Synthesis on Qualitative Research Methods

Research has been defined by Kerlinger (1970) as the systematic, controlled, empirical
and critical investigation of hypothetical propositions about the presumed relations among
natural phenomena. Meanwhile, educational research is the application of the scientific
approach to the study of educational problems and is the way in which people acquire
dependable and useful information about the educative process to formulate scientific theory.
Educational research, far from being a mechanistic exercise, is a deliberative, complex, subtle,
challenging, thoughtful activity and often a messier process than researchers would like it to be
(Ary et.al, 2006).

The planning and conduct of research is some degree an art, an iterative and often
negotiated process (Cohen, 2010). Ontological assumptions (assumptions about the nature of
reality and the nature of things) give rise to epistemological assumptions (ways of researching
and enquiring into the nature of reality and the nature of things); these, in turn, give rise to
methodological considerations; and these, in turn, give rise to issues of instrumentation and
data collection. Added to ontology and epistemology is axiology (the values and beliefs that we
hold) (Hitchcock and Hughes, 1995).

Educational research is typically classified into two broad categories: quantitative and
qualitative research. Quantitative research uses objective measurement to gather numeric data
that are used to answer questions or test predetermined hypotheses and requires a well-
controlled setting. Qualitative research, in contrast, focuses on understanding social
phenomena from the perspective of the human participants in natural settings. It does not begin
with formal hypotheses, but it may result in hypotheses as the study unfolds. In this paper, we
will focus on qualitative research which is based on a different philosophical approach, which
sees the individual and his or her world as so interconnected that essentially the one has no
existence without the other. It sees social reality as unique; thus, researchers can only
understand human behaviour by focusing on the meanings that events have for the people
involved. Researchers must look not only at what people do but also at how they think and feel,
and you must attempt to understand their reality. The intended result of a qualitative research
study is a narrative report so rich and comprehensive that you can understand the social reality
experienced by the participants (Ary et. al, 2006). Proponents of qualitative inquiry argue that
the quantitative approach to the study of human experience seeks to isolate human behaviour
from its context; it engages in context stripping.
 The research design is the researcher’s plan of how to proceed to gain an understanding
of some group or some phenomenon in its context. Individuals, families, and a variety of groups,
organizations, industries, and more can be examined using qualitative research. The qualitative
researcher begins from a conceptual framework—a “system of concepts, assumptions,
expectations, beliefs, and theories” (Maxwell, 2005) that informs the design. The research
question may be one that comes from the investigator’s observations and experiences with
particular topics, settings, or groups. Qualitative problems examine the context of events, real-
world setting, subjects’ perspectives, unfolding and uncontrolled events, reasons for the events,
and phenomena needing exploration and explanation and that is why the background of the
study is included which tells the genesis of the problem or the area of opportunity you wish to
address.

Moving forwards, the choice of the research question is crucial because the question
determines the design. Particularizing questions ask about a specific context and are less
concerned about generalizing but, rather, focus on developing rich descriptions and
interpretations. Generic questions can be used in qualitative research, such as with multisite
studies, but must be used with caution. Process questions examine how things happen—the
process by which a phenomenon takes place. Questions asking about meaning, influences, and
context are process oriented. Realist questions treat unobserved phenomena (feelings, beliefs,
intentions, etc.) as real and are common is qualitative studies. The following are some practical
suggestions for choosing a qualitative research problem (Bogdan & Biklen, 2006). Choose a
problem that (1) interests, (2) significant because it will contribute to the body of knowledge or to
solving some educational problem, (3) reasonable in size and complexity so that you can
complete a study of it within the time and with the resources to which you will have access, and,
if possible, (4) one in which you are not directly involved. It is more difficult, though not
impossible, to assume the role of researcher in contexts in which you are directly involved.

The formulation of research problems should have significance that will need to convince
the reader that your proposed study is important. Tell what can be learned as a result of the
study and why that is worth knowing. State the implications of the findings for educational
practice and/or theory. Although, remember that the findings may be unique to the single group
or specific context of your study and may not necessarily transfer to other contexts or groups.

Another content for the first chapter are: a.) Assumptions - statements of what the
researcher believes to be facts but cannot verify; b.) limitations - those conditions beyond the
control of the researcher that may place restrictions on the conclusions of the study and their
application to other situations. Administrative policies that preclude using more than one class in
an experiment, a data-gathering instrument that has not been validated, or the inability to
randomly select and assign subjects to experimental and control groups are examples of
limitations. c.) Delimitations are the boundaries of the study. A study of attitudes toward racial
minorities may be concerned only with middle-class, fifth-grade pupils, and conclusions are not
to be extended beyond this population sampled; d.) Definition of terms – clarifications of the
main terms in context with the study.

On supporting the study with established various sources, the review of the literature
helps identify what is already known, how it relates to your question, how your study might
contribute to greater understanding of the topic, and the potential theoretical frameworks that
might inform the study. Reviewing the literature helps prevent duplication of research and helps
avoid making mistakes in research design that others have made. In a qualitative proposal, it is
not assumed that the literature review is complete in the proposal. Reviewing literature in a
qualitative study is an ongoing process and involves an interactive process throughout the
study. As the study evolves, new literature may need to be explored. This literature review stage
serves several important functions: 1.) Knowledge of related research enables investigators to
define the frontiers of their field; 2.) A thorough review of related theory determining whether
your endeavours are likely to add to knowledge in a meaningful way. Studies with no link to the
existing knowledge seldom make significant contributions to the field; 3.) Reviewing related
literature helps researchers to limit their research question and to clarify and define the
concepts of the study - researchers become familiar with previous efforts to clarify these
constructs and to define them operationally; 4.) Through studying related research, investigators
learn which methodologies have proven useful and which seem less promising.

Also, it is appropriate here to formulate a major hypothesis and possibly several minor
hypotheses. This approach further clarifies the nature of the problem and the logic underlying
the investigation and gives direction to the data-gathering process. A good hypothesis has
several basic characteristics: 1.) It should be reasonable; 2.) It should be consistent with known
facts or theories; 3.) It should be stated in such a way that it can be tested and found to be
probably true or probably false; 4.) It should be stated in the simplest possible terms. The
research hypothesis is a tentative answer to a question. The gathering of data and the logical
analysis of data relationships provide a method of confirming or disconfirming the hypothesis by
deducing its consequences. It is important that the hypothesis be formulated before data are
gathered.
 Moulded by qualitative paradigm, qualitative research is classified into types. First, the
narrative research has its roots in different humanities disciplines and focuses on stories
(spoken or written) told by individuals about their lives. The researcher emphasizes sequence
and chronology and a collaborative re-storying process. The key question is, “What is the story
and its meaning as told through this person’s experience?” The researcher seeks to understand
the lived experience of an individual or small group. Narrative research evolved from diverse
disciplines in the humanities, such as literature, art, and film, as well as from disciplines such as
psychology and sociology. The narrative can be a first-person account or a collaboratively
constructed narrative and use documents and other sources of personal information to
illuminate a person’s life.

Narrative research has been claimed to be the best qualitative approach for capturing
detailed stories of life experiences of either single individuals or small. Information is collected
about the context of stories in order to situate the individual stories within personal experiences
(home, job, and school), within culture (ethnic and racial), and within history (time and place).
The researcher then analyses and reorganizes the data and shapes them into a framework with
a chronological sequence with beginning, middle, and end.

Another type is the phenomenological study - designed to describe and interpret an

experience by determining the meaning of the experience as perceived by the people who have
participated in it. What is the experience of an activity or concept from the perspective of
particular participants? - the key question in phenomenology. Rooted in philosophy and
psychology, the assumption is that there are many ways of interpreting the same experience
and that the meaning of the experience to each person is what constitutes reality and the
subjective experience is at the center of the inquiry unlike other types. Phenomenology moves
from individual experience to a universal essence and always asks what is the nature or
meaning of something. Interviewing multiple individuals is the typical data collection approach.
Phenomenological studies begin with the assumption that multiple realities are rooted in
subjects’ perspectives. Thus, an experience has different meanings for each person.

Phenomenological researchers also typically explore their own experiences related to

the topic of interest prior to conducting interviews in order to examine their own biases and
assumptions. The concept of bracketing is used in phenomenological research. Bracketing
involves the researcher intentionally setting aside his or her own experiences, suspending his or
her own beliefs in order to take a fresh perspective based on data collected from persons who
have experienced the phenomenon. The bracketing or suspension of belief is also referred to as
epoche. From an analysis of the interview data, the researcher writes descriptions of the
participants’ experiences and how those experiences were perceived. From the analysis, the
researcher derives an overall description of the general meaning of the experience. This is done
through a process called reduction. It is a phenomenological device that aims to bring aspects
of meaning into nearness or focus.

Third, the ethnographical studies - Ethnography is the in-depth study of naturally

occurring behaviour within a culture or entire social group. It seeks to understand the
relationship between culture and behaviour, with culture referring to the shared beliefs, values,
concepts, practices, and attitudes of a specific group of people. What are the meanings of these
human actions and interactions within this context? Ethnographers typically describe, analyse,
and interpret culture over time using observations and fieldwork as the primary data collecting
strategies. The final product is a cultural portrait that incorporates the views of participants (emic
perspective) as well as views of researcher (etic perspective). Anthropologists immerse
themselves in the lives of the people they study, using primarily extended observation
(participant and nonparticipant) and occasionally in-depth interviewing to gain clarification and
more detailed information. The ethnographer undertakes the study without any a prior
hypotheses to avoid predetermining what is observed or what information is elicited from
informants. The success of a researcher in conducting ethnography depends on the skill with
which the researcher builds rapport and exhibits appropriate behaviours in the setting. The
researcher must build a level of acceptance and trust with the informants.

Creswell (2007) describes two approaches to ethnography. Realist ethnography is the

more traditional approach. In realist ethnography, the researcher tries to provide an objective
account of the situation, typically from a third-person point of view. Standard categories are
used, and factual information and closely edited quotes are presented as data. The researcher’s
interpretation occurs at the end. In critical ethnography, the researcher takes an advocacy
perspective and has a value-laden orientation. The researcher is advocating for a marginalized
group, challenging the status quo, or attempting to empower the group by giving it voice.

Meanwhile, a case study focuses on a single unit to produce an in-depth description that
is rich and holistic. The underlying question is “What are the characteristics of this particular
entity, phenomenon, person, or setting?” Case studies typically include multiple sources of data
collected over time. As indicated, case studies provide an in-depth description of a single unit.
The “unit” can be an individual, a group, a site, a class, a policy, a program, a process, an
institution, or a community. It is a single occurrence of something that the researcher is
interested in examining. Case studies can answer descriptive questions or attempt to explain
why something happened by looking at a process. They are particularistic (focused on a
particular phenomenon, situation, or event), descriptive (providing as an end result a thick rich
description), and heuristic (focused on providing new insights). The unit is defined within specific
boundaries, referred to as a “bounded system.” To be bounded, the phenomenon must be
identifiable within a specific context. If it cannot be described in such a way, case study may not
be the best approach to study it. In some ways, case study is not so much about how the
phenomenon is studied but, rather, about the choice of what to study. In an instrumental case
study, the researcher selects the case because it represents some other issue under
investigation and the researcher believes this particular case can help provide insights or help to
understand that issue. The multiple or collective case study uses several cases selected to
further understand and investigate a phenomenon, population, or general condition.
Researchers conducting case studies provide a detailed report that may build on narratives,
vignettes, tables, charts, figures, visual displays, text, and more. Typically, the report is written
to provide both an emic, or insider, perspective (the perspective of the individuals who are part
of the case) as well as an etic, or outsider, perspective (the interpretations of the researcher).

Next is the grounded theory that has its roots in sociology. Its goal is to inductively build
a theory about a practice or phenomenon using interviews and observation as the primary data
collection tools. Glaser and Strauss (1967) developed grounded theory as a way of formalizing
the operations needed to develop theory from empirical data. This research approach focuses
on gathering data about peoples’ experiences in a particular context and then inductively
building a theory “from the bottom up.” The theory is grounded in the data and “conceptually
dense”; that is, it presents many conceptual relationships that are stated as propositions
pertaining to a particular context, situation, or experience. Grounded theory moves beyond
description to generate or discover a theory that emerges from the data and that provides an
explanation of a process, action, or interaction.

The study may include as many as 20 to 25 subjects who are interviewed on the topic
until no new information is forthcoming (data saturation). To confirm or refute the theory that has
developed, researchers sometimes interview another group that has had different experiences.
Data are analysed by searching for similarities and differences among the participants’
responses about the experience. The constant comparative method of analysis is typically used
in grounded theory. In this method, the researcher compares units of data with each other to
generate tentative categories, eventually reducing these to conceptual categories that evolve
into an overall framework or theory. Open coding is used to develop major or core categories
with axial coding to develop categories around the core. A visual model is developed called an
axial coding paradigm. Selective coding is then used to develop propositions or hypotheses
based on the model, showing how the categories are related. The resulting theory can take the
form of a narrative statement, a picture, or a series of hypotheses.

Last is the content or document analysis is a research method applied to written or

visual materials for the purpose of identifying specified characteristics of the material. The
materials analyzed can be textbooks, newspapers, web pages, speeches, television programs,
advertisements, musical compositions, or any of a host of other types of documents. Content
analysis is widely used in education. The following are some of the purposes of content analysis
in educational research: 1.) To identify bias, prejudice, or propaganda in textbooks; 2.) To
analyse types of errors in students’ writings; 3.) To describe prevailing practices; 4.) To discover
the level of difficulty of material in textbooks or other publications; 5.) To discover the relative
importance of, or interest in, certain topics. The preceding study illustrates the steps involved in
a content analysis: 1.) Specifying the phenomenon to be investigated; 2.) Selecting the media
from which the observations are to be made; 3.) Formulating exhaustive and mutually exclusive
coding categories so that the verbal or symbolic content can be counted; 4.) Deciding on the
sampling plan to be used in order to obtain a representative sample of the documents; 5.)
Training the coders so that they can consistently apply the coding scheme that has been
developed and thus contribute to the reliability of the content analysis. If the inter-rater reliability
is satisfactory, you can proceed to the next step. If it is less than satisfactory, the coders may be
useful in revising the category definitions to make them clearer and more complete and; 6.)
Analyzing the data, which may involve just the frequencies and percentages in the various
categories or may involve more descriptive accounts.

In gathering data for all types, we need sampling, or selection of participants or sites, is
important in qualitative research just as it is in quantitative research - typically it is not a random
sample. Qualitative researchers are purposeful in selecting participants and settings. They
select purposive samples believed to be sufficient to provide maximum insight and
understanding of what they are studying.

Purposive samples are typically small. There is no general rule about the number of
participants to include in a qualitative study but practical considerations such as time, money,
and availability of participants influence the size of the sample. However, the primary criterion of
sample size is redundancy of information. Sampling should be terminated when no new
information is forthcoming from new units. A unit is an individual participant, group, organization,
event, setting, document, or artifact selected as part of the qualitative study. This point is
referred to as data saturation. Several variations on purposive sampling are used in qualitative
research. These strategies can be combined or mixed within studies as well to meet multiple
needs or to provide triangulation: 1.) Comprehensive sampling - every unit is included in the
sample and is used when the number of units is small; 2.) Critical case sampling involves the
selection of a single unit that provides a crucial test of a theory or program; 3.) Maximum
variation sampling where units are included that maximize differences on specified
characteristics and reveals differences but may also identify commonalities across the units; 4.)
Extreme, deviant, or unique case sampling - units that are atypical, special, or unusual. Such a
study might identify practices, teaching methods, and student characteristics that may be
relevant to their superior performance; 5.) Typical case sampling selects units that are
considered typical of the phenomenon to be studied. This approach highlights what is normal or
average; 6.) Negative or discrepant case sampling selects units that are examples of exceptions
to expectations - confirming and disconfirming sampling; 7.) Homogeneous sampling selects a
subgroup that is considered homogeneous in attitudes, experiences, and so on. This approach
may be used with focus group interviewing; 8.) Snowball, chain, or network sampling occurs
when the initially selected subjects suggest the names of others who would be appropriate for
the sample. These next subjects might then suggest others and so on. Such sampling occurs
when potential respondents are not centrally located but scattered in different sites; 9.) Intensity
sampling involves selecting participants who exhibit different levels of the phenomenon of
interest to the researcher; 10.) Stratified purposeful sampling attempts to ensure that subgroups
are represented so that comparisons can be facilitated; 11.) Random purposeful sampling -
when the potential purposeful sample is too large, the credibility of the study can be enhanced
by randomly selecting participants or sites from the larger group; 12.) Theoretical or theory-
based sampling - the researcher begins by selecting a person or site that exemplifies the
theoretical construct and continues to select new cases that reflect the developing theory to
include as the research unfolds and the theory emerges; 13.) Criterion sampling - the
researcher sets the criterion and includes all cases that meet that criterion; 14.) Opportunistic
sampling takes advantage of new leads or unexpected opportunities; 15.) Convenience
sampling is choosing a sample based on availability, time, location, or ease of access but is not
recommended because it may produce evidence that is not credible.

One of the distinguishing characteristics of qualitative research is the methods used to

collect and analyse data. In qualitative studies, the human investigator is the primary instrument
for the gathering and analysing of data. Because qualitative research studies human
experiences and situations, researchers need an instrument flexible enough to capture the
complexity of the human experience, an instrument capable of adapting and responding to the
environment. It is believed that only a human instrument is capable of this task - talks with
people in the setting, observes their activities, reads their documents and written records, and
records this information in field notes and journals. Qualitative inquiry relies on fieldwork
methods (interviewing, observation, and document analysis) as the principal means of collecting
data, avoiding the use of paper-and-pencil tests, checklists, mechanical instruments, and highly
structured observational protocols.

An instrument commonly used in qualitative research is the interview. Interviews are

used to gather data from people about opinions, beliefs, and feelings about situations in their
own words. They are used to help understand the experiences people have and the meaning
they make of them rather than to test hypotheses. Interviews may provide information that
cannot be obtained through observation, or they can be used to verify observations. The
qualitative interview is typically more probing and open ended and less structured than the
interview used in quantitative research but varies considerably in the way it is conducted. At one
extreme is the unstructured interview, which is a conversational type of interview in which the
questions arise from the situation. It is sometimes described as “a conversation with a purpose.”
The interview is not planned in detail ahead of time; the researcher asks questions as the
opportunity arises and then listens closely and uses the subjects’ responses to decide on the
next question. The subjects in the setting may not even realize they are being interviewed.
Using the who, what, when, where, why, and how categories is generally enough guidance for
the researcher to follow in asking questions. On the other hand, the more structured interview,
scheduled for the specific purpose of getting certain information from the subjects. Each
respondent is asked the same set of questions, but with some latitude in the sequence. In the
qualitative approach, the list of questions is generally more limited in length and most questions
cannot be answered with yes or no or limited word responses. In between the unstructured and
structured interview is the semi- or partially structured interview, in which the area of interest is
chosen and questions are formulated but the interviewer may modify the format or questions
during the interview process. Questions are typically open and the questions are designed to
reveal what is important to understand about the phenomenon under study. Qualitative
interviews might involve one-time interviews with a subject or subjects, multiple interviews with
the same subject or subjects, or group interviews or focus groups.
 A focus group, which is like a group interview, typically centers on a particular issue; the
trained interviewer elicits the views of the group members while noting interactions within the
group. Listening to others helps people form their own opinions. The researcher gains insight
into how the participants are thinking and why they are thinking as they do. The respondents are
free to answer in their own words and can answer either briefly or at length. The questions
asked may even vary from individual to individual. The responses are recorded by taking notes,
either during the interview or immediately afterward, or with an audiotape. Focus groups are
more socially oriented than individual interviews and can increase the sample size in the study,
but they allow less control than individual interviews and data can be more difficult to analyse.
Focus groups should not be used in emotionally charged environments or in environments with
unbalanced power dynamics. There are other approaches as well that are related to particular
types of qualitative research: ethnographic interviewing, grounded in anthropology, attempts to
understand the participants’ worldviews through gathering cultural knowledge and includes
descriptive questions, structural questions, and contrast questions; phenomenological
interviewing, grounded in philosophy, attempts to examine lived experience through three in-
depth interviews, with the first focused on the past, the second on the present, and the third on
connecting the previous two to understand the “essential” experience; elite interviewing selects
individuals based on their expertise—those who are considered particularly influential or well
informed. In another light, the number of types of interview is frequently a function of the
sources one reads. For example: (i) standardized interviews; (ii) in-depth interviews; (iii)
ethnographic interviews; (iv) elite interviews; (v) life history interviews; (vi) focus groups; (vii)
semi-structured interviews; (viii) group interviews; (ix) structured interviews; (x) exploratory
interviews; (xi) informal conversational interviews; (xii) interview guide approaches; (xiii)
standardized open-ended interviews; (xiv) closed quantitative interviews.

Next is the direct observation of behaviour - the most desirable measurement method.
When observations are made, the product of those observations is notes or narratives, the
research is qualitative. The researcher identifies the behaviour of interest and devises a
systematic procedure for identifying, categorizing, and recording the behaviour in either a
natural or a contrived situation. The behaviours observed in quantitative studies may be
categorized as high inference and low inference. High-inference behaviours such as teacher
warmth or creativity require more judgment on the part of the observer. Low-inference
behaviours require less judgment by the observer. Examples of low-inference behaviours
include teachers’ asking questions, praising students, or accepting students’ ideas. In
educational research, one of the most common uses of direct observation is in studying
classroom behaviour. 1. Select the aspect of behaviour to be observed; 2. Clearly define the
behaviours falling within a chosen category. Know what behaviours would be indicators of the
attribute; 3. Develop a system for quantifying observations e.g. divide the observation period
into brief time segments and to record for each period; and 4. Develop specific procedures for
recording the behaviour. Record the observations immediately after they are made because
observers’ memory is not sufficiently reliable for accurate research. The best solution is a
coding system that allows the immediate recording of what is observed, using a single letter or
digit.

A highly structured observation will know in advance what it is looking for and will have
its observation categories worked out in advance. A semi-structured observation will have an
agenda of issues but will gather data to illuminate these issues in a far less predetermined or
systematic manner. An unstructured observation will be far less clear on what it is looking for
and will therefore have to go into a situation and observe what is taking place before deciding
on its significance for the research. In a nutshell, a structured observation will already have its
hypotheses decided in advance and will use the observational data to confirm or refute these
hypotheses. On the other hand, a semi-structured and, more particularly, an unstructured
observation will be hypothesis-generating rather than hypothesis-testing. There is a well-known
classification of researcher roles in observation, which lie on a continuum: the complete
participant: a member of the group who conceals her/his role as an observer, whose knowledge
of the group/situation may be intimate and who may gain ‘insider knowledge’, but who may be
viewed with suspicion or resentment by the other members when his/her true role comes to light
and who may lack the necessary objectivity to observe reliably; the participant-as-observer: a
member of the group who reveals her/his role as an observer, whose knowledge of the
group/situation may be intimate and who may gain ‘insider knowledge’, but who may lack the
necessary objectivity to observe reliably and with whom confidences and confidential data may
not be shared or given respectively; the observer-as-participant: not a member of the group, but
who may participate a little or peripherally in the group’s activities, and whose role as researcher
is clear and overt, as unobtrusive as possible, without those being observed always knowing
who is the researcher, and whose access to information and people may be incomplete or
restricted; the complete observer: who only observes (overt or covert) and is detached from the
group, for example, an outside observer, or where the observer is not covert but whose
presence is unnoticed by the group, for example, an observer in a crowded location.
 Researchers use checklists, rating scales, and coding sheets to record the data
collected in direct observation. Checklists, the simplest device used is a checklist, presents a list
of the behaviours that are to be observed. The observer then checks whether each behaviour is
present or absent. The behaviours in a checklist should be operationally defined and readily
observable. Rating scales, are often used by observers to indicate their evaluation of an
observed behaviour or activity. Typically, rating scales consist of three to five points or
categories e.g. point scale of 3 (very well prepared), 2 (prepared), or 1 (not well prepared).
Scales with more than five rating categories are not recommended because it is too difficult to
accurately discriminate among the categories. Coding Systems are used in observational
studies to facilitate the categorizing and counting of specific, predetermined behaviours as they
occur. Two kinds of coding systems are typically used by researchers: sign coding and time
coding. Sign coding uses a set of behaviour categories; each time one of the behaviours occurs,
the observer codes the happening in the appropriate category. In the second type of coding,
called time coding, the observer identifies and records all predetermined behaviour categories
that occur during a given time period. The time period might be 10 seconds, 5 minutes, or some
other period of time.

Another is, questionnaires - offer benefits of standardized and open responses to a

range of topics from a large sample or population. They can be cheap, reliable, valid, quick and
easy to complete. The questionnaire is a widely used and useful instrument for collecting survey
information, providing structured, often numerical data, able to be administered without the
presence of the researcher and often comparatively straightforward to analyse. There are
several kinds of question and response modes in questionnaires, including: dichotomous
questions; multiple-choice questions; rating scales; constant sum questions; ratio data; and
open-ended questions. It is essential that question types are fit for purpose (Champagne, 2014),
being suitably focused and concrete, yielding useable and relevant data, measuring what they
are intended to measure and avoiding questions to which the researcher already knows the
answer.

The open-ended question is an attractive device for smaller-scale research or for those
sections of a questionnaire that invite an honest, personal comment from respondents in
addition to ticking numbers and boxes. Here, the questionnaire puts the open-ended questions
and leaves a space (or draws lines) for a free response. Further, it puts the responsibility for,
and ownership of, the data much more firmly into respondents’ hands.
 The first stage in analysing qualitative data collected through any of the instruments
involves familiarization and organization so that the data can be easily retrieved. Initially, the
researcher should become familiar with the data through reading and rereading notes and
transcripts, viewing and reviewing videotapes, and listening repeatedly to audiotapes. Field
notes, audiotapes, videotapes, observer comments, and other data must be put into a form
ready for analysis. Preferably, transcriptions should be made of all data, including tape-recorded
interviews, focus groups, video recordings, and handwritten field notes. Words should be
transcribed directly to avoid potential bias in selection or interpretation that may occur with
summarizing.

After familiarizing yourself with the data and organizing them for easy retrieval, you can
begin the coding and reducing process. This is the core of qualitative analysis and includes the
identification of categories and themes and their refinement. Coding is about developing
concepts from the raw data. The first step in coding is referred to as axial coding, open coding,
preliminary coding, or provisional coding. The most common approach is to read and reread all
the data and sort them by looking for units of meaning—words, phrases, sentences, subjects’
ways of thinking, behaviour patterns, and events that seem to appear regularly and that seem
important. Each unit of meaning label should be understandable without any additional
information. When coding initially, use as many codes as needed. These will be reduced later.
The initial coding leads to the development of tentative categories. These may be refined and
reconceptualised as the analysis process continues. Once coding of a transcript is completed
and all items with a particular code are placed together, review the sets of items to ensure they
belong together.

Once categories have been established, consider whether some categories may be
linked to create major categories or themes. This process of coding, categorizing, and
developing themes will be repeated for each transcript or set of data. Then merge these sets
together, reviewing categories and themes. The categories developed from the coded data
should be internally consistent and distinct from one another. Categories can be: substantive
categories - primarily descriptive and not generally related to more abstract theories; emic
substantive categories are those from participants’ perspectives and words. However,
substantive categories are more likely to be based on the researcher’s interpretation of what is
going on (etic categories). Theoretical categories are more abstract and can be from prior
theory or from inductively developed theory. They are more likely to be etic categories. The
number of categories developed will depend on the type of data collected, the focus of inquiry,
and the researcher’s analytic skills. If the number of categories is very large, they should be
collapsed into a manageable number. Generating themes is the most difficult and challenging
process of analysis. Perhaps the best known qualitative analysis strategy is the constant
comparative method, which combines inductive category coding with simultaneous comparison
of all units of meaning obtained. The researcher examines each new unit of meaning (topics or
concepts) to determine its distinctive characteristics. Then he or she compares categories and
groups them with similar categories. If there are no similar units of meaning, form a new
category. Another approach used in analysis is the negative case analysis or discrepant data
analysis. Negative cases contradict the main category or pattern, and discrepant cases provide
a different perspective on a category or pattern. This approach provides a counterbalance to a
researcher’s tendency to hold on to first impressions or hunches.

Once data have been completely analysed and themes developed, the next step is
interpretation, going beyond the descriptive data to extract meaning and insights from the data.
Researchers state what was found that was important, why it is important, and what can be
learned from it. Interpreting involves reflecting about the words and acts of the study’s
participants and abstracting important understandings from them. It is an inductive process in
which you make generalizations based on the connections and common aspects among the
categories and patterns. You may develop hypotheses that have evolved during the analysis.
Interpretation is about bringing out the meaning, telling the story, providing an explanation, and
developing plausible explanations. Interpreting qualitative data is difficult because there are no
set rules to follow. The quality of the interpretation depends on the background, perspective,
knowledge, and theoretical orientation of the researcher and the intellectual skills he or she
brings to the task. The interpretation cannot be just a figment of your imagination but must be
supported by the data. The approach to interpretation is clearly affected by the particular
qualitative approach used in the study. In narrative inquiry, the researcher is describing the
larger meaning of the story. In phenomenology, the researcher is attempting to examine the
“essence” of something using textual and structural descriptions. In grounded theory,
interpretation involves developing propositions. Ethnographers are trying to make sense of how
a culture works. A case study develops generalizations. Each of these qualitative approaches
evolved from differing disciplinary bases that clearly influence how the data are interpreted.

In regards with the final results, representation involves how the data are presented -
graphs, pictures, diagrams, figures, or frameworks. The procedure used most frequently by
qualitative researchers is to report by themes, topics, or cases and demonstrate these through
descriptive detail. Some findings are best represented visually, and newer technologies are
providing alternatives for how qualitative research is shared. Performance ethnography and
some other approaches often present findings in alternative formats—for example, using
multimedia presentations or theatre. Generally, ethnographies have narrative augmented by
tables, figures, and sketches. Case studies present a detailed view of a case using narrative,
tables, and figures. Grounded theory usually concludes with a visual model of a theory or a set
of propositions. Phenomenology provides a narration related to the essence with discussion of
meaning; it describes what the participants experienced and how they experienced it. Narrative
inquiries result in life stories or similar presentations and are likely to include such elements as
plot, characters, setting, and interaction and often are written chronologically.

To put security on the research study, validity cannot be taken for granted. It is relative—
related to the purpose and circumstances of the research. The integrity of qualitative research
depends on attending to the issue of validity. Validity concerns the accuracy or truthfulness of
the findings. Eisner (1998) defines structural corroboration as a “means through which multiple
types of data are related to each other to support or contradict the interpretation and evaluation
of a state of affairs”. The use of multiple sources of data, multiple observers, and/or multiple
methods is referred to as triangulation. Structural corroboration uses different sources of data
(data triangulation) and different methods (methods triangulation). A combination of data
sources, such as interviews, observations, and relevant documents, and the use of different
methods increase the likelihood that the phenomenon under study is being understood from
various points of view. In data triangulation, the researcher investigates whether the data
collected with one procedure or instrument confirm data collected using a different procedure or
instrument. The researcher wants to find support for the observations and conclusions in more
than one data source. Convergence of a major theme or pattern in the data from these various
sources lends credibility to the findings. Methods triangulation uses more than one method in
the study. The assumption is that the combination of methods results in better evidence. When
these different procedures or different data sources are in agreement, there is corroboration

Validity is primarily demonstrated through two methods: peer review and investigator
triangulation. In peer review, also called peer debriefing colleagues or peers are provided with
the raw data along with the researcher’s interpretation or explanation. Discussions then
determine whether the reviewer(s) considers the interpretation to be reasonable, given the
evidence. Investigator triangulation involves having multiple researchers collect data
independently and compare the collected data. The key questions here are “Did what was
reported as taking place actually happen? Did the researcher accurately report what was seen
and heard?”

Referential or interpretive evidence of validity refers to “accurately portraying the

meaning attached by participants to what is being studied by the researcher” and “the degree to
which the participants’ viewpoints, thoughts, feelings, intentions, and experiences are accurately
understood and portrayed”. Two primary strategies are used to enhance referential adequacy:
member checks and low-inference descriptors. Theoretical adequacy or plausibility concerns
the degree to which a theoretical explanation developed from the study fits the data and is
defensible. There are three key strategies for promoting theoretical adequacy: extended
fieldwork, theory triangulation, and pattern matching. Extended fieldwork provides time for the
researcher to observe a full range of activities in order to identify patterns and relationships and
their typicality, which will contribute to valid interpretations. An extended time in the research
setting enables the researcher to gain the participants’ trust and thus to obtain more detailed
and honest responses. Theory triangulation involves consideration of how the phenomenon
under study might be explained by multiple theories. Considering different theories, the
researcher may gain better insights. Interdisciplinary triangulation uses other disciplines to infer
processes and understandings of the findings. Pattern matching involves making predictions
based on theory and testing those predictions. Based on theory, the researcher predicts a
certain pattern of results. If the predictions occur, the researcher has strong evidence to support
his or her explanations.

Researcher bias is a source of invalidity in qualitative studies. Bias may result from
selective observations, hearing only what one wants to hear, or allowing personal attitudes,
preferences, and feelings to affect interpretation of data. The most common strategy to control
for bias in qualitative studies is reflexivity. Reflexivity is the use of self-reflection to recognize
one’s own biases and to actively seek them out. The researcher should refer to his or her
journal reflections during the process of data analysis. Another strategy used to control for bias
is negative case sampling, in which researchers intentionally seek examples of the opposite of
what they expect.

Another is transferability - the degree to which the findings of a qualitative study can be
applied or generalized to other contexts or to other groups. In quantitative research, the term
external validity is used to refer to the generalizability of the findings. This is referred to as
descriptive adequacy. The researcher must strive to provide accurate, detailed, and complete
descriptions of the context and participants to assist the reader in determining transferability.
One strategy to enhance transferability is to include cross-case comparisons. The researcher
may investigate more than one case. If findings are similar, this would increase the possibility of
transferability of findings to other settings or contexts. In some cases, even a single case can be
compared with other cases in the published literature that might demonstrate transferability. Be
aware that there are threats to transferability, such as selection effects (the fact that the
constructs being investigated are unique to a single group), setting effects (the fact that results
may be a function of the specific context under investigation), and history effects (the fact that
unique historical experiences of the participants may militate against comparisons). The
researcher should recognize limitations of the study in the description.

Qualitative researchers speak of dependability rather than reliability. Recall that reliability
in quantitative research has to do with consistency of behaviour, or the extent to which data and
findings would be similar if the study were replicated. However, qualitative studies expect
variability because the context of studies changes. Thus, consistency is viewed as the extent to
which variation can be tracked or explained. This is referred to as dependability or
trustworthiness. Some strategies to investigate dependability are using an audit trail, replication
logic, stepwise replication, code–recoding, interrater comparisons, and triangulation. Audit trail
documents how the study was conducted, including what was done, when, and why; contains
the raw data gathered in interviews and observations, records of the inquirer’s decisions about
whom to interview or what to observe and why, files documenting how working hypotheses were
developed from the raw data and subsequently refined and tested, the findings of the study, and
so forth. Dependability can be demonstrated by showing consistent findings across multiple
settings or multiple investigators. The more times a finding is found true with different sets of
people or in different settings and time periods, the more confident the researcher can be in the
conclusions. In this strategy, two investigators divide the data, analyse it independently, and
then compare results.

Intrarater and interrater agreement are strategies for assessing dependability (reliability).
An intrarater method is the code–recode strategy: A researcher codes the data, leaves the
analysis for a period of time, and then comes back and recodes the data and compares the two
sets of coded materials. Because much qualitative research involves observation by multiple
observers, some researchers suggest interrater or interobserver agreement methods for
assessing dependability. Confirmability in qualitative research is the same as the quantitative
researcher’s concept of objectivity.
 Lastly, a research study should uphold ethical considerations in terms of: 1. Kind of
information obtained. The data collection methods involve personal interviews and observations
of participants during a prolonged time at a site. Depending on the nature of the research, there
is a great likelihood of situations arising that present ethical dilemmas to the researcher. A
researcher may come into unforeseen possession of knowledge about illegal activities or in
some cases may have actually participated in the illegal behaviour. 2. Researcher’s relationship
to participant. After spending a great amount of time observing or interviewing, the researcher’s
relationship to participants may gradually become less that of researcher and researched and
more like friendship. Because the researcher is regarded as a friend, the participants trust him
or her and may forget a research study is going on. Somewhat related are the issues of
anonymity and confidentiality in qualitative research. Participants may expect and have a right
to anonymity, but it can be problematic in qualitative research. The researcher cannot promise
anonymity if, as is usually the case, he or she knows the names of the participants. In field
notes, they can use false names or code numbers to keep track of what information came from
whom without revealing identities. The confidentiality issue may arise in educational research. 3.
Reciprocation. The people in the research setting have given of themselves to help the
researcher, and he or she is indebted. Qualitative researchers need to give participants
something in return for their time, effort, cooperation, and just tolerating their extended
presence. For example, assume a researcher has conducted a lengthy study of students’
achievement, parental involvement, and the teachers in an elementary school. At the end of the
study, it would be appropriate for the researcher and his or her team to give something back to
the school. 4. Getting permission to conduct research. Especially if minors are included in the
research. It is often more difficult for a qualitative researcher to inform the committee about the
specifics of the research project because in qualitative research one does not always know what
is going to happen. Qualitative research plans may be altered as you get the study under way,
and unanticipated ethical issues may arise that you will need to resolve. Usually, qualitative
researchers inform the committee in a general way about the project. If it is necessary to
change the focus or design of the study, then the researcher needs to get an amendment from
the committee (Ary et. al, 2006).
 References:

Ary, D., et. al (2010). Introduction to research in education. Wadsworth CENGAGE Learning:

USA. 8th ed. ISBN-13: 978-0-495-60122-7 ISBN-10: 0-495-60122-5

Cohen, L., Manion, L., & Morrison, K. (2018). Research Methods in Education. Routledge: New

York. 8th ed. ISBN: 978-1-138-20986-2 (hbk) ISBN: 978-1-138-20988-6 (pbk) ISBN: 978-
1-315-45653-9 (ebk)

Best, J. and Khan, J., (2006). Research in Education. Pearson Education Inc.: United States of

America. 10th ed. ISBN 0-205-45840-8

 Reflection on Qualitative Research Methods

Although the nature of qualitative research methods highly differs from its quantitative
counterpart, this holds the same importance, same function, same fragility against errors, and
same demands for proper and appropriate execution. While quantitative can be purely objective,
qualitative research can go both objective and subjective.

Even if this research design does not involve much numerical values and statistical
computations, this also has its complexities. As discussed, collecting the data can be done
through interviews and questionnaires that are asked open-endedly. Meaning, the analyzation
and interpretation of data can be complicated as it involve feelings, opinions, or accounts that
has a high risked of being misinterpreted if not done right. There is also a higher chance of
having biases in the data collection as it can be determined by the researchers’ point of view.
That’s why this kind of design requires professionalism.

Moreover, this can take time depending on how much information is collected within the
expected time range and on the natural settings for observations and assimilation. Certain
factors should also be considered in collecting data – settings should work for both participants
and researchers to avoid external influences. Another thing is, as this is somehow an in-depth
research based very personal accounts, upholding ethical considerations is imperative.

Personally, a qualitative research design exhibits more complication than the

quantitative one as this can be lengthy, time-consuming, and critical. However, despite these
unaccounted opinions, this design serves same importance as the latter. Through this
framework, social, political, cultural, personal, and other aspects in society has been founded
and. This laid strong theories, arguments, and principles that run our current societal network
today. The undeniable work ad function of qualitative research in our community has been
apparent since its establishment.