An introduction to Research Methods (STUDENT SPECIAL)

Hi, you're listening to a student focused episode of the Sociology Show podcast. If you're studying the subject at GCSE, A Level Foundation degree, or any other discipline, then this podcast could be of help to you. If you have a question for the show, then you can email the Sociology Show podcast at gmail.com and one of the teachers, lecturers, examiners or experts will answer your question for you. You can sponsor the show on the GoFundMe page and subscribe on all the usual podcast platforms. If you're going to class, going to the gym or just chilling, put your headphones in and let's be Sociology Beats together, eh? Hello, and welcome to the Sociology Show. The Sociology Show podcast is brought to you in association with tutor to you, Sociology, the exam performance specialist for A Level and GCSE Sociology students and teachers. So you can visit tutor to you dot net forward slash sociology and there you can pick up revision guides, flashcards, revision videos and everything else that you need for your A Level or GCSE Sociology studies. The focus of this episode is an introduction to research methods and I'm delighted to welcome back Ben Hewittson from the All Sociology Take One podcast. Hi, Ben. Hello, Matthew. How are you? Very good. Thanks. All right here. Thank you. All good. Thanks. Very good. Very good. And thanks for helping out and also I'm going to do a bit of plugging you might want to as well. You've just started a GCSE Sociology YouTube channel, haven't you? Yes, I have. I've just ventured onto the world of YouTube and if anyone's interested to find some of the videos I've been doing, they're all about GCSE Sociology. You can just search all Sociology, all one word, or you can use the short link bit dot l y slash g c s e s o c. Brilliant. Great plug in there. I love that. Thanks, mate. And they're mostly about the family at the moment, aren't they? Yeah. So basically the whole reason I started these is really they're for my own students, but I thought I might as well put them in a format that will appeal to everybody because we've got my U11s, we've got their mock exams coming up at the moment. So I've done basically a small video on hopefully every single part of the specification. So there's been one on kind of like introductory key terms and stuff. I think five on family, I've done three on education and one on methods as well. And that wasn't so long ago. I did the methods one. So hopefully I should be fairly familiar with that stuff when we talk about stuff today. Brilliant. Excellent. Thank you for that. So let's let's get on to methods then. So we're going to do an introduction to research methods and we're going to start right from the very basics. Maybe you are a GCSE student or even if you're doing it at a level and I don't know what your thoughts are on it, Ben, but some people kind of see research methods as a bit boring and a bit dry, but it's kind of essential, isn't it? Yeah, I would agree with that. Nearly I thought I would say it's fair to say that most student groups I've taken through the A level sociology and GCSE have found research methods to be the least interesting part of it. But I think that one reason for that is it's not particularly a tangible subject. So it's not like the family where we can compare to our own experiences or our own educational experiences. But the thing I always say to students is that if you are interested in sociology, if you like the subject and you want to take it further, go into university and maybe doing it as a career, ultimately very few people are going to pay you to speak about Marx, Weber and Dirkheim, but they will pay you an awful lot of money if you understand people and if you understand how to elicit information from people, which effectively is what research methods is all about. And so what I always say to students is that this is a worthwhile topic if you are looking to get into the field of sociology to work as a sociologist, because you have to know this stuff. And the beauty of it is is that there are so many companies and employers that want social and market researchers because data is a hugely valuable commodity at the moment. And understanding people's data is what sociologists do. Yeah, definitely agree with that. And you just mentioned whichever level you take sociology at, actually, you do need to know it, don't you? Because even if you go on to degree, you'll probably do research methods again. Oh, absolutely. I mean, this is this is where sociology becomes vocational and where you can actually start to put into practice some of the stuff that we learn in lessons and lectures and seminars and stuff. And for me, that's quite exciting, but perhaps we can make it exciting for students today as well. I think sometimes applying some of the more interesting studies livens it up a little bit. So if we can integrate any of those, that will be useful too. Yeah, sure. So sure. Great. Okay, let's start from the very basics then. Probably what you'd first learn in class is the difference between primary and secondary research. So go for it, Ben. Okay, so it's really simple. This one, primary research is research that you personally have conducted and you've got your own data. Secondary research is when you will use someone else's research. So for example, if I go out and do a questionnaire on what toppings of pizza everyone who's listening to this podcast like, I am getting that that's primary data because I've written the questionnaire. I've asked people and I've got that data back. However, if I was to read your diary, Matthew, and read about how you gorged yourself on a pepperoni feast pizza at the weekend, that would be secondary data because you put that together. It's not for my purpose. I've managed to use that diary to understand that you love pepperoni pizza, but you may like ham and pineapple. I don't know. Pepperoni, you're right the first time. There you go. And similarly, if you were to look at domino statistics, that would be secondary because it's not your first fresh piece of research, isn't it? Absolutely. So this is a good example. So dominoes, I actually did a job for Pizza Hut, not dominoes, a few years ago. And they wanted me to do my primary research and ask people pretty much about how they ordered pizzas. It might sound ridiculous, but that's what they wanted to find out. But they already had a whole load of their own data from previous years about how people ordered pizza. So are they using the app or using the phone? Are they going in store? And so my research was primary, but their own stuff. I used that. It was their stuff. So it was secondary. So there we go. The pizza analogy coming back again. Wow. Never knew that about you, Ben. You worked doing research for Pizza Hut before becoming... And many others, far less interesting brands. I will bore you another day. Thank you. So primary and secondary, fairly straightforward. And then students probably will be introduced to the terms qualitative or quantitative. We often abbreviate them to qual and quant for ease. So do you want to explain that one? Yeah, sure. So quantitative is the easier one to explain because if you understand what quantitative is, qualitative is anything that isn't quantitative. Quantitative comes from the word quantity or quant, the amount of something. And when we talk about quantitative data, we're always thinking about numerical data. So it would be statistics and figures and numbers. So by process of elimination, qualitative data is anything that's non-numerical. And that could be words, pictures, feelings, emotions, anything that is non-numerical and that cannot be reduced down to statistics. Brilliant. Thank you. And I always remember my sociology teacher taught me this many years ago. Look at the fourth letter, L for letters and N for numbers. So you shouldn't get them muddled up. If it's written down, qualitative, quantitative, look at the fourth letter, L for letters, N for numbers. That's awesome. I'm stealing that myself. Yeah, yeah. Please do. And what about sort of if we link that to research methods? So what sort of methods would be considered quantitative? So again, we're thinking about any research method here that can generate numerical data for you. And I always sort of think that there are kind of three biggies. There are questionnaires, which you might produce yourself, so primary sort of data. We can talk about structured interviews, which are effectively like doing a questionnaire but face-to-face with someone where you ask them questions and you get back their answers and you jot them down. And then we've also got a secondary form of data, which is official statistics, which are any statistics or any research produced by the government. So that you might use some official statistics to look at, I don't know, let's say the cases of coronavirus at the moment, you can use official statistics, they're going to give you numbers. When you ask people questions through a questionnaire, their closed answers are or can be turned into numbers. So we'll go back to the example of what kind of toppings you like on pizzas. I could send the questionnaire out to 100 people and ask them to tick all of the toppings that they like. When I get my questionnaires back, I just count up how many people tip pepperoni, how many people picked ham, how many people picked mushrooms or whatever else. And then I've got my numbers from that. So questionnaires, structured interviews and official statistics would be my kind of free main quantitative methods. Brilliant. Excellent. Let's flip it over then. What kind of methods that produce qualitative data? So again, what we're thinking about here is what kind of methods are going to get you really rich, detailed, explained answers. And again, I'll pick free out. So we'll go for primary is unstructured interviews. So whilst the structured interview is very kind of formal and very kind of closed and very short answers and unstructured interviews very much like a conversation that can go in any direction that you want it to go to or even that your participant wants it to go to. In these interviews, you're going to basically have a conversation with somebody. How do you turn that into knowledge? You can't. It's qualitative data. The other thing we can talk about is potentially participant observations. So perhaps you are getting involved with a group you are studying and you are watching what they do. As you are watching it, the data you're getting is visual and you might also be making some notes as well, but you're not getting numbers. So you might be, you know, immersing yourself like James Patrick did in his study at Glasgow Gang Observed. He wasn't looking for how many people were in the gang or how many times they committed crimes. He was looking for what they did, why they did and how they did it. And those are the kind of questions that we can answer with qualitative methods. And the last thing is what I mentioned earlier, Matthew's diary, if he has one, personal documents would be another form of something we can gain quantitative data from. And again, that's secondary because it's someone else's property and you're using it for your purpose of research. So unstructured interviews and participant observation for primary qualitative methods and personal documents for secondary qualitative methods. Excellent. Good. Yeah, I was just thinking actually, there's a good piece of research by Valerie Hay called the company she keeps and it's called ethnographic research. It's all based on qualitative research and she asked school girls to keep a diary about their kind of interactions, their behaviors, their relationships, and she also interviewed them as well. So, very much a good piece of research on that qualitative side. Okay. Then we get a little bit more detail to them then. So you often require the sociological thought process to consider if something is positivist or interpretivist. So how do we know if a piece of research falls into one side or the other? I think you can say there's a number of ways you can tell this. A positivist is basically a social scientist and someone who wants to produce really robust, almost scientific data and they will conduct their sociological research in a similar way to a natural scientist would do in a lab. So things are very controlled as much as possible and really the key to understanding what positivist research is, is through the method that they will use. And again, what they're trying to do is generate numerical, statistical, quantitative data. So again, a positivist would be using methods like questionnaires, like potentially laboratory experiments and like structured interviews. And the reason for this is because it means that they're not allowing their own subjectivity and their own values to interfere with the research process. So when we think about a questionnaire, the respondent, the person who's answering the research, the respondent will answer the question that's asked of them and nothing more and nothing less. Whereas, if we're, and that's kind of what a positivist wants to get out, they want to get factual data, but an interpretivist is the other way around. An interpretivist will use qualitative data and look not to make statistical judgments about things, but just to get an understanding of something. So like I mentioned earlier, James Patrick's study on Glasgow gangs back in the, I think it was the 1960s, early 1970s. He was trying to understand what it was like to be in a gang. He wasn't looking to, as I say, quantify anything about that. And his interpretations were a key part of the research. And that's where we get that phrase interpretivist. So an interpretivist, we look to try and sort of interpret and analyze the meanings of what people are doing or what they're saying. Whereas a positivist literally records the data as they get it and that's the data. Does that kind of make sense? It does. Yeah. And you can also kind of link it to the size as well, can't you? So as a general rule of thumb, that the positivists tend to look at macro large scale data and the interpretivist more in depth micro data. That's correct. Again, the reason for that is, is positivists are generally trying to make representative almost cause and effect relationship statements about society. And positivists will also often start their research with an idea of what's going to come out of it. So they will start it with a hypothesis and their research will often look to try and prove or disprove their hypothesis. And by speaking to large samples of people, not only do you gain a representative view of ideally getting some findings that can be generalizable up based on the sample you've spoken to, but also to a wider group. But the idea as well is for them to have kind of sort of hard and fast robust data. Whereas interpretivists, as you say, we're looking at much smaller samples of people because interpretivists aren't looking to make generalizations. They're just looking to see or understand how or why a particular group behaves that they do. Right. Thank you, Ben. And you did mention the word representative, which we'll come back to if, if people are a little bit unclear on that, we will come back to that bit. So actually, we're going to, going to focus on those kind of three big words within research methods that every piece of research is analyzed on the terms of validity, reliability and representativeness. So we're going to take them one by one. So can we start with validity, Ben? How do you introduce what we mean by the term valid? Okay. So valid in a nutshell. When we talk about whether research or research findings are valid, what we mean is how true to life they are and does your find or do your findings reflect reality? Because when we think about valid basically means, is it truthful? Has the respondent given you an accurate, truthful picture of what they think or what they do? Or has it been possible for them to lie? So when we think about validity, we're thinking about getting to the root of the truth of what someone's talking about. And we tend to associate high validity with qualitative methods because what they do is they firstly, they take a bit longer, they go in depth a lot more on what someone's talking about. And they allow the both the researcher and the participant to steer the research in whatever way they want. This might be particularly valid because let's say you've come up with, you're doing some research on pizzas and say, sorry, I've got food on the mind at the moment. And you're, you're trying to understand what the ideal pizza is. What makes the ideal, the perfect pizza? Now, you might have your own idea in a questionnaire, you might say, what's the basic like? What's the source look like and all these topics do you have on it? And that's all well and good. But you're only ever going to get back the answers that you've basically given to people in the first place. And you've put these boxes that they can fill in and tick, yes, okay, I like it when there's a thick base or when it's got a tomato base. But when you do qualitative research, you can get the participant to guide that conversation or that research in any way they want. And there might well be things that you've never, ever thought about. So it could well be that the box that the pizza comes in is incredibly important or the source that you get with it. And that you would not get from a positivist data because that is valid. It's coming from the mouth of the respondent themselves and they're telling you what the truth is rather than you trying to input your ideas onto them. So when we talk about validity, we tend to think of qualitative methods having high levels of validity, but quantitative methods perhaps having lower levels of validity, but that is also cancelled out by something I'm sure you're going to mention in a minute, which is liability. Yeah, we come, we come back to that because a lot of students actually muddle the two don't they? But just thinking what, what a classically do sociologists have to think about in terms of validity? Where do they face issues where they're not getting back truthful answers? Okay, so some classic cases might be let's use questionnaires, but questionnaires are not a particularly valid method for several reasons. Number one, as I mentioned a moment ago, you're the answers and the closed answers that you're giving people may not actually fit with what they want to say, but they feel that they've got to put their answers into them anyway. The other thing is, is that you don't actually know who has filled that questionnaire out because this is something that the question has generally done self completion, so you'll do it on your own. So perhaps if I want to send one out to, if I sent the questionnaire out to you, perhaps your other half would have filled it out, I'm never going to know that's the case because, because I don't know. So the validity might be in question for that. And the other thing as well, it's, we tend to think of quantitative methods lacking validity because they include closed questions rather than open questions. So a closed question is where there's just potentially one answer or a select bunch of answers they could give, whereas an open question could be any number of answers. And it's those, the inclusion of open questions in interpretive as qualitative methods that allow participants to basically tell us what they want to rather than ask asking them what we think might be the answer. Yeah. Yeah. That's a really, really good point. And if the questions are a little bit too personal, that's an issue, isn't it? And you've also got this issue of what's called social desirability, where people answer how they think they should rather than how they actually feel. And, you know, the classic example, if you had a question there which said, are you a racist? Even if you were, you would have hoped that most people would tick no because they know that's socially unacceptable to put them. Absolutely. You've also got the other issue of interviewer bias as well. So potentially in qualitative research, it kind of works the other way. So you can lead respondents down a particular path to say something, as well as them thinking you want to hear something as well. So the bias thing can work both ways, both from a respondent and from a researcher as well. But of course, highly trained, well trained qualitative interviewer wouldn't do that. Great. Thank you, Ben. And now, you've already mentioned reliability. Be honest, Ben, how many students confuse validity and reliability when you're marking papers? A lot. And I would say that goes all the way up from, you know, year 10 starting at GCSE, which you'd kind of expect, but I've still seen it in year 13 papers as well. And I kind of understand it because, you know, we do throw an awful lot of terms at students, but to get reliability and validity round the right way is really quite crucial. And I think once students get it, you've basically got almost like a couple of paragraphs for pretty much every methods essay you'll ever be asked to write, because when we talk about methods, some methods are valid, some methods are reliable, and it tends to be that a method that's valid tends to be low in reliability, and a method that's high reliability tends to be low in validity. So if you understand those things, pretty much any method you get, you're going to be able to have something to say about it. Agreed. Yeah. And I can understand why people do get it confused, because the way in which we use the word reliable in everyday conversation is quite different from sociology, isn't it? Oh, absolutely. I mean, we use the term in everyday parlance reliable. It's just meaning, you know, you can count on it, that kind of thing, or it's good. Or it's even sometimes we even think of reliable as meaning truthful or as, you know, factual. So that we can really get confused with validity. But when it comes to sociology, when we're talking about the reliability of a method or the reliability of a data, effectively, what we're doing is we're asking the question, how easy would it be for another person, another researcher to go and repeat the process that you've just done with your participants that say you've just handed out a questionnaire, how easy would it be for them to repeat it and get the same data back? So effectively, when we talk about reliability, we're asking how watertight is your method that almost anybody can do it and the result you'll get back will be the same thing. So with reliability, I mentioned earlier that we tend to think of qualitative methods as having very low reliability. And this is because they're so subjective. So unstructured interviews, as I mentioned earlier, are like a conversation. It's very, very, very hard to repeat a conversation. Even if you listened to every single word that somebody said, you've then got to consider about the tone or the timing of when you ask particular questions. So it's very difficult to repeat that. However, if we consider a structured interview instead of an unstructured interview, where all of the questions are written down verbatim, so word for word, anybody theoretically should be able to repeat a structured interview because you're just asking exactly the same questions in exactly the same way again. So we tend to think of reliability as being, can this method be replicated or repeated by another sociologist and theoretically get the same results from the same sample? Is that a good representation of reliability? It definitely is. It definitely is. I was just wondering how students can remember the difference between the two because you will have seen this. A lot of students say, therefore, this method is valid and reliable, and they're not going to pick up marks because they haven't made a distinction. Have you got a way that students can remember the difference between the two? Do you know what? I'm looking now for the letters to see if there's any letters that might signify something, but do you know what, Matthew? I don't, to be honest, but I do hammer this home a lot. So I know at GCSE level, when I'm teaching methods, I introduce this in the first lesson and almost every lesson thereafter, I give my students some kind of question or quiz or, you know, tell me, is this reliable, is this valid? Because it is very confusing and even the top students can get this just mixed up in the heat of the moment when they're, you know, caning off an essay, finishing off their conclusion and they go to, like you say, they're not quite sure. So they hedge their bets and wrote, it's both reliable and valid. It's very unusual to get one method that is both reliable and valid. So by saying that, you're technically incorrect. Yeah. And actually, you should talk about them separately, shouldn't you? Because otherwise, that's where it does become a little bit confused. Oh, yeah. Absolutely. I mean, I know I'm guilty of this myself because I've said, well, if a method's reliable, it's not like to be valid. But I think when we're thinking about the issues of reliability and validity, as you say, take them in turn, talk about why something is reliable. What does it actually mean? Can you give an example? Like I gave the example of structured interviews and how you are literally just repeating word for word. That's the sort of thing that I suppose examiners will want to see, especially in those, you know, when we come to think of methods essays, but also methods in context as well, because you're going to need to think about not only the methods you use, but what information or what data are you going to elicit and get out of your respondents when you're doing whatever method is you're doing. And the way you do it is almost sometimes just as important as what you actually ask as well. Good. Yeah. I remember my sociology teacher putting it in some quite blunt terms. They said, right, think of a car. What do we mean by a reliable car? Well, reliable car means it starts every time each morning. It doesn't mean, you know, the car's not going to lie to you. It means it can not fit. And the other one that really stuck in my mind, so it is a good example, my teacher said, think of a contraceptive pill. Every time you take it every morning, it better bloody work, it better be reliable. It's a good way to remember it though, right? I'll try that one. I'll try it. And there's one other term then that we need to cover. So we've talked about validity, we talked about reliability, and then we have representativeness. So do you want to go with that one then? Yeah. It's one of those really sort of hand-fisted long words that we use in sociology, representativeness. It basically means to what extent does your data and the findings that you've got from your sample to what extent is that are we able to almost blow that up and have that generalize the whole population or of a wider population. So if we're thinking about, let's say, pizza toppings, sorry, I've really gone on a ramp this one today, haven't I? But let's say you're doing pizza toppings. To make my research representative, if I only spoke to men in my sample, that would not be representative because the UK is made up of men and women, roughly 50/50. So I need to make sure that my sample has roughly equal numbers of men and women. I'd probably want to make sure that in that sample, by sample, I mean a smaller group who you conduct your research with, I'd probably want to make sure that I've got different age ranges, different ethnicities, different people from around the country, people who buy pizza, people who perhaps don't buy pizza. So I've got to think of all of the types of people if I want to make it representative that are going to be allowing me to say, ah, from the 2000 people that we spoke to, we can confidently make a generalization to say that, yes, pepperoni passion or pepperoni other pizzas are available, pepperoni is the nation's favourite, that kind of thing. Yeah. So is it general rule then that the bigger the sample, the more representative it's probably going to be? Yeah, absolutely. I mean, the ideal case, I would say this to students. The ideal case is on pretty much every kind of quantitative piece of research, you would interview or speak to or conversely with every person in the country. But that just ain't possible because of practicalities and time and cost and all these kind of things. So you've got to think, who can I speak to that would best represent the wider population? As I say, if you get a kind of representation from different kinds of groups in society, then you're not going to be too far off, but definitely always quantitative and qualitative often doesn't seek to be representative. I think this is one of the things that students tend to forget as well, that it's not necessarily a weakness of qualitative research that it isn't representative because it's not trying to be representative in the first place, but by qualitative research, we're just trying to understand why people do things. And I think sometimes I put it in those terms, quantitative will answer you the questions like what, how much, qualitative answers to the questions like how and why those kind of things. I think that's a really good point to make. So if you're just doing one really long in-depth interview, you're not trying to be representative, are you? You're just trying to get as much information out of that individual as possible. Oh, absolutely. And, you know, it's very common to, you know, on a qualitative level to be working with samples of below 50, sometimes on a quantitative level, we tend to want samples to be somewhere in the region of around 2000 upwards in order for us to make statistically significant inferences about representativeness. So for quant research, generally, it needs to be about 2000 plus, but for qualitative it could be any number. It tends to be about 30, 40, something like that. Yeah. And I think there's just one other thing that I'd really recommend to students is that a lot of students define the word using the same word. A lot of students say representative is where it represents a group. So maybe use a word like reflect or something like that instead, just because, you know, sometimes the examiners are a little bit picky about using the same word to define a term, which makes sense. So we're going to conclude this one. This is going to be part one of an introduction to methods. In the future, we are going to delve a little bit deeper and look at some more complex terms as well. And Ben, have you got any other advice for students to help to start to understand research methods and get a basic concept or understanding of it? Yeah. Do you know what I'd say is that I think that when you first start a topic of research methods, it can be very daunting because, as we mentioned right at the top of this, for one, it's not a particularly tangible subject. But the other thing is that, I guess, you've got to be aware that these things aren't going away. And if you like sociology and if you are thinking of studying or taking exams and you're going to have to talk about methods, I think one of the things I can suggest is breaking down because in any exam with its GCSE, A level or undergraduate, even beyond that, and I've done a postgraduate in research methods, which was very boring. But what you have to do is understand the strengths and the weaknesses of these different methods. And I'm sure that you will talk about this in a future episode, Matthew, but thinking about how you can group those different factors up. And so we often talk about practical issues, ethical issues, and theoretical issues. And once you start to know that actually those are the same issues, it's just when you talk about a different method, you're going to talk about it in a slightly different way. So things like knowing what reliability and validity are, because those things will always be something you can say about methods. So I highly recommend not freaking out, first of all, and secondly, see if you can try and group up some of the positives and negatives of the methods you look at, because there will be overlaps with them for sure. And so it feels like there's a load to learn. But once you realize there's a lot of overlap, you will only need to learn some of the key things and you can kind of fill in the gaps a bit when you get a bit closer towards it. Yeah, I think that's good advice and pets, practical, ethical, theoretical. And also, maybe read or look at or watch some videos of some research first before you start to dissect as to whether it's primary, secondary, qual, quant, whatever. I think just enjoy the sociology first and then do the grouping bit afterwards. Yeah, absolutely. And I think one of the exciting things about this and one of the things that used to excite me about sociology is that you'd actually, this is a job, someone will pay you to go out and talk to people for a living, but you have to be really good at it. So if you are someone who wants to do that job when you're a bit older, then get good at methods and you will not be too far away from actually being paid to talk to people for a living. How wonderful would that be? Oh, hang on a minute, we're teachers, aren't we? But as you said, you do have to be good at your methods to be able to do that job. You do. You do. You do. Great. And I really appreciate you coming to help out and I might drag you back for a future episode on methods if you don't mind. I'd love to, mate. Give me a shout anytime. Brilliant. Thank you very much, Ben. Take care. Thanks, Matthew. Cheers. Bye. If you are a regular listener to the sociology show, then you could help with the cost of promoting and hosting the podcast. If you can spare even a small amount, then you can donate on the gofundme.com website by searching for the sociology show. There is no obligation, of course, and all future downloads will continue to be free. A huge thank you to all those that have already donated. Your kind gesture will help to continue keep the show going and growing. Best wishes and keep enjoying the show. Thank you for taking the time to listen to the podcast. If you would like to contact the show or be interviewed, then please email the sociology show podcast at gmail.com.