8- Data collection and analysis.ppt

educational

1. Data collection, recruitment &
analysis

2. Data collection in qualitative research Data collection in qualitative research
 Qualitative researchers don’t start with narrowly
specified questions
Starts with general questions and allow
respondents to tell their narratives in a naturalistic
fashion
 The study tools/ instruments are often
unstructured
Mostly contain general and open ended questions
because the phenomenon under study is subjective
or poorly understood

3. Methods Methods
Commonest methods
 In-depth interviews
 Focus groups
 Participant observation
• Others
Photographs, document analysis, videos etc.

4. In-depth interviews In-depth interviews
Qualitative interviews tend to be conversational.
Interviewers encourage respondents to define the
important dimensions of a phenomenon and to
elaborate on what is relevant to them
Unstructured interviews: Used when no
preconceived view of the information to be
gathered exist.
 Researchers begin by asking a grand
tour/general question
Subsequent questions are guided by initial
responses.

5. •Semi-structured (or focused) interviews: used
when researchers have a list of topics or broad
questions that must be covered in an interview.
• Interviewers use a written topic guide to ensure
that all question areas are addressed.
•The interviewer’s function is to encourage
participants to talk freely about all the topics on
the guide.
• Interview guide is applied flexibly with deletion
or additions of questions in the process of dta
collection

6. Use semi-structured interview if the topic
under study is reasonably studied, which
guided identification of initial
questions/topics to be covered in the
interview
 An initial brief review of literature enables
researcher identify issues relevant to the
current study

7. Focus groups [interviews] Focus groups [interviews]
Focus group interviews involve groups of about
5 to 10 people whose opinions and experiences
are solicited simultaneously.
 The interviewer (or moderator) guides the
discussion using a topic guide.
Using is efficient and can generate a lot of
discussions on topic of interest
 The study topic should be appropriate for focus
group e.g. where diversity of opinions is required
Usually structured with topic guide with pre-
identified questions

8. Some considerations during interviews Some considerations during interviews
•Used participant friendly language.
•Questioning and listening skills important
•Attention enables one to ask useful follow-up
questions
• Interview are often recorded rather than taking of
notes to ensure conversation is spontaneous and
nothing is missed out
• Transcription of the interviews is performed
later when audio recording is converted into text
data which is analyzed

9. Participant observation Participant observation
•Aim to understand the behaviors and
experiences of people as they occur in
naturalistic settings.
•Skillful observation permits researchers to
see the world as participants see it
•Participant observation is characterized by
prolonged periods of social interaction
between researchers and participants
• Usually unstructured observation tool used

10. •It could be overt participant observer or convert/
concealed participant observer
• Review question: advantages and
disadvantages of either of these
• Hawthorne effect is a major weakness of overt
participant observation
• Covert observation goes against informed
consent, and privacy principles
• Observers collect data by recording, taking field
notes, photographs etc.
• What to be observed and how should be clear

11. Data collection in quantitative research Data collection in quantitative research
•Data could be original data or uses existing data
from existing records
•Data for quantitative studies tend to be
quantifiable and structured, with the same
information gathered from all participants in a
comparable, prespecified way.
• Quantitative researchers generally strive for
methods that are as objective as possible.
• Several methods for collecting quantitative data
exist

12. Structured interviews Structured interviews
Structured self-report data are collected with a
formal, written document— an instrument.
The instrument is known as an interview schedule
when the questions are asked orally face-to-face
or by telephone or as a questionnaire when
respondents complete the instrument themselves.
 In a totally structured instrument, respondents are
asked to respond to the same questions in the
same order. Closed-ended (or fixed-alternative)
questions are ones in which the response
options are prespecified.

13. •Some structured instruments, however, also
include open-ended questions, which allow
participants to respond to questions in their own
words
•Questions must be carefully worded for clarity,
absence of bias, and (in questionnaires) reading
level of participants

14. Structured observation Structured observation
•Observational methods can be used to gather
such information as patients’ conditions, or
behaviours
•Observations can be made through the human
senses and then recorded manually, but they can
also be done with equipment such as video
recorders
•Structured observation involves the use of formal
instruments and protocols that dictate what to
observe, how long to observe it, and how to
record the data

15. •Structured observation is not intended to capture a
broad slice of life but rather to document specific
behaviors, actions, and events.
• Structured observation requires the formulation
of a system for accurately categorizing, recording,
and encoding the observations
•E.g. Category systems are the basis for
constructing a checklist—the instrument
observers use to record observations

16. Reliability and validity of study instrumentsReliability and validity of study instruments
 Related to quantitative approaches
Reliability: Cconcerns consistency in measuring a
stable attribute for an individual. It involves a
replication to evaluate the extent to which scores for a
stable trait are the same.
Validity : degree to which an instrument is measuring the
construct it purports to measure. When researchers
develop a scale to measure resilience, they need to be
sure that the resulting scores validly reflect this construct
and not something else
 Instruments should be both reliable and valid

17. Trustworthiness strategies Trustworthiness strategies
Related to qualitative approaches.
Transparency: audit trail eg use journals
 Reflexivity: researcher awareness and openness
about how he/she influenced research
 Thick description: derailed description to
promote transferability to other contexts
Peer-debriefing : regular team discussions
 Member checking/participant validation

18. Pilot studies Pilot studies
 Applies to both Qual and Quant studies
 Implementation slightly differs in Qual vs Quant
 No discarding of pilot study data in qual
 Differentiate pretesting instrument and pilot
studies
 Demonstrate how it informed the study
 Results may be reported

19. Review question Review question
 Difference in data collection between qualitative
quantitative research
Examples
 Level of structure of study instruments
 Nature of data
Implementation of the instrument
 Content of the instruments e.g. nature of questions

20. Triangulation Triangulation
Triangulation refers to the use of multiple
referents to draw conclusions about what
constitutes truth.
Overcome the intrinsic bias that comes from
single-method, single-observer, and single-theory
studies
When applied to data collection there are two
types: Data triangulation and method
triangulation

21. Data triangulation involves the use of multiple
data sources for the purpose of validating
conclusions. There are three types of data
triangulation: time, space, and person.
Method triangulation involves using multiple
methods of data collection e.g. in-depth interviews
and focus groups or observation and interviews
 Other triangulation types: investigator, and
theory

22. Recruitment Recruitment
Process of by which participants are first informed
about the study, and requested to join the study if
meets the eligibility criteria
 Should not be confused with sampling
 Methods used depends on sampling strategy,
characteristics of the population, ethical issues,
resources among others

23. Recruitment strategies Recruitment strategies
Media-based e.g. Posters/flyers, radio, TV,
internet
 Investigator-initiated: email, door-to-door,
phone
 Socially-based: chain referral, external referral
 List of participants e.g. panels or HCPs,
patients attending services

24. ConsiderationsConsiderations
Several recruitment challenges and
opportunities might emerge based on some
consideration
•Practical logistics : demand for resources of
time and money
• Population/sample characteristics e.g. are
literacy, dispersal area
• Ethical issues: based on topic sensitivity
• Sample characteristics

25. Qualitative data management & analysis Qualitative data management & analysis
Introduction
•Main difference between qual and quant approaches
• Starts at start of data collection
• Analysis starts with data collection and ends when
results are reported
• Iterative process
• Results used to promote emergent nature of qual
approaches
• Differentiate preliminary & formal analysis

26.  Unlike quant approaches, no conventional
analytic rules in qual approaches, hence
openness on analytic strategy important
 Consider research aims, methods and theoretical
underpinning relative to analysis decisions
 Qualitative research designs such as ethnography,
grounded theory, phenomenology, case studies
have some analytic conventions based on their
philosophical worldviews and theoretical
foundations

27. Preparation for qualitative data analysis Preparation for qualitative data analysis
 Form of data: textual
 Depends on form of analysis
Use computer-Assisted Qualitative Data Analysis
as manual difficult
 Transcription: speech to written text
 Translation : source language to target language
 The population characteristic esp. linguistic
characteristics considered
 Consider linguistic details required

28. Data management Data management
 Researcher interacts with data from data
collection to reporting of findings
Designed structure for filling and categorising
data for easy retrieval
 Helps with data retrieval, security, monitoring
and transparent reporting
 Some strategies: labelling transcripts, information
back up, use password, folders/subfolders and use
of computer software

29. Data analysisData analysis
 No criteria for choosing analytic approach
 Thematic analysis[key themes] commonly used
 Others include content analysis[words, themes,
concepts] and discourse analysis[language]
 Inductive process driven by data
 Moves from descriptive to analytical and abstract
levels
 Starts with labelling of coding of data

30. Common steps Common steps
 Familiarisation with data
 Iteratively developing initial thematic
framework /codebook
 Each code should be operationalized for
consistency in coding
 Start with few transcript to develop codebook and
modify when coding rest as need be
 Review data under each theme and subtheme to
improve coherence in coding

31. Read all data extracts under each theme and
subtheme and note range of elements and
dimensions in the form of perceptions, views,
experiences, emotions, behaviours and action
related to the research questions
 Continuously revisit the themes and subthemes
during the entire reporting of the research results
new ideas keep emerging

32. Dissemination of research findings Dissemination of research findings
 This is done through
 Publication in journals as articles
 Publication in news papers
 Presentations in conferences, seminars

33. Implementation science (research)Implementation science (research)
 It takes between 10-20 years for research
evidence to be implemented into practice and
policy
 There are attempts to shorten the duration
IS attempts to advance knowledge about how to
adopt and integrate evidence-based health
interventions into practice & policy

34. IS defined as scientific study of methods to
promote the systematic uptake of research
findings and other evidence-based practices to
improve the quality of service delivery in
routine care
 Efforts to narrow the gap between intervention
development and its implementation requires
implementation strategies
Implementation strategy is a systematic
intervention process to adopt and integrate
evidence-based health innovations into usual
care

35. Implementation StrategiesImplementation Strategies
 Change of infrastructure: Policies, change
communication, organisation structure
 Financials strategies: resources allocation,
involve policy makers and health administrators
who control resources
 Engage consumers: patients, family members,
community members
 Support clinicians: further education, working
conditions
 Train & train stakeholders for their active
participation

36. Develop stakeholder partnerships such as identify
champions, first adopters, obtain formal
commitments
 Adapt and tailor intervention to context
uniqueness
 Develop monitoring & evaluative strategies