were purposively selected. Sub- sample of the study participants (male/female, 10- 14/15- 19 years, urban/rural), divided into males and females 10- 14 years and 15- 19 years groups. The focus group sessions were held among separate groups of male and female adolescents 10- 14 and 15- 19 years 21

22. Study Instruments The FGD guide explored barriers to the utilization of reproductive health services. Purpose • Confidentiality • Duration • How the interview will be conducted. Permission to record. The introductory part was on the age categorized into 10- 14 and 15 to 19 years age groups of males and females in both urban and rural communities. The introductory questions were about whom they resided with, their relationship with them if they have heard and can describe the RHS facility, and the source(s) of their information. The transitory question was can you tell me about your visit to a reproductive health service facility in the last 12 months. Some of the main questions are what makes it hard to visit a reproductive health service facility? How long it takes to get there? How long it takes to be attended to and so on. The end questions include the ways their school, services, and location affected their access and utilization of services. Is there anything more you would like to add? 22

23. Data Collection Methods and Analyses Data Collection Methods - The FDG guide was administered to collect. the moderator, timekeeper, and the recorder Use of field notes for each of the sessions. The average time for completion of the focused group discussions was 30-45 minutes It was audio recorded after getting the consent of the participants. Assurances of confidentiality were given; the conversations were heard in private places and no personal identifiers were used. Data Analyses The focus group discussions were recorded, transcribed verbatim, and coded using NVivo 11. The thematic analysis was employed in analysing data and included familiarization through repeated reading, coding (open coding approach), theme generation, applying the codes to the transcripts, and interpretation. Data were coded in NVivo, version 11. Qualitative data analysis software include MAXQDA , Dedoose N v ivo QDA Miner 23

24. Advantages of mixed- methods approach Both have weaknesses. Quantitative research is weak in understanding the context or setting in which data is collected. Qualitative research may include biases and does not lend itself to statistical analysis and generalization. • Mixed method strategies can offset these weaknesses by allowing for both exploration and analysis in the same study. Researchers can use all the tools available to them and collect more comprehensive data. This provides results that have a broader perspective of the overall issue or research problem. The results may include both observations and statistical analyses. Therefore, the results are validated within the study. Using both approaches in one study provides additional evidence and support for the findings. Mixed methods combine inductive and deductive thinking and reasoning. The researcher can use both words and numbers to communicate the results and findings and thus, appeal to a wider audience. Combining methodologies helps to reduce the personal biases of the researcher. 24

25. Challenges It is more time- consuming to collect both quantitative and qualitative data. The study may require more resources to collect both types of data. The research procedures are more complicated and may be out of the comfort zone of the researcher. Investigators are often trained in quantitative or qualitative methods and may need assistance crossing over. Methodology requires a clear presentation when published or presented so that the audience can accurately understand the procedures and the findings. 25

26. Conclusion Mixed- methods research is widely used in various fields, including social sciences, education, healthcare, and psychology, where it helps researchers address complex research questions and gain a richer understanding of human behavior and social phenomena. Learn more: https://cirt.gcu.edu/research/develop/research _ready/mixed_methods/3 26

27. 27