Introduction to Qualitative Methods for Public Health Research

SPEAKER: So to do qualitative researchyou need to take five steps, and I'mgoing to walk you through each of those today.Number one, define your research question.Number two, identify the appropriate research methodto answer that question.Number three, collect data.Number four, analyze the data.

SPEAKER [continued]: And number five, communicate or write upor present your research findings.In this video I'm going to walk youthrough each of these tips in a little bit more detail.Now, obviously, we're not going to get into absolutelyeverything, but you are at least goingto feel comfortable with the ideas thatunderpin qualitative research.So let's get going.So the starting point in this videois that I'm assuming that people havea general understanding of what qualitative research is.

SPEAKER [continued]: But just in case you're watching this and you're not sure,let me give you a quick overview.Qualitative research, as distinctfrom quantitative research, looks for meaning in the worldaround us using non-numeric data.In other words it's not using numbers,it's using other kinds of datas like speaking to people,getting people's opinions, getting to knowwhat people think, for example.OK.Let me illustrate that.

SPEAKER [continued]: If I were to count up the number of patientsthat I have in one of the wards in a hospital,that number would be quantitative research.That would be quantitative research.That number would be quantitative research.We've counted it up.It's a quantity.If, however, I was interested in howthose patients felt about their treatment,there's not a number there.I need to speak to them.I need to extract a different kind of data from them.

SPEAKER [continued]: And that is what we call qualitative research.And that is what we call qualitative research.Got it?OK, let's keep going.So let's get into these five stages or these fivecomponents of your qualitative research project.Now, the first thing you want to dois you want to generate a well developed research question.Your starting point for generating research questionis some sort of interest.

SPEAKER [continued]: You're interested in, for example, mental health.Fantastic.Now you want to start adding details,and you want to be as precise as possible.And you want to start adding parameters around that interestso that somebody else reading your research questionknows exactly what it is that you're trying to understand.Let me give you an example.You might want to understand the use of alcoholas a coping mechanism among small business owners in London

SPEAKER [continued]: during the COVID-19 pandemic.Now we've got a much clearer objective or question.If you're not sure if you've got a good research question,look at the literature.Who's been writing?Who's been studying in this area?What questions did they ask?Look at the literature.Look at the literature.Has this question being asked before?OK.So get the question right, let's move on.Next we're going to talk about methods or what

SPEAKER [continued]: some people call methodology.Now, you've got to get this right for a few reasons.First of all, if you don't get the methods rightyour actual research won't produce meaningful results.Your results won't be valid or reliable.Secondly, you need to be able to describe your methods wellso that other people, people reading your research,will understand how it is that you got to where you got.Your research should be replicatable.And, of course, people need to understand

SPEAKER [continued]: that what you've done follows a scientific method.So there's three aspects to methodsthat we're going to talk about.And I'm not going to get into a lot of detail on all of them,but I want to give you this synopsis.First of all, there are different kinds of research.Secondly, there are different waysthat data can be collected.And then, thirdly, there are different waysthat data can be analyzed.So let's talk about all three buckets.

SPEAKER [continued]: First, the types of qualitative research.Now, there are many, many types of qualitative research.We can't get into all of them in this video.The four main types-- the four most commonly used typesof research are case studies, ethnography, phenomenology,and grounded theory.And we're to talk about each of these quickly.Case studies look at a single example of somethingin a lot of detail.So a case study might explore a process, or an event,

SPEAKER [continued]: or an individual, or an experience,but the point is it goes into a lot of detailfor that single entity.So, for example, you might do a case studyinto how a specific or particular hospital dealtwith the COVID-19 pandemic.Next, ethnography.Ethnography studies people in their own setting or sometimespeople say in their natural setting,and it studies them over a period of time.

SPEAKER [continued]: So, for example, you might want to look at health care workersin a hospital setting during the COVID-19 pandemic.Phenomenology explores the human experiencearound a particular phenomenon.And that phenomenon could be an object, an experience,an event, or an idea.So, for example, you might want to explorehow patients experience and understand the carethat they're given while they're in a hospital.

SPEAKER [continued]: And the last one I want to talk about is grounded theory.Now, grounded theory is a little bit different in thatthe output of your research is a theory,is a generalizable or abstract theoryabout how the world works.And that theory as to how the world worksis meant to be grounded in your data--usually the opinions and thoughts of your researchsubjects.So, for example, you might want to interview people

SPEAKER [continued]: to get a general theory as to how it isthat people react in a crisis.Let's talk about data.Three kinds of data or three sources of data--it's going to be something you hear,so it could be an interview or a focus group.It could be something you see, so observational data.Or it could be something you read, so content analysisor document analysis.Those are three main sources of data in qualitative research.

SPEAKER [continued]: Interviews are conversations that we have with one or moreof our participants, and we explore their thoughts,their feelings, their opinions about a particular phenomenon.Now, usually when we do an interviewwe record that interview so we have an audio recordingof that interview.And then we'll transcribe it so that we'vegot a written transcript that we can use for our analysis.We're going to talk more about that in just a minute.

SPEAKER [continued]: Interviews can be structured, semi-structured orunstructured.And what we mean by that is the extentto which the interview follows a very prescribed order, a verypredefined set of questions.And for some qualitative research, that's appropriate.Most of the time, however, most of the qualitative researchthat I've done is, you allow the interviewto take on a little bit of a life of its ownwithin certain parameters.

SPEAKER [continued]: In other words, we want it to go in a certain direction.We've got certain big themes and headings that we want to cover.But we're not going to prescribe every single secondof the interview and tell the person whatquestions to answer.We allow them to guide a little bit of how things unfold.Does that make sense?Focus groups are very much like interviews,but now you've got more than one person in the room.Now, the nice thing about focus groups

SPEAKER [continued]: is it allows people to interact with each other.So you can have a really rich source of ideas.You can have ideas develop, unfold, change, merge.It's all quite interesting.The whole thing is moderated by the researcher.And, again, it could be structured, semi-structured orunstructured.The next type of data is observationsor observational data.This is where we observe people directly

SPEAKER [continued]: and the researcher would then take field notesor have some other way of recording whatit is that they've observed.For this kind of research with observations,the researcher, him or herself, maybe a participant in the activity that's being observedbut they don't need to be.And the subjects may or may not knowthat they're being observed.And, finally, document or content analysis.This is any recorded communication

SPEAKER [continued]: that can be analyzed.Now, you might have more than one source of dataand, believe me, that's not a bad thing.Very often that's a very good thing.In other words, you for example, mightbe looking at documents between content analysis and documentsand you get a lot of facts and figures straight outof the documents.You complement that with interview data.You speak to people.You get their thoughts, their opinions, their insightsinto these facts and figures.

SPEAKER [continued]: Can you see how different methods can be mixed together?Different sources of data can be thrown into the pot.You can even add in a bit of numbers,get some quantitative research in there.The point is, what you need to dois make sure that you're very transparentabout your methodology, about the methods that you use.Make sure that a person reading your researchunderstands how it is that you got there

SPEAKER [continued]: and why it is that you chose the sources of data that you did.Now let's talk about collecting the data itself.The first thing that you want to get rightis you want to select the right participants.You want to get the right people into the study.You want to get people with a perspective, with the opinions,with the thoughts, with the knowledge,with the insights that are going to give youa rich understanding of the phenomenonthat you're studying.Of course you want to get consent

SPEAKER [continued]: from the participants or, in the case of documentanalysis and content analysis, youmight want to get permission to use and analyzethose documents.If you're going to be doing an interviewor you're going to be doing focus groups,you might want to develop an interview guide.This is basically a document thatoutlines the general themes that you're going to explore.And sometimes, depending on the kind of researchthat you're going to be doing, it

SPEAKER [continued]: might actually delineate what the actual questionsare that you're going to ask.Now, as well as the information that youcan extract from the participants with respectto their thoughts and feelings, et cetera,et cetera about the phenomenon that you're studying,you might want additional information about them.In other words, you might want to, in your study,be able to say we spoke to people within this age groupand they were mostly males or mostly females

SPEAKER [continued]: or they were people from South Africaor these were people from Europe.There might be metadata about themthat you want to identify as important that'll give contextto the answers that they give.You also want to understand and describethe cohort of people who decided not toor declined to participate in your study.That might be important.There might be some sort of systematic difference

SPEAKER [continued]: between the people that did and the people that did notparticipate in your study.And that's important in terms of interpreting the resultsthat you get and the information and the datathat you extract from the interviewsor from the focus groups that you undertook.Then, from a practical point of view,you want to find a time that works for them, a time thatworks for you, so nobody feels rushed during this interviewor during the focus group.

SPEAKER [continued]: You want to find a place or a venue that'squiet, that there's not going to be interruptionsor distractions.And you want to make sure that your recordingdevice, if you're going to record the interviews,is ready and working and you've tested it.And then you're good to go.Now let's talk about the analysis.The first step with the analysis isyou want to prepare your data.Very often that'll involve, for example, youtake your recordings, your recorded interviews

SPEAKER [continued]: or your recorded focus groups, and you transcribe them.You want to get all of that transcribed.And then, quite often, we'll take those transcriptsand it'll be an electronic form.And we'll upload them into software that helpsus do the thematic analysis.Now, you don't have to do that.People sometimes just do it in Microsoft Excel.Some people just do it old school with paper and penand highlighters on a desk.There's nothing wrong with that.

SPEAKER [continued]: There's no rules.The most important thing is that you'retransparent about the method that you usedand how it is that you got to the conclusionsthat you draw from the data that you've gotOnce your data is prepared, a popular methodof analyzing qualitative data is somethingthat we call thematic analysis.So the first thing we do with thematic analysisis we code our data.And I'm not going to get into the details of how to do this.

SPEAKER [continued]: But, very briefly, what it means isyou're reading through your data and you read it again,and again, and again, and again.You identify meanings that are common in different places,and you apply labels or codes to that bit of text.So the same idea starts emerging in different places,and we start seeing that.And the same bit of text may have more than one codeapplied to it.

SPEAKER [continued]: Once we've done that-- once we've coded the data,we then extract all of the data that has the same code.And we look at it separately and westart seeing different themes emerging.Now, depending on the kind of research you're doing,your starting point even before you start looking at the datamight be that you've got a set of codes and a set of themesthat you're going to use.And when you look at the data you basically

SPEAKER [continued]: want to stick the right ideas into the right buckets.Or you might have those codes and those themesemerge from the data itself.So both methods are absolutely valid.Once you've analyzed your data, youwant to communicate your findings eitherin a presentation or quite commonly youwant to write a paper.Depending on the discipline within whichyou might find yourself, there may be different styles

SPEAKER [continued]: and formats that papers written in that discipline follow.I'm going to talk you through how a typical medical paper iswritten in the medical sciences.But the principles apply, and I'msure what it is that I'm going to sayis more or less applicable no matterwhat discipline you're from.The first part of your paper is the introductionand background.This is where you talk about your question.

SPEAKER [continued]: Why is this an important question?What do we already know?What's in the literature?What are the gaps in the literature?And how will it be that by answeringthe question you're asking, those gaps will be filled?Next is going to be your method section.Now, in the method section you'regoing to describe in detail what it is that you did--how you identified people, how you collected your data, wheredid you keep your data, how did you

SPEAKER [continued]: analyze the data, et cetera, et cetera.Not only what you did but why you did it--what was the rationale?Why did you do it this way instead of that way--so that a person reading the paperunderstands your thought process completely.And your method section needs to describehow it is that you ensured scientific integrity-- howit is that you ensure the validity and the reliabilityof your results.

SPEAKER [continued]: Next, the results section.Now in as much as possible, keep the results sectionabout your results, the data that you collected,the output of your analysis.Don't make the results section about what other people saidin other research, in interpreting your resultsand extrapolating, et cetera, et cetera.All of that can come into the discussion.Now the discussion section.Now, this is where things really get interesting.

SPEAKER [continued]: First of all, in your discussion,you want to make sure that you addressthe fact that your results have addressedyour original research question.So reiterate your research question.This is what we asked.This is what we found.This is how that answer relates to the question firstly.Secondly, address how does that your results relateto the gaps in knowledge that you found in the literature?So when you looked at the literature,

SPEAKER [continued]: these is what the literature didn't address.This was the gap in knowledge that I identified.And that's now been filled by the resultsthat I've identified, by the outcomesof this piece of research.You can also talk about the extentto which what it is that you found is consistentor maybe at odds with other findings in the literature.

SPEAKER [continued]: So that's a great thing to bring into your discussion.The discussion is also a great place for youto interpret your results a little bit.Speculate, extrapolate, talk about where the research mightbe going, what additional questions should be asked, whatadditional research could be done because of whatit is that you've found.And, of course, importantly, also talkabout the limitations of your study.And the conclusions section is usually short and punchy.

SPEAKER [continued]: What can we conclude?What are our actual take-home messages?What do we know now that we didn't know before?And, importantly, what are the practical implicationsof your findings?