Designing IA for AI - Information Architecture Conference 2024
Qualitative research second copy corrected
1. By Dr Vishal Agrawal&
Dr Menaal Kaushal
JR-1, Department of SPM
SN Medical College, Agra
2. Introduction
Difference Between Qualitative and Quantitative Study
The Concept Of Qualitative Research
Scope Of Qualitative Research
Where To Use
Where Not To Use
As Complementary To Quantitative Study
Steps in Qualitative Research
Types Of Sampling In Qualitative Research
Types Of Data& Forms Of Data
Qualitative Research Methods
Data Analysis
3. Consists of an investigation that:
seeks answers to a question
systematically
uses a predefined
procedures to answer the question
set
of
collects evidence
produces findings that were not determined in
advance
produces findings that may be applicable beyond
the immediate boundaries of the study
5. depends on:
- The type of research question
- The
nature
investigated
of
the
problem
being
6. Includes collecting, analyzing, and interpreting
data by observing what people do and say.
Qualitative research refers to the meanings,
concepts, definitions, characteristics, metaphors,
symbols, and descriptions of things.
It helps to understand the perspectives of the local
population, thus providing the culturally specific
information about opinions, values& behaviors.
(Social Context)
7. The Scope of Qualitative Research is to explore&
understand through
Participant’s ‘Lived experience’
“Insider’s” perspectives
And to understand In the Context
8. Developing and delineating program elements before a
quantitative evaluation
Generating theory
Boosting the power of quantitative design& Broadening the
observation field
Analyzing process and individual cases to explain the how
and why of an outcome
9. Qual
Quant
• Generate hypothesis
• Tests Generalisability
Quant
Qual
• Guides Purposive Sampling
• Helps Interprets Results
Then Test it
10. Seek to explore phenomena
Instruments use more
Seek to confirm hypotheses
about phenomena
flexible, iterative style of
Instruments use more rigid,
eliciting& categorizing responses
inflexible style of eliciting and
to questions
categorizing responses to
questions
Use semi-structured methods
such as interviews, focus
Use highly structured methods
groups, and participant
such as questionnaires,
observation
surveys& structured observation
To describe variation
To quantify variation
To describe& explain relationship To predict causal relationship
To describe individual
experiences& group norms
To describe characteristics of a
population
11. Open- ended
Closed- ended
Textual (obtained from audiotapes, Numerical (obtained by
videotapes, and field notes)
assigning numerical values to
responses)
Some aspects of the study are
Study design is stable from
flexible
beginning to end
Participant responses affect how
and which questions researchers Participant responses do not
influence or determine how and
ask next- Researcher may use
which questions researchers ask
probes
next- Researcher is limited by the
Questionnaire
Study design is iterative, that is,
data collection and research
questions are adjusted according Study design is subject to
statistical assumptions and
to what is learned
conditions
12. Example:
Example:
Open Ended Questions
Close Ended Questions
What are your opinions about Do you give OPV to your
giving OPV to children,
child during PPI rounds?
during PPI rounds?
Yes, Always
No, Never
Probe:
Why do you think that giving
OPV can harm your child?
Sometimes/Occasionally
Most of the times
13. To inform what people are doing, thinking, and saying
about a problem
To identify the important problem to be solved at
community/ local/ policy levels
Generate a list of options for interventions
To
investigate
interventions
how
best
to
implement
promising
To monitor response to interventions and assess how best
to present its results to public and scientific community
14. When numbers are needed to make a decision
(what proportion of people )
Results are to be projected to the total population
(unless generalisability ensured by researcher
through appropriate measures)
15. Sampling
Techniques
Purposive
Sampling
data review and
analysis is done
in conjunction
with data
collection
Chain- Referral
Snow-ball
Sampling
Quota
Sampling
Is a Subtype of Purposive
sampling, when we pre-fix
the number of participants
from every category in the
study
Finds& recruits
“hidden
populations,” not
accessible
through other
sampling
strategies.
16. Unlike quantitative studies, here sample size Is not
determined by conventional formulae, but depends on the
point of Saturation.
Goal is to understand phenomena, not to represent
population, so Select information-rich cases for intensive
study
Minimum samples based on expected reasonable
coverage, given the purpose of the study and constraints
17. 1. Establish the general problem to be investigated
Of interest to the researcher
2. Stating the purpose of the study
Based on problem analysis
Arises from previous studies
Guided by literature review
Determined by who will use the research results
3. Develop a conceptual/theoretical framework for the study
4. Formulate general and specific research questions (aims
and objectives)
5. Select a qualitative research design
18. 6 Select a sampling strategy
Establish site of the research
Selection of participants
7. Ensure trustworthiness of the study
8. Determine data collection methods and develop data
collection tools
9. Establish how data will be managed and analyzed
10. Interpretation and discussion of findings
11. Prepare research report
23. Strong Rapport Building over
long time through prolonged
(>1yr) field work.
Captures field in Holism
Rapport building over short time
through ice- breaking activities
Concept of Engaging in Selection
rather than Holism (see next
slide)
Researcher increasingly
merges with the field& becomes Researcher remains a “Short
time Visitor” or Can use data
“as one of them” i.e. a part of
collectors from the local
the real people‟s lives to
community
understand the community “as
naturally as possible”
Data collection is not iterative but
can be staggered (see next slide)
Data collection is iterative&
over prolonged time periods
Analytical process is Terminal
Analytical process is Iterative
25. Types Of Qualitative
Qualitative Data
Data
Indirect
Semi- Direct
Inferential/
Abstract
Participant‟s
Report
May not be
Accurate
May differ with
Different
Perceptions
Validity of
Representation
Direct
Represents
The Actual
Phenomenon
Very Closely
26. Forms Of Qualitative
Data
Field Notes
Of:
Participant
Observations
Observations
made during
FGDs& InDepth
Interviews
Audio- Video
Recordings
Of:
• Focused
Group
Discussions
• In- Depth
Interviews
Diagrams
E.g.
Sociograms
29. A
qualitative
research
method
in
which
researchers gather data either by observing or by
both observing and participating, to varying
degrees, in the study-community‟s daily activities,
in community settings relevant to the research
questions. Ex at bars, brothels, and health clinic
waiting areas, religious settings.
31. Appearance might indicate membership in groups or in subpopulations
of
interest
to
the
study,
such
as
profession, social status, socioeconomic class, religion or
ethnicity
Verbal behavior& interactions Who speaks to whom and for
how long; who initiates interaction; languages or dialects
spoken; tone of voice, Gender, age, ethnicity, and profession
of speakers; dynamics of interaction
32. Personal space How close people stand to one another. What
individuals‟ preferences concerning personal space suggest about
their relationships.
Human traffic People who enter, leave, and spend time at the
observation site. Where people enter and exit; how long they stay;
who they are (ethnicity, age, gender); whether they are alone or
accompanied; number of people.
People who stand out Identify people who receive a lot of attention
from others. What differentiates them from others; whether people
consult them or they approach other people; whether they seem to
be strangers or well known by others present.
33. Complete observer
Behind one-way mirror, invisible role
Observer as participant
Known, overt observer
Participant as observer
Pseudo-member, research role known
34. It provides Familiarity with the cultural milieu
It can Uncover The Unknown Factors- It Can provide
information previously unknown to researchers that is
crucial for project design, data collection& interpretation of
other data.
But:
It is Time- consuming- at least 1yr in the field, (as in
traditional study pattern)
It is Memory based, so discipline yourself
It is Inherently Subjective so, Practice to be objective
rather than narrating subjectively.
35. Determine :
the purpose of the participant observation activity
as related to the overall research objectives.
the population(s) to be observed.
the venues in which you would like to observe them.
Investigate
possible
sites
for
participant
observation.
Select the site(s), time(s) of day, and date(s)
36. Decide how field staff will divide up or pair off to
cover all sites most effectively.
Consider how you will present yourself, both in
terms of appearance and how you will explain
your purpose to others if necessary.
Plan how and if you will take notes during the
participant observation activity. Remember to take
your field notebook and a pen.
37. Schedule time soon after participant observation
to expand your notes.
Type your notes into computer files using the
standard format set for the study
38.
39. A qualitative research method in which a
researcher/interviewer
gathers data about an
individual‟s perspectives on a specific topic(s)
through a semi-structured exchange with the
individual.
40. The researcher/interviewer engages with the
individual by posing questions in a neutral
manner, listening attentively to responses, and
asking follow-up questions and probes based on
those responses.
41. Elicit feelings
Thoughts
Opinions
Previous experiences
The meaning people give to certain events
43. Experience and behavior questions
Opinion and value questions
Feeling questions
Knowledge questions
Background/demographic questions
44.
45.
46. Focus
group
discussions
are
group
discussions with a small group of individuals
from a well defined target population on preselected topics that rely on interaction
between group members, under the guidance
of a trained facilitator.
Each participant is stimulated by the
comments of others and in turn stimulate
them.
47. It is a qualitative method which helps to find
out the „How‟ „Why‟ of human behaviour
It can provide insight into how a group thinks
about an issue, the range of opinions and
ideas,
and
variations
the
that
inconsistencies
exist
in
a
and
particular
community in terms of beliefs and their
experiences &practices.
48. Get a variety of perspectives/reactions to a
certain issue
In a short time
Mainly for eliciting opinions, values, feelings in
the group- the group norms
49. The topic should is narrowly focused
Selection of participants is also focused by
targeting individuals who meet specific criteria
Topic should be of interest to both the investigator
and respondents.
The emphasis should be on interaction between or
among the group members.
50. Setting the objectives
Determine the target population
Plan the number of of sessions
Follow the guidelines regarding selection of
participants, role of moderator/facilitator etc
Developing F.G.D. guide
Conducting F.G.D.
Analysis and interpretation of result.
51. Cost-effective
Quality of data enhanced by group participants
Can quickly assess the extent to which there is
agreement or diversity on an issue
Enjoyable for participants
52. Outside of natural setting
Silences the minority view
Responses by each participant may be
Constrained
Restricts number of questions that can be asked
Requires group process skills
Confidentiality not assured
Explores major themes, but fails to catch subtle
differences
53. Homogenous
Strangers
6-10 people
Discussion Time: 1- 2 hours
2 FGD per type of respondent
Facilitators: Moderator and note taker
Prepare discussion guide
54. Adequate knowledge on background information about the
topic and experience in conducting FGD
Good listening skills
Leadership skills
Relationship with the participants
Patience and flexibility
Clothing
55. Orient the group in a proper manner.
Put forth issues/ sub issues in appropriate
questions.
Create a non-judgmental environment in which
group members feel free to express.
Encourage interaction between participants.
Encourage quiet participants to speak up and
quieten garrulous talkers.
56. Guide the direction of discussion so that it does
not wander too far from the designated focus.
Pace the discussion appropriate for the participant
Subtly control the time allotted to each question
and to the entire discussion.
57. Primarily an observer, tape record the session.
Observe
the
nature
of
interaction,
record
non-verbal
communication & level of consensus
Should know what type of data she/he is expected to collect.
If facilitator has omitted a question from the guide, the recorder
can point them out (at the end).
Identify the speakers. Note down the first few words every time
a new person speaks and make brief notes of the content.
58. Diagrammatic representation of entire session of
FGD
Offers a useful method of conceptualising group
dynamics drawing comparisons between focus
groups & reflecting on moderating technique
63. Qualitative data analysis is a non-linear/
iterative process
Numerous
rounds
of
questioning, reflecting, rephrasing, analysing, th
eorising,
verifying
observation,
interview,
Discussion
after
or
Focus
each
Group
64. During data collection
Reading – Data Immersion – reading and re-reading
Coding – listen to the data for emerging themes and
begin to attach labels or codes to the texts that
represent the themes
After data collection
Displaying – the Themes (all information)
Developing hypotheses, questioning and verification
Reducing – from the displayed data identify the main
points
65. At all stages – searching for core meanings of
thoughts, feelings, and behaviours described
Overall interpretation
Identify how themes relate to each other
Explain how study questions are answered
Explain what the findings mean beyond the
context of your study
66. I.
Reading / Data immersion
1. Read for content
Are you obtaining the types of information you intended
to collect
Identify
emergent
explanations
themes
and
develop
tentative
Note (new / surprising) topics that need to be explored in
further fieldwork
2. Note the quality of the data
Have you obtained superficial or rich& deep responses
How vivid and detailed are the descriptions of
observations
Is there sufficient contextual detail
67. 3. Develop a system to identify problems in the
data (Audit Trail)
- Read identifying patterns
- After identifying themes, examine how these are
patterned
Do the themes occur in all or some of the data
Is there a relationship between themes
Are there contradictory responses
Are there gaps in understanding – these require
further exploration
68. Problems in the quality of the data require a
review of:
How you are asking questions (neutral or
leading)
The venue
The composition of the groups
The style and characteristics of the
interviewer
How soon after the field activity are notes
recorded
69. II. Coding –
No standard rules of how to code
Emergent
Borrowed
Record coding decisions
Record codes, definitions, and revisions
Usually - insert codes / labels into the margins
Building theme related files
Cut and paste together into one file similarly coded
blocks of text
NB identifiers that help you to identify the original
source
Identify sub-themes and explore them in greater
depth
70. Coding qualitative data does not mean reducing it to
numbers, rather it is a means of indexing your data
While all grounded theory involves coding, not all coding
is grounded theory, again it is often assumed that because
you are coding you are ‘doing’ grounded theory, this is not
always the case
It is a common misconception that computers can code
qualitative data for you, to put it simply, they can’t.
71. III. Displaying data
Capture the variation or richness of each theme
Note differences between individuals and sub-groups
Return to the data and examine evidence that supports
each sub-theme
72. IV. Developing hypotheses, questioning and verification
Extract meaning from the data
Do the categories developed make sense?
What pieces of information contradict my emerging
ideas?
What
pieces
of
information
are
missing
or
underdeveloped?
What other opinions should be taken into account?
How do my own biases influence the data collection
and analysis process?
73. V. Data reduction
i.e. distill the information to make visible the most
essential concepts and relationships
Get an overall sense of the data
Distinguish primary/main and secondary/subthemes
Separate essential from non-essential data
Use visual devices – e.g. matrices, diagrams
74. VI. Interpretation
i.e. identifying the core meaning of the data,
remaining faithful to to the perspectives of
the study participants but with wider social
and theoretical relevance
75. Consistent with data collected
Verified with respondents
Present multiple perspectives (convergent and
divergent views)
Did you go beyond what you expected to find?
76. As
a
way
to
increase
credibility,
but
also
check
transferability, dependability and confirmability triangulation
is often used.
Triangulation
=
a
cross-checking
of
information
and
conclusions in research, brought about by the use of
multiple procedures or sources. If there is agreement
between these, there is support of the interpretation of data.
Using triangulation does not mean you get a certain truth,
but you get closer to it – reflexivity is still necessary.
77. Method triangulation: Comparing data that come from the
use of different methods. These could be both quantitative
and qualitative. Eg. first using a questionaire to ask about
eating habits in a school, and then conduct focus group
interviews afterwards.
Researcher triangulation: Involves using different people
as researchers. This increased the confirmability and
credibility of conclusions. Without this data collection and
conclusions might be affected by researcher bias.
Other triangulation thechniques include data triangulation
and theory triangulation.
78. NATURALISTIC – Natural setting as source of data
INDUCTIVE – It seeks to build theory from data & avoid imposing
researcher‟s own categories of analysis
HOLISTIC – It looks at the phenomenon in totality & takes an overall
perspective
TRIANGULATION – Comparing data collected from different methods
FLEXIBLE DESIGN- Emergent design as opposed to pre-determined
design in quantitative methods
INTERPRETIVE – Aimed at discovering the meaning the events have
for the individuals who experience them & interpretation of these
meaning by researcher
PARTICULARISTIC - Guided by objectives
PROBES - are neutral questions, phrases, sounds, and even gestures
interviewers use to encourage participants to elaborate on their
answers and explain why or how
80. If a Questionnaire developed in a different cultural setting is “imported”
and used, then one must check its validity in one’s setting – Qualitative
methods in the initial phase, work towards avoiding this type III error.
e.g. In U.S. It may be fairly normal to ask a single lady how many
children she has, but this Question would create a havoc if asked by the
researcher in India or Pakistan.
Conversely, in U.S or U.K, it may again be a valid question to ask a
mother of two, whether she is married, but this would prove to be
disastrous if asked to an Indian mother.
Thus, in essence, Type III errors deal with right answers to wrong
questions!