Transcript for:
Exploring Research Methodology Essentials

In this video we're going to jump into the murky world of research methodology and answer some of the most common questions that we receive here at GradCoach. We're going to look at one, what exactly is research methodology in simple language? Two, what's meant by qualitative, quantitative, and mixed methods in terms of research methodologies?

Three, what exactly is and what are the options around sampling, around data collection methods, and around data analysis? And And four, most importantly, how do you as a student go about choosing the right research methodology for your project? So grab a cup of coffee, grab a cup of tea, and let's jump into it.

Hey guys, welcome to another episode of Grad Coach TV, where we demystify and simplify the sometimes seemingly bizarre world of academia. My name is Derek, and today I'm joined by one of our very own coaches. Karen Karen.

Karen is a seasoned researcher. She's been published in various peer-reviewed journals. She's authored textbook chapters. She's got a PhD, MSc, BSc. Basically, Karen knows what she's talking about when it comes to research.

So today I'm going to be picking Karen's brain. about all things research methodology related. We're going to unpack those topics so that you can understand it better and that you can make informed choices about your own research methodology. So let's get into it.

All right, Karen, so welcome back to Grad Coach TV. Thank you for sharing that brilliant mind of yours with us once again. So we're going to talk about research methodology, which is something that is a common pain point for the students that come through to us here at Grad Coach.

So let's start with the basic question. What is research methodology? What does this mean in simple terms?

Yeah, that's actually a really good question. But the simplest, nastiest way of... of talking about research methodology is to just talk about it as the doing chapter, the how did you do it? And just to sort of set the scene, to kind of get us all on the same page.

At this point, you would have already... probably undergone two chapters, maybe a bit more depending on the nature of your thesis, where you would have looked at your introduction and you would have discussed the previous literature throughout the literature review. And so at this point, it would have been clear to the reader, the examiner, the reviewer, what your research is about, the general questions or objectives of your research. And it would have...

also been relatively clear, hopefully, to them that you understand the background of your research, that you understand the previous literature, the ways in which other people have undergone thinking about this problem, researching this problem or similar problems. And it will be very clear that you probably know what your research is and why it is important. So that is essentially just sort of setting the scene as to where you end up being by the time you get to your research methodology.

But at this point, now you've got to say where you fit in, right, in terms of what you are going to be doing to help you answer that question that you've spent two or more chapters justifying and explaining. So that is why. That is important because that means that this chapter is very, very essential at sort of placing where you are and what you're going to do.

And so it's the how chapter. So just to jump in, to simplify this a little bit, what you're saying is the introduction chapter is about the what. It's about what are you going to be researching and perhaps a little bit of the why. And then the literature review chapter is essential. assessing what other people have done and reinforcing your why, reinforcing why your research is important, why your angle is important.

And then the methodology chapter naturally comes after that. And the research methodology is about how, in other words, how are you actually going to undertake your own research? Is that a fair assessment? Exactly, exactly. It's the practical.

You would have probably spent the whole of the previous chapter going into all these theories and paradigms. bigger picture ideas. And now you're going to be like, well, well, what am I going to do about this? How am I going to go about this? And so what you tend to really want at the end, by the time when you start writing your methodology, and to be honest, you want to know these things as early on in your research as possible anyway.

So you actually kind of want to know already, probably by the time you've already started, but definitely the reader and the examiner wants to know what data... you collected right who or where you got that data from and by and by data I mean those units of information that are going to help inform or help answer your question whether it is things like that you measured or questions that you asked people or the answers to the questions that you asked people that is all data right and and so it's the it's the what data is where it comes from or who it comes from, the how you collected it, right? Was it through, for instance, surveys or interviews?

But we'll talk about that in a moment. And the how you analyzed it. So what did you do about that data to make sense of it and to help you answer that question? And if you've got those components, then generally speaking, you should be fine in terms of having...

explicitly described what your methodology is. But by the time you've answered those sort of general ideas within the methodology, remember that it should be very clear to whoever's reading your thesis what you did. To the point that they should technically be able to repeat or mimic what you did to as high a level as possible, to as low a level as possible, if you will.

Right, right. So you're basically providing, to really oversimplify this, you're basically providing your recipe to how you approach this research so that others can go and reassess results using a similar recipe or using the same recipe. Yes, I've hesitated to use the word recipe. Fair enough. Fair enough.

because a recipe is technically what you're wanting to. If you think about the methodology chapter, it's actually different from what you would see as a series of methods. So it would be different of a recipe because a recipe is just the different steps, right? And yes, that is important. It's important to know that method, that series of steps.

But you actually want to go a step further and to sort of say... justify it, right? Justify each of those steps.

So in the case of a recipe, you're going to whip the egg whites, right? But in the case of a methodology, you're going to say why you whipped the egg whites, right? It's in order to get some kind of volume into your recipe or whatever the reason is. So you also want to be able to justify those steps as well. You don't always have to hop on about it, but you do want to make it relatively clear.

that you know why these steps were crucial in terms of your understanding of what you're doing. It's a great point and something that I've often mentioned to students, that the research methodology that you adopt and the research methodology chapter that you write needs to be more than just an account of what you did and how you did it, but it needs to explain why you made those choices. We'll discuss some of those choices.

a little bit later in this video, but there are many choices that you are going to have to make in terms of your research methodology. And just as important as it is to choose the right one, you need to justify why you chose those because ultimately what you are doing in explaining why you went left as opposed to going right, what you are doing is you are showing that you understand research skills, that you understand how to approach research in a... Exactly. in a systematic fashion, in a formal fashion. So it's really important for students to remember that it's not just about saying what you did or what you will do, but also saying why you made those choices.

Yes, and also don't be afraid. of simple reasons as well. There are the higher level reasons, like these really important foundational reasons as to why you would maybe choose a specific method within your broader methodology.

But sometimes convenience is just a good one, right? Convenience, cost, those are... Perfectly valid justifications for a step.

Yeah. So don't be shy of those justifications, even though they might initially feel, ooh, but is this the best possible method to help answer this question? Actually, sometimes the most convenient method is fine. Okay, so just to recap, the research methodology, this ominous thing called research methodology, we can simplify that down to essentially being the how, the practicalities of how you went. and undertook your specific research.

The how in terms of what data you collected, perhaps numerical, perhaps textual, perhaps visual data that you collected, who you collected it from, how you went about collecting it, and then how you went about analyzing it. So those are the important components of the methodology. Is that a fair summary? That's a pretty fair summary, I think.

Awesome. So. So let's jump on to the next question. Alright, so one of the first pieces of terminology that students will stumble upon when undertaking their research for the first time is the likes of qualitative, quantitative and mixed methods. So Karen, what is this all about in simple terms?

Yes, so I just want to sort of backtrack one step and think about who is most likely to have that problem. Because for instance, if you're in the sciences or in engineering, it's very likely that You only have one of those problems, but if you're doing business or social sciences, then yes, you do have to be very explicit about the kind of data that you have. And a very sort of nasty general definition, quantitative would mean that your data is in the realm of numbers and is quantifiable, is measurable, is numeric, essentially, and many different ways of saying the same thing. Sure. And qualitative is in the realm of ideas and words and phrases and trying to make sense of those.

Right. So that's sort of just, as I said, a general nasty overview of the two. Right.

A mixed methods means that you're coming at it from you. probably going to have a mix of both. Some of your data is going to be in the form of numbers and some of them are going to be in form of ideas or words or phrases. In the sciences, in engineering, what tends to happen is they'll think of qualitative data still as things that you can count.

So because they're associated with words like there are so many green peas versus purple peas. There's no green peas or purple peas. White flowers or purple flowers on pea plants, for instance. And they'll think of those as qualitative because they're worded or they're categorized. but in almost every other discipline that will still be considered as quantitative because you're going to be counting those you're going to be using using those words in order to create a proportion or a percentage of what you're looking at.

So that's the sort of general nasty definitions of those. But it's also important to know why it's important to be explicit about that. The reason it's important to know whether your data is... quantitative, qualitative, or mixed is because that essentially tells you a little bit about your philosophy. And once again, in business studies and in social sciences, it's very common for you to be explicit about the philosophy.

And by philosophy, I'm just meaning the angle at which you're going to be looking at your research question, at your research problem. The two extremes. of the philosophies that people typically end up coming to us with is that they'll either be doing something in the realm of the positivist paradigm, the positivist philosophy, which is that they have a theory or a hypothesis or an inference, and they're going to use the data to test it, or they're going to have an interpretivist paradigm or philosophy.

And that is the approach they're going to look at their data at. And in that realm, they're going to actually do it the other way around. They're not going to come at their data with any assumptions or with minimal assumptions. They're just going to use their data to make an inference or a hypothesis or a theory. Right.

So it's almost like the reverse. The data drives the theory, whereas for positivists, the theory or the hypothesis will be tested by the data. Right.

So if that makes any sense. It's sort of the angle and the direction of which your data plays a role in answering your question. Just to try to recap and simplify that. So if we talk about research philosophies at a simplified level, we can say the two broad approaches are positivist and interpretivist.

And positivist is very much about testing hypotheses, confirming. theories about measuring and cutting and so it might sort of lean towards the the quant side whereas interpretivist is is very much about sort of starting from scratch and letting the knowledge emerge letting the theories emerge which then at a later stage might then sort of circle back and be tested with the positivist approaches is that fair to say exactly exactly and what you said was important as well which is that positivist tends to lean towards quantitative data right because quantitative data can be measured and you can get You could do some statistics on it, throw some statistics on it, and then kind of spew out a yes, no, this hypothesis is proven or disproven or at least supported. And that's very much in line with the positivist approach. And an interpretivist approach is a little bit more exploratory. It's a little bit more like, oh, well, let's actually kind of see, let's kind of generate a theory from the data.

So the data doesn't prove a theory. It actually almost generates the theory. Right.

you will. And there's neither is right or wrong. It's just, it depends really on the kind of information that you find valuable and that will help best answer your question. And as you said, you could almost do a little bit of an iteration between the two, right?

You can have the sense of, well, we think it might be that this is the hypothesis, but we don't just want to test the hypothesis. We want to take it a step further. Maybe you want to sort of...

ask people how they're feeling about something, right? Maybe a product on the market. You want to ask people how they feel about it and you could get a very easy quantifiable answer.

You know, maybe they're very happy with it or very sad with it and you can sort of quantify where they fit in that space. And then you can take it a step further and say, well, what will make you happier, right? In which case you're kind of giving them an option to give you a series of ideas.

Yeah. words, phrases of which you can generate a new theory. So that's sort of just a general nasty... overview of why it's important to understand quantitative or qualitative or mixed methods, because it really tells us the angle at which you're going to be answering that question and the kind of data that you're going to be using to answer that question.

Right, right. I think, let me just give you another dirty example. One thing that sort of just sprang to mind is this idea of, let's say your research question is, what is the value of...

YouTube for helping graduate students understand their research methodology. Very typical. Yes. In which case you could have several ways of answering that question, right?

Maybe we're coming at it with this idea of we just want to know the value, right? And we're just wanting to analyze the value in these ways. And we're just going to give people a sort of a very valuable, not very valuable scale.

scaled way of answering the question. In which case we could just send out a whole bunch of surveys, get that scaled data back, that numbered data back, and then we just quantify it. Yes, so many people think of YouTube as a very valuable resource, right, in the graduate school programs.

But there is also that qualitative interpretivist way of looking at it, which is we don't just want to just sort of know a number, we actually want to sort of explain explore that a little bit more, in which case we're probably going to want to interview people, for instance, maybe interview graduate students about their broader use of YouTube in trying to understand research methodologies, in which case you will have that qualitative data associated with it. Right, right. All right. So to recap, qual, quant, and mixed methods.

Qualitative is essentially At least for most areas of study, qualitative data is about words, it's about images, it's about data which is not numeric, whereas quantitative data is all about the numbers. And then mixed methods is sort of mixing the two together. And then if we extend that through to how that links to research philosophies, there's a link between the qual and quant and positivist and interpretive. Is that fair to say? Pretty much, yeah.

Cool. Awesome. So let's jump on to the next question. Okay, so one of the important parts of research methodology, as we discussed earlier, is understanding or choosing what data you're going to be looking at.

In other words, what portion of data one is going to be looking at, and that is the field of sampling. And this can get really, really complicated and really, really intimidating for a lot of students. So can you give us a... a basic overview of what is sampling all about and what are the main decisions that a student needs to make in the space? Yes, and that's a very important question because ultimately that is it's going to determine the sort of accuracy if you will of your data and so understanding what a sample is is crucial.

But I think in order to understand sampling you want to understand population. The population is every possible person who could be of interest to your research. So if my research is on South Africans, then the population is 50 million people. Right, everyone.

But for me, that would be ridiculous. there's no way I'm going to be able to connect to everyone so what you want what you will get for research is only a proportion and that proportion, that small proportion of South Africans in the case of my potential research will be the sample and so that's just a general nasty once again a nasty overview of that and the assumption is that that is proportion, that sample is generally speaking representative. And that really is dependent and that will determine the kind of sampling you do, the likelihood.

that you will have a fully representative sample. Right. So before we jump into more detail there, I guess to simplify this down into an analogy, we spoke about the population being a... essentially all the people that are of interest to your research and the sample being the group of people that you actually have access to or you choose to connect with.

To create an analogy, yeah, if we had a big chocolate cake and we wanted to taste what the chocolate cake tastes like, we would cut a slice out of that and we would taste that slice. And so the analogy would be that the cake itself is the population. We're interested in...

what their population tastes like, but we certainly don't need to eat the whole thing. It would be impractical. So a sample is essentially a slice of that cake. And then to your other statement about generalizability, depending on what part of that slice we ate, that would give us either a view that gives us a view of what the whole chocolate cake tastes like, or if we just ate the icing, for example, we wouldn't have an accurate view of what the whole cake tastes like. Is that a decent analogy?

Oh, that's nice. I'm just imagining somebody going around eating the icing, no matter how much you want to. And exactly, exactly, the people in the medical school, matter how much they want to, there's a lot of barriers to eating the entire cake. There's a lot of barriers to assessing the entire population.

And those are generally speaking in the realm of convenience, annoying the people you're wanting to get information from. There's a lot of reasons to not do that. But having said that, you very rarely are in a position where you actually need to look at everyone in a population. It's not only not essential, but it's actually just quite often a waste of money and a waste of resources, a waste of time.

So you only need a small proportion of that population. But that proportion, as you mentioned, is important. Yeah. or at least you need to be explicit about where that proportion comes from.

So where that sample comes from. And that is why the sampling section of a methodology is very important. What you tend to get are two different kinds of sampling methods, two broad kinds of sampling methods.

You get a probability sampling and a non-probability sampling. In the context of a probability sampling, it means that your sampling method of choice has been designed in such a way that the people who you sampled are representative of the broader population, right? Right.

So if I were for instance, as you mentioned earlier, you need to be able to have dissected that entire cake, made sure that you've got a little piece of every section in order to understand, in order to be able to generalize the cake. And that is, it's very tempting to think of that as the most important or the most appropriate way of doing research. But there is another kind of sampling, which is the non-probability sampling, which is that you actually are using a sample that is convenient to you, that is targeted, that is perhaps well thought out for multiple other reasons, be it convenience and time.

In fact, convenience sampling is an example of non-probability sampling, just because those people are a available to you. And while you cannot generalize non-probability sampling to the same extent as you can probability sampling, that doesn't mean that it's the wrong way of going about it. Sometimes a convenience sample means that you're not annoying the people that you are interviewing or surveying.

It could mean that you are saving resources and you're creating and you're allowing yourself. to explore a certain topic a little bit further, right? Before maybe doing something a little bit more random.

Right, right. Random sampling, probability sampling has... therefore also has its inconveniences and one of them is that it's actually not very easy to to get everybody that is representative within a population if I think about um once again I think I use that analogy of South Africa as an example it's going to be very difficult to be able to draw from at uh the different socio-economic racial gender all of these different think as ways in which people can be categorized in a convenient, appropriate way that can therefore make it representative of South Africa.

But if I am thoughtful about my sample and am explicit about the sample that I used, then I can be, that I could still make a a valid research level, academic level statement based on the data that I get from that sample. I guess it's very much about being well aware of what the limitations of your... sampling approach are so that you can draw insights out but you are careful about the extent to which you generalize them.

So to drag us back to the cake example, you can see what I have on my mind here. To drag us back to the cake example, let us just say we did cut um the bottom half or let's just say we took a slice off the bottom that had a little bit of cake and a little bit of icing and a little bit of the underside of the cake that would that would not be a probability sample because it's only a cut it's it's not a it's not a clean slice a full slice of the entire cake um but if we acknowledge that and we say okay well we know that we we are looking at this portion of yeah then we can be careful about what we say and the conclusions we draw um and we know that okay what whatever our findings are, they apply to that portion of the cake as opposed to the whole cake. It doesn't mean that the findings are useless. It just means that you need to appreciate what the limitations are. Precisely.

And especially in a time like like now we've got coronavirus and all those things. So a lot of research is being done online. That means that you're not going to be able to maybe delve into the population or the populations of interest that you would have normally done. That doesn't mean that that research is invalid.

It just means that you're using what is available now. And that's fine. That's completely justifiable as long as you're explicit.

Okay, so to recap sampling. is about deciding which slice of the of the pile of the the cake one is going to take a slice of you're making me so hungry when when undertaking research you're interested in a population whether that's a population a large population such as the population of a country or perhaps a group of managers within an industry or a group of whatever you will always have a population that you're interested in But as a researcher, it's always going to be highly unlikely that you can reach that entire population. And it's also just unnecessary to reach the entire population. So you'll take a sample thereof.

And to recap what we discussed, the two main approaches are that you can take a probability sampling approach where you get a truly random sort of cross section, a slice of that population. So that might be a perfect full cut slice of the cake. All right. Or you could potentially go the non-probability route where it's just about convenience or your access might be limited in some other way, in which case the data is not useless.

You just need to know what the limitations of the data are and that you can't treat that as representative of the entire population, as representative of the whole cake. Is that a fair summary? Yes, that is a very fair summary. And acknowledging those limitations are important. Important examiners and reviewers are very happy when they see that you know what's going on and that those limitations.

Awesome. So let's jump on to our next question. All right, so the next important decision that students need to make in terms of research methodology is around data collection methods.

In other words, how they go about getting the information, how they go about getting the data. So let's talk a little bit about that. What are the most common sort of data collection methods?

Yeah, well, there's... A lot. I mean, basically, basically, there's, if you think of the world as data, thinking about collecting those methods, there's just an infinite number of possibilities.

But I think generally, speaking the kinds of data collection methods that are relatively common that we get from multiple different disciplines is in the realm of interviews or focus groups or surveys or observations those are probably the most common that we get here and once again that links back to a previous section where we're talking about quantitative and qualitative data, you will probably find situations where the collection methods for qualitative data are typically things like interviews, where you have something like this line, you're sort of, if you were interviewing me about my knowledge of research methodologies, it would be something along those lines, a sort of a back and forth between. the researcher and the participant of interest. Yeah. And this is useful if you're wanting to, say, get an idea or explore the ideas within a person's mind.

Right. Around a specific topic. So interviews are one approach. Another approach to the data collection for qualitative data is focus groups.

Yeah. Focus groups. Focus groups, if you're interested in still exploring people's minds, are useful, but they're useful for a different reason. Focus groups almost allow people to feed off of each other, if you will.

So in an interview… What's valuable about an interview is that you're sort of getting a glimpse into a single person's mind without them being influenced by external forces other than you as a researcher. And in an interview that's valuable because then you just have to observe and watch and listen to this one person. But in a focus group, people are sort of feeding off of each other. They're getting energy and ideas and thoughts from each other and building on them. And that creates a different kind of a value in the sense that maybe...

Maybe people forget about a certain idea and then somebody can sort of stimulate. stimulate that idea further or take it a step further in which case a focus group can be incredibly valuable and rich for that reason right but then the participants are influencing each other of course so it's like a there's a I catch you two of those sorts of things but that those are the two sort of very common qualitative uh approaches you also get things like document analyses uh analyzing texts maybe uh in terms of understanding history or historical records that is a very important way of doing things. You get ethnographic or observation data collection methods where the researcher is sort of stepping outside of a particular situation and maybe just making observations, making recordings or jotting down notes about what's happening in a particular setting. So there's many different ways of collecting that kind of data, but those are probably the primary or fundamental ways of collecting data. of collecting qualitative data.

Right. In the context of quantitative data, relatively straightforward ways of collecting quantitative data include surveys. Right.

Surveys can collect qualitative data. So I don't want to completely put it into the quantitative realm, but you can sort of just ask people open-ended questions within the context of a survey. That will still be qualitative in nature, but you can also allow people to sort of on- answer categorical data, such as maybe things like their race or their understanding about something, in which case you can give them a scale of one to five and they can say, yes, I feel very strongly about a certain statement, which will also be pretty quantitative.

So that's one way of collecting a lot of quantitative information from people who are participating in your surveys. You could do measures, measurements, right? Measurements.

are very much in the quantitative realm. And they can be done using a number of different instruments that are available from a simple ruler through to a CT scanner. And so there are many different ways of collecting quantitative data, but it really just depends. on once again your research question right so to just to recap on on the data collection methods i think it's important for for students to understand uh we're going back now to to the broader the broader thing of philosophy what is it that you you're trying to understand what is your research question and how are you going to approach it and that will heavily influence what your data collection methods are is that fair to say totally yes and never forget never at any point in your writing of your methodology, forget that this all has to answer your research question. Right.

So as, once again, you can justify whatever collection you used. Convenience is always an appropriate way of justifying your data collection methods. Now we see a lot of interviews being done on Zoom.

Yeah. Not always done that way. It's convenient, but it also allows you to get access to people who might you might never have had access to previously just because they're too far away as you mentioned exactly exactly as now um it's a little bit easier but it also there are limitations there too um obviously seeing people's ways of answering questions can sometimes give the research a little bit of a clue as to whether a topic is uncomfortable, for instance.

And that's a little bit less easy to see on an online call. But having said that, interviews and surveys are definitely of choice right now, I think, for obvious reasons, as you said. And yeah, I think that's important. But once again, never forget that at the end of the day, you're directly... collection method is still feeding into answering your question.

Yeah. And it needs to be appropriate for that. Awesome.

Awesome. So I think the key takeaway, yeah, or really important takeaway is that you can't, you can't go about your research starting with your data collection method. We see this sometimes students come to us and they say, I I like Survey Monkey and I want to undertake a survey for my dissertation.

And that's really, really the wrong starting point. As you said, the research topic and the research questions... those are the things that dictate what approach you're going to take. Those are the things that dictate the methodology and therefore the data collection methods.

Yes, exactly. And I think what you said is an important thought, right, is that you cannot have the how before you know what it is you're looking at. And quite often, I think you actually mentioned earlier that an interview itself comes from a very specific point of view.

comes with you having a good understanding about your research question. It's very rare, unless in very specific kinds of approaches, that you will come into an interview kind of just having a chat. You really want to be at least somewhat prepared.

And this could generally be in the realm of you coming in with a set of questions that need to be answered, a very structured approach, a semi-structured approach where you have a general set of questions that you allow. allow the participant to go into a certain series of directions with but you'll sort of rein them in or a very unstructured approach where you kind of will just generally have a chat and just generally see where it goes in a certain direction yeah but exactly so to kind of see where it goes and so you do want to make sure that you are clear about linking that collection to your question because the question is what's important all right so now that we have spoken about data collection, the next question is naturally about data analysis. So let's jump into that. Okay, so let's talk data analysis or data analysis methods.

What are the sort of main approaches there and what do students need to know? Yes, in my experience, by the time I get a methodology chapter, This is usually the smallest section in everyone's methodology chapter. And for good reason.

It's very difficult to be sure about what you're doing, right? And I think that analysis is the part that scares people the most. So what you tend to find, as you said earlier, is that people know very much that they're wanting to do a survey or that they're wanting to do an interview.

And then when it comes to the analysis, they're like, actually, we all actually, especially the first time we've done a data analysis of this kind, don't really know how we are meant to describe it, or what we're meant to say, or what is appropriate. And I think, once again, just to always keep in mind this idea of every piece of this research needs to be able to answer that research question, your research, and be justified and justifiable. And this includes the analysis. Right, right. And in the case of qualitative data, you at this point have had your interviews, done your focus groups, maybe gotten some open-ended answers within your surveys.

And now you need to make sense of it all. But you need to make sense of it in light of your research question. And so what you tend to have are a number of ways of approaching your data. and looking at the information, looking at the words and the ideas.

Once again, I'm keeping into the qualitative for the moment, the words and ideas, and trying to sort of group them into a sort of a series of understandings, if you will, a series of themes. And that is the typical way of going about qualitative data is to essentially read, listen, and read and listen. over and over and over again until you're so sick of hearing those people's voices or reading those sentences but you are so sure about what they're saying and how that fits into your research question so in the in the case of qualitative data we have a series of approaches that you can take on how to think while you're listening to your interviews over and over again or reading the responses over and over again And a couple of examples of this is to do a content analysis where you are just listening to what they're saying and theming them according to what the participants in your interviews, for instance, were saying. Right.

So if they were, for instance, talking about if you were doing research on the value of YouTube for research methodologies and people went on a little bit of a tangent about supervisors, for instance, then that could be a theme. Yeah. Right.

And and the. way that people speak about their supervision at university level might vary from interview to interview, but you can always sort of say, well, actually, this is how they've discussed, how we've discussed the value, and supervisors tend to come up in multiple different ways, but they all end up talking about supervisors. So that's one of the ways in which you could sort of, for instance, theme or approach the analysis.

is in theming it by the content that naturally comes up during the interview or the focus group process. You could actually be coming at your data with a very specific lens, such as in understanding the discourse between two people, for instance, the discourse analysis. So, for instance, if you're observing two people having a discussion or in the focus group observing people... participating in your focus group and you see that some people are a little more shy and some people seem to be a lot more authoritative that might be because of varying power dynamics within that space and that could also be of interest to you and then of course you've got another way of approaching your data which is to look at the narrative style of the data how are people talking you about their experiences.

I see narrative approaches are being used a lot more among nurses now because people are interested in hearing about how nurses feel about their about the situations in which they've been placed in to help other people. And whether they're coming at it from a very hopeful direction or whether they're coming at it from a very uncertain direction is very important. And the way they tell that story is important. So these are various ways of sort of analyzing qualitative data, if you will.

And there's a very nasty examples. I think hopefully they give you a broad overview. So just to recap that on the qualitative content analysis side, the three common approaches that you've mentioned are one, content analysis, where it's really about picking up on themes and ideas that emerge from review of the data, whether that's interviews or transcripts, whatever the case may be.

There's discourse analysis, which is about... understanding how people speak to each other and what that reveals about the the environment and then there's narrative analysis which is about just understanding the stories that are being shared and the meaning of those. Is that a fair sort of summary of the three common approaches we see?

Exactly. Exactly. I think you did it better than I did. Cool. So on to the quants.

What's quant all about? Well, I mean, I love quant. So I'm always going to be valuable.

Well, not valuable. I'm always going to be excited about this kind of stuff. So you probably have to rein me in a bit. But a quick and nasty way of thinking about it.

about quantitative analyses is to think about whether what you're doing is describing your data or almost sort of comparing your data, making inferences with your data. And in the case of describing your data, you know, you've taken a whole bunch of measurements, for instance, about on something, or people have answered a whole bunch of questions, then you can use descriptive statistics to describe how people answered or what you measured. And descriptive statistics include things like taking the the average or the mean, right? And saying, well, the average person feels very happy about using YouTube as a medium to understand their research methodology. Or the average person is 1.6 meters tall.

In my mind, everyone's really short because I'm really short. So. So that's a perfect average for me. And that's just a general quick and nasty way of looking and showcasing your data. But that's nice.

I love descriptive statistics because it tells me that you know what your data looks like. And a lot of people tend to go, tend to jump a whole bunch of steps and go right through to inferential statistics and comparing various different kinds of data. Right. Or using things like regressions.

or correlations or even sort of ANOVAs and t-tests and chi-squared tests which are just a lot of fancy ways of saying how certain kinds of data sets compare with each other or the kinds of trends that they make together and um and that's important and valuable to know because quite often the magic or the way in which you answer a question actually lies in how you compare different kinds of data but that cannot be done unless you know what you're your data looks like or it shouldn't be done unless you know what your data looks like so descriptive statistics should always be considered and then the ways in which you infer or make sense of your data further are other ways in which you are the ways that you can link the data to answering your question right if that's sort of a general once again a big nasty overview of quantitative statistics yeah so to to sort of recap the uh the the two sort of pods or not pods, the two areas that one would delve into with quantitative analysis is first descriptive statistics, understanding just the nature of the data, understanding the nature of the sample, looking at what the averages are, who is involved and really getting a feel for what the data is all about. And then the inferential statistics such as regressions and so forth, that's really about understanding the relationships between different variables, understanding how things compare to each other. And that might be sort of at the core of what your research question was about, for example, understanding the relationship between. Watching Grad Coach YouTube videos and the end mark on a dissertation, one can collect that data and see, okay, what is the relationship that would be inferential, right?

Yes, exactly. And that's usually where a lot of the very foundational sort of quick summaries of what you ended up getting out of your research ends up giving you. It's from the inferential statistics. Right. But you shouldn't.

do inferential statistics unless you know what your data looks like. Right. And so even if it doesn't end up making it into your thesis, make sure that you know what your data looks like. I cannot stress this enough, how many people do certain kinds of very exciting methodologies, but they actually fundamentally don't know what their data says. Yeah.

And then they're very confused by the outcome. because it's like it makes no sense and it's like but you don't but it's because you actually didn't look at your data to say okay well where were the nuances in terms of how people might have answered each specific question where is the um what are the proportions in which people looked at uh specific or the measurements for instance or or the proportions of counts and that's actually really important information to just get a sense of okay well this is where I'm standing and this is and therefore I can use these kinds of analyses and another important thing is that a lot of those inferential statistics cannot be done done on data that looks a certain way. One of the things that we find, especially in survey questions, is that you tend to see a lot of people suddenly answer a certain kind of question in a very specific way. And you will find, for instance, in a Likert-based question where you've got a scale of one to five, the threes tend to go way above the roof, right?

Everyone sticks to the safe answer. Yeah, because they don't actually, but it's not because they actually feel that way it's because they just don't know how to answer that question or because it's not applicable to them so they just clicked three and that is very important information to know before you start comparing your data to each other because you want to know well why why is everyone suddenly answering three for question five and and another another important function of of that descriptive data especially the demographic data of age and gender and ethnic group and so forth. And an important way to use that is to sort of use that as a sense check for whether or not the data that you've collected is representative of the population. So if you know some details about your population and you know that the population should be 50% female and 50% male, then you can check your descriptive statistics to see, okay, well, does... does my data actually line up with this or do I have a bit of a skew towards one side or the other?

So you can use that descriptive data, that basic data that a lot of students just kind of float over. You can use it to check, okay, am I representative? If I'm trying to create data that's representative, do I actually have that?

Exactly. I don't think I've ever been in a situation where people have answered a survey 50% male, 50% female, especially with online surveys. It was always like at least a 60% female.

sort of lean and understanding that is important because that really allows you to sort of take that step further in the sense of maybe you actually don't want to compare the different variables to each other for the entire sample maybe you want to actually break it up into males and females just to see what's happening a little bit of a further level but you don't usually know that you want to do that until you've had an exploration of your different data using what is otherwise very simple statistics. Yeah, yeah. All right. So to recap on the data analysis methods, obviously, as we've discussed, the methods will be heavily heavily heavily influenced or completely influenced by whether or not you're going the qual or the quant route.

You can't go and analyze quantitative data in a qualitative fashion. So on the qualitative side, content analysis is sort of the simplest level. And we spoke about discourse analysis and narrative analysis. And then on the quant side, descriptive statistics, don't overlook them.

They are the foundation of your data and a really important way of making sure you understand your data. And And then, of course, the exciting stuff is the inferential statistics to try to understand the relationships and the differences between different variables. Yes. And to be honest, it's actually a little bit more if you were to be very, very, very fancy and you're wanting to do sort of supervised or unsupervised machine learning and that sort of thing with your quantitative data. That is all possible.

But at a very basic level, you probably don't need a lot more than descriptive and inferential statistics. Right. Right.

All right, so let's jump on to our final question. All right, so on to our final question for this video and possibly the most important one. The big question is, how does a student that is fresh into research, that is starting out, how do they go about choosing their research methodology?

We've spoken about all these different choices today and all these different directions that one could go. So how do they decide or at least how do they start thinking about what research methodology to adopt. Yeah, I actually have a very particular way of going about with new clients on going about doing this. And that is right at the beginning.

You know, they've just gotten, they've just freshly gotten their new research topic or decided on their new research topic. And they kind of have a vague idea as to where they're going, but they don't know too much else. In which case, what I'll do is we'll sit together and we will plot.

These four things being the title and the question, or five things really, the aims and objectives and the methodology. And the reason I do that is, and it really is just a page of work, right? It's just a single page where we make sure that...

that the title and the question go hand in hand, and that the question and the methodology go hand in hand, and that the aims and objectives sort of fill that intermediate space of... repeating the question in an actionable way and making sure that that action is linked to the methodology so the reason this is important is because we allow each other to interrogate what they're going to do if you have a title and a research question in mind whatever methodology you use needs to be able to answer that question right and by iterating between these different components we're making sure sure that we are using the best possible methodology to answer that question and that those that question is therefore tied into those aims and objectives right if that makes any sense right okay so so students i mean we we understand this broader principle of um the the research question the research objectives etc etc these things dictate and at least heavily influence how one design science and methodology but what's the starting point and what what are the first sort of things that a student should think about in terms of the nature of their research and what might that mean for the direction they take and that's actually really important it really sort of filters back into that sort of idea of research approach are you going to be having an exploratory or interpretivist approach to your research in which case you're wanting to sort of figure things out in which case you are probably heading into that qualitative um interview focus group realm right and then it just becomes about practicality who do you have access to what's straightforward and easy and appropriate to do, maybe given the people that you're trying to figure out information from. And in which case, that ends up just being the big dictator. And it is important to just think about those details right from the beginning, because you do want to know, well, if I'm doing interviews, do I have access to those people to interview? Do I have the space to do the interview?

Should I be recording the interview? These are little practicalities that become really important at some point for you because you have to set up the interviews, right? So that's one way.

Another direction is once again into that sort of confirming hypothesis testing, a positivist realm where you are going to probably be using quantitative data. In which case, once again, you need to make sure that you have the instruments to do. do the measurements that you need to measure, or you need to understand that you might need to do a survey of sorts, right? And where to set up that survey.

So those are the kinds of things that you need to think about is the direction, exploration, theory building, or theory testing, and confirming direction. And then the practicality. What is the most reasonable way of doing this considering my budget and the instruments that are available to me? Right, right.

Okay, so just to sort of recap that and emphasize the important bits is the starting point for someone. who is just starting to think about their research and has no idea in terms of methodology would be to to first understand the thing that you're researching is the nature of that exploratory. Are you sort of starting from a blank canvas and you're wanting to sort of uncover and try to build a theory from nothing, or at least from a very sort of nascent start?

Or do you already have something, a theory that you're wanting to go and test, something that you're wanting to confirm in a specific context? Exactly. exploratory versus confirmatory and if you are on the exploratory side chances are you're probably going to err down the road of qualitative data and interviews and potentially focus groups and if you're on the confirmatory side you're probably going to take a more quantitative approach and whichever side you're on as you say Give some thought to the practicalities and how you're going to actually collect data and what your limitations are as a researcher. Is that a fair summary?

Absolutely. Awesome. Well, I think that pretty much recaps this video of research methodology.

We've thrown a lot at our poor viewers today. Hopefully some of that sticks. Karen, thank you so much for your time and for dropping the knowledge bombs on us again.

And we'll see you in the next video. Bye, Derek. Well, that wraps up this episode of Grad Coach TV. Hopefully, you're feeling a little bit more confident. Hopefully, you're feeling a little bit more informed.

And you can now tackle the research methodology of whatever research project you're working on. Remember that Karen is... of our grad coaches that help students like you every day with their research so if you're interested in getting a helping hand with your research visit gradcoach.com and you can book a free consultation with one of our friendly grad coaches just like karen so there you have it in this video we have unpacked research methodology just a little bit we've looked at one what is research methodology Two, what exactly are qualitative, quantitative and mixed methods approaches?

Three, what are the main choices around sampling design, around data collection, and around data analysis? And four, the most important one, how do you go about choosing a research methodology for your research? If you have any questions about anything we've discussed in this video today, please do leave a comment below.

We'll do our best to reply. And you might also want to check out the Grad Coach blog where you'll find loads loads of other free resources, free information covering research methodology, covering pretty much everything research related from literature review through to analysis through to writing up. So check out the Grad Coach blog at gradcoach.com forward slash blog.

If you enjoyed the video, please do give us a thumbs up. And you might also want to subscribe to the Grad Coach YouTube channel because we are doing our best to produce more content like this. If you're writing a dissertation, thesis or any sort of research project, you'll probably find.

our other content to be quite useful. So from me, Derek, thanks for watching. This is Grad Coach signing out.