WEBVTT Kind: captions; language: en-us NOTE Treffsikkerhet: 87% (HØY) 00:00:01.900 --> 00:00:12.200 Dear students, welcome back to our sixth lecture of this qualitative research module. Today we're 00:00:12.200 --> 00:00:21.800 gonna have quite a longer video introduction because we are going to discuss a very relevant topic 00:00:21.800 --> 00:00:30.450 within the topic of qualitative research. Today we are going to discuss a little bit about how NOTE Treffsikkerhet: 84% (HØY) 00:00:30.450 --> 00:00:36.000 to analyze the qualitative data. NOTE Treffsikkerhet: 86% (HØY) 00:00:36.400 --> 00:00:47.500 So first, what you will find on canvas has resources for for today topic. I ask you kindly 00:00:47.500 --> 00:00:53.600 to read chapter 20 of the handbook data; "analysis and interpretation: emergent issue in linking 00:00:53.600 --> 00:01:00.800 qualitative and quantitative evidence". And then I provided you with an optional reading which is 00:01:00.800 --> 00:01:05.500 about the different ways in which you can NOTE Treffsikkerhet: 88% (HØY) 00:01:05.500 --> 00:01:18.800 verify the reliability of your qualitative analysis. So let's start with a simple game. 00:01:18.800 --> 00:01:26.900 So, in this picture, how many fingers do you see? Of course, probably your first answer will 00:01:26.900 --> 00:01:34.550 be 5, but then let me ask you. And now how many fingers do you see? NOTE Treffsikkerhet: 83% (HØY) 00:01:34.550 --> 00:01:43.000 You will probably answer five. So what is the difference between this 00:01:43.000 --> 00:01:52.000 two sets of five fingers? What is the transformation that happened from the first 00:01:52.000 --> 00:02:01.600 slide to the second? Well, of course, you may say that in the first slide you could see the 00:02:01.600 --> 00:02:05.000 hand and the five fingers. NOTE Treffsikkerhet: 91% (HØY) 00:02:05.000 --> 00:02:16.900 Here you can see only the five fingers. In this transformation something has been lost. But 00:02:16.900 --> 00:02:25.150 first of all, let me explain what is this transformation. So we have taken full phenomena. So the 00:02:25.150 --> 00:02:34.950 initial picture and we have applied and created or overimposed a homogeneous NOTE Treffsikkerhet: 78% (HØY) 00:02:34.950 --> 00:02:42.800 classes or categories. So, first we had a hand. Here we have five fingers because we have applied 00:02:42.800 --> 00:02:50.300 the categories finger. What are the consequences of this categorization? Well, there are many. First 00:02:50.300 --> 00:02:57.000 of all, we have created a category fingers which is by definition homogeneous. 00:02:57.000 --> 00:03:05.000 So we have lost the specificity of each figure. Indeed, NOTE Treffsikkerhet: 89% (HØY) 00:03:05.000 --> 00:03:11.400 in our hand each finger has some characteristics and features and cannot do some things. 00:03:11.400 --> 00:03:22.400 If we classify them all as fingers, of course, we achieve some kind of modernity that 00:03:22.400 --> 00:03:30.400 allows us for instance, to treat this example of different categories statistically. But we have 00:03:30.400 --> 00:03:34.950 also lost another very important feature. We have lost the NOTE Treffsikkerhet: 85% (HØY) 00:03:34.950 --> 00:03:42.600 systemic organization of the five fingers. So the fingers are connected together in a system that we 00:03:42.600 --> 00:03:52.300 call the hand. And again we have lost the perspective of this system. So we must be aware 00:03:52.300 --> 00:03:58.800 that each act of categorisation is at the same time a way of organizing the phenomenon, but at the 00:03:58.800 --> 00:04:05.000 sametime, on the other hand, is a way of losing some features and this is a very common operation NOTE Treffsikkerhet: 72% (MEDIUM) 00:04:05.000 --> 00:04:23.600 in social sciences. So now we enter in the field of elaborating from our data collection to 00:04:23.600 --> 00:04:34.900 our data analysis. So far we have discussed about methods to cut out a portion of NOTE Treffsikkerhet: 78% (HØY) 00:04:34.900 --> 00:04:42.100 of the phenomenon to capture some dimensions of the phenomenon, as we see in every action of 00:04:42.100 --> 00:04:48.900 capturing is also an action of losing something else. So it's a compromise and we have to 00:04:48.900 --> 00:04:57.600 make always some choices. So, why we do decide to collect data and analyze data in a certain 00:04:57.600 --> 00:05:04.000 way, and not in another. Well, I will say that our guiding principle is always our research 00:05:04.000 --> 00:05:05.050 question. NOTE Treffsikkerhet: 79% (HØY) 00:05:05.050 --> 00:05:16.600 So, if you remember the mythological cycle, our starting point is our research question, our 00:05:16.600 --> 00:05:25.800 curiosity and our research question can of course be of very different kinds. Our research question 00:05:25.800 --> 00:05:34.900 can, for instance, be about different phenomena so we can have a question about people, about NOTE Treffsikkerhet: 84% (HØY) 00:05:34.900 --> 00:05:42.750 cases or about objects. So for instance, we can have a research question about children with autism 00:05:42.750 --> 00:05:51.000 about a single case of that specific child with autism and his developmental trajectory or we can 00:05:51.000 --> 00:05:57.800 have a research question which is about autism as a concept or as an object. Our research 00:05:57.800 --> 00:06:04.500 question can involve different levels of analysis. We can have a research question that 00:06:04.500 --> 00:06:05.000 looks at NOTE Treffsikkerhet: 69% (MEDIUM) 00:06:05.000 --> 00:06:12.800 micro phenomena, like for instance child-mother interaction, verbal interaction. We can have a 00:06:12.800 --> 00:06:20.700 research question about a meso Level phenomenon. Like for instance how a classroom works or we can 00:06:20.700 --> 00:06:28.800 have how school works or we can have macro levels research question, how the society views 00:06:28.800 --> 00:06:34.850 disability for instance, but we can also have a research question that works across the NOTE Treffsikkerhet: 86% (HØY) 00:06:34.850 --> 00:06:43.800 levels. In the chapter 20 of the handbook you will find a very rich discussion about this. 00:06:44.400 --> 00:06:55.300 Another very important aspect of our research question is what kind of concept we use when we asked 00:06:55.300 --> 00:07:02.500 our research question. For instance our research question can be about a concept. Again, we want to 00:07:02.500 --> 00:07:04.850 understand NOTE Treffsikkerhet: 76% (HØY) 00:07:04.850 --> 00:07:13.800 how the concept of autism developed over time, but in our research question can also be about 00:07:13.800 --> 00:07:25.700 variables. Can be about, for instance, the factors that influence autism. What we must be very 00:07:25.700 --> 00:07:34.850 careful about is to not mix the concepts in a wrong way. So for instance, if we ask, if in our NOTE Treffsikkerhet: 83% (HØY) 00:07:34.850 --> 00:07:41.700 research question we use variables as a concept. Of course we cannot do a qualitative 00:07:41.700 --> 00:07:49.800 research, because variables are about quantitative research. Another important dimension is what kind 00:07:49.800 --> 00:07:58.200 of phenomena we are looking for in our research question. So for instance, we can look for patterns, 00:07:58.200 --> 00:08:04.450 regularities, we can ask about the effects of some NOTE Treffsikkerhet: 90% (HØY) 00:08:04.450 --> 00:08:13.500 variable on an another variable, it can be a research question about relationship, can be a research 00:08:13.500 --> 00:08:24.000 question about a case that is perfectly representing a more general category, it can be a 00:08:24.000 --> 00:08:31.000 research question about meaning making so how people tell the story of their life, for instance. But also 00:08:31.000 --> 00:08:34.799 our research question has a NOTE Treffsikkerhet: 78% (HØY) 00:08:34.799 --> 00:08:43.299 temporal dimension. So we can study phenomena that happen here and now or can study phenomena that 00:08:43.299 --> 00:08:52.600 develop over time. So we have also to locate our research question in space and time, somehow. And 00:08:52.600 --> 00:09:01.000 this work on research question will also help us to formulate our methodology. NOTE Treffsikkerhet: 91% (HØY) 00:09:02.900 --> 00:09:19.800 So what is that data analysis for? The idea is that starting from your research question you will 00:09:20.200 --> 00:09:29.600 consider some different theoretical aspects that will lead you to observe the phenomenon in a 00:09:29.600 --> 00:09:30.250 certain way, NOTE Treffsikkerhet: 68% (MEDIUM) 00:09:30.250 --> 00:09:36.700 through a certain method. But then when it comes to the analysis of your data, your work 00:09:36.700 --> 00:09:44.950 will be to create new knowledge by relating the data with the theory with your research 00:09:44.950 --> 00:09:53.700 question. So basically data analysis consists of trying to answer your research question by 00:09:53.700 --> 00:09:59.600 looking at the data through the lenses of your theory, but what is very important is that this 00:09:59.600 --> 00:10:00.300 process in NOTE Treffsikkerhet: 85% (HØY) 00:10:00.300 --> 00:10:07.400 qualitative research is always circular. So the more you learn from your data, the more your 00:10:07.400 --> 00:10:15.650 research question also will develop your understanding of your criticism of theory will also 00:10:15.650 --> 00:10:22.700 develop. And with this new knowledge you can go back to the data and learn something more about the 00:10:22.700 --> 00:10:25.900 data. This is a circular process. NOTE Treffsikkerhet: 73% (MEDIUM) 00:10:27.600 --> 00:10:37.700 So well, let's go back to the research question. And the way the research question is related to the 00:10:37.700 --> 00:10:47.550 data. The first thing we have to learn is that if we ask a question, we also have to 00:10:47.550 --> 00:10:53.900 be able to answer this question. That's why we do research. So we don't do empirical research, 00:10:53.900 --> 00:10:55.850 qualitative or quantitative, NOTE Treffsikkerhet: 77% (HØY) 00:10:55.850 --> 00:11:02.100 to answer impossible questions. Like what is the meaning of my life? This is not a question that can 00:11:02.100 --> 00:11:10.200 be answered empirically, maybe philosophically but not empirically for sure. So no method and no 00:11:10.200 --> 00:11:20.200 analysis, can answer this kind of question. So the first effort that we have to make 00:11:20.200 --> 00:11:25.900 is to try to formulate our research question in a way that is answerable. NOTE Treffsikkerhet: 86% (HØY) 00:11:25.900 --> 00:11:37.349 The second important thing that you have to keep in mind is that when you study a phenomenon 00:11:37.349 --> 00:11:53.200 you must define the minimal unit of analysis. So it means what I actually going to study. 00:11:53.200 --> 00:11:56.800 Just imagine that you want to study a NOTE Treffsikkerhet: 86% (HØY) 00:11:56.800 --> 00:12:12.900 text. So a text is a complex phenomenon, which is full of meanings of words. In 00:12:12.900 --> 00:12:19.600 order to study it empirically we have to decide which level we are going to study. Are we going to 00:12:19.600 --> 00:12:26.100 study, for instance, the way the words are combined together or are we interested in study NOTE Treffsikkerhet: 76% (HØY) 00:12:26.100 --> 00:12:34.300 how people make express meaning in the sentences? So this will constitute different unit of 00:12:34.300 --> 00:12:43.400 analysis. When we choose the unit of analysis we can have very different types. So for instance, we 00:12:43.400 --> 00:12:51.500 can study people, our unit of analysis can be people. So I'm interested in studying children in a 00:12:51.500 --> 00:12:56.000 specific age or NOTE Treffsikkerhet: 76% (HØY) 00:12:56.000 --> 00:13:07.100 people with disabilities in a specific workplace or we can want to study a process. So our 00:13:07.100 --> 00:13:13.700 unit of analysis would be the process itself, like the process of learning and we can study the 00:13:13.700 --> 00:13:20.900 the process of learning under different conditions for existence, or we can decide that 00:13:20.900 --> 00:13:27.000 our unit of analysis is the rule of a single case, like one single school. NOTE Treffsikkerhet: 87% (HØY) 00:13:27.300 --> 00:13:38.250 Or we can choose this unit of analysis is also some objects; like an interview, a Facebook post. So 00:13:38.250 --> 00:13:49.200 these are the conceptual units that we choose to define, and to collect empirical data 00:13:49.200 --> 00:13:56.000 about. But then not all, in order to study NOTE Treffsikkerhet: 87% (HØY) 00:13:56.000 --> 00:14:05.800 empirically a phenomenon we need to be able to observe this phenomenon acting in real life. So we 00:14:05.800 --> 00:14:19.700 have also to choose the kind of observable event that we are going to collect in real life. So 00:14:19.700 --> 00:14:26.050 we have to define also a unit of observation, which is the minimal unit of data NOTE Treffsikkerhet: 78% (HØY) 00:14:26.050 --> 00:14:35.300 that we are going to collect and the unit of observation must never be confused with the 00:14:35.300 --> 00:14:42.700 unit of analysis. Why? Because for instance, if our unit of analysis are all the children at a 00:14:42.700 --> 00:14:51.500 certain age or in a certain school grade, we may want to observe different aspect of this child, of this 00:14:51.500 --> 00:14:56.400 children. Like for instance, we can want to observe NOTE Treffsikkerhet: 78% (HØY) 00:14:56.800 --> 00:15:07.100 their verbal expression or you want to observe their micro interaction between themselves, between 00:15:07.100 --> 00:15:13.800 them and the schoolmates. So, our unit of observation is not all completely overlapping 00:15:13.800 --> 00:15:19.900 with unit of analysis. They are different concept and we don't have to confuse them. NOTE Treffsikkerhet: 85% (HØY) 00:15:20.900 --> 00:15:36.650 Now what I really just provide you is some examples of different strategies of data analysis. 00:15:36.650 --> 00:15:45.600 How we can apply the strategies to our data. NOTE Treffsikkerhet: 81% (HØY) 00:15:45.700 --> 00:15:56.500 And as we have discussed in the previous lectures, in qualitative research our 00:15:56.500 --> 00:16:03.500 data can be multimodal. So the strategies of analysis that I'm just presenting to you can 00:16:03.500 --> 00:16:09.300 be applied to different kind of data; at verbal interaction, transcription of interviews, visual 00:16:09.300 --> 00:16:15.650 image, and I don't know and observation of gestures, for instance. NOTE Treffsikkerhet: 82% (HØY) 00:16:15.650 --> 00:16:26.000 So the first strategy is the thematic analysis. So basically the principal of thematic 00:16:26.000 --> 00:16:34.500 analysis at that you consider your unit of observation and you apply to your units of observation 00:16:34.500 --> 00:16:46.000 some categories that we call code. So you code your unit of analysis with categories and NOTE Treffsikkerhet: 81% (HØY) 00:16:47.600 --> 00:16:58.600 the sum of the unit of analysis that are coded with the same category will give you some information 00:16:58.600 --> 00:17:07.099 about the phenomenon. So the most common way of doing this categorization is called thematic 00:17:07.099 --> 00:17:15.500 analysis. So you start from some categories, some codes, that for instance, emerge from NOTE Treffsikkerhet: 81% (HØY) 00:17:15.500 --> 00:17:22.300 your theory, from the theory that you use. Like, I want to observe classroom behavior, children 00:17:22.300 --> 00:17:30.700 behavior. Well, I can start already from some codes that belonged to the theory I choose. Like 00:17:30.700 --> 00:17:38.300 for instance, I want to study the code, the kind of interaction like cooperation or conflict between 00:17:38.300 --> 00:17:45.400 the children. So I will code the different behavior of NOTE Treffsikkerhet: 83% (HØY) 00:17:45.400 --> 00:17:55.650 children with two categories: conflict and cooperation. But during the process of coding 00:17:55.650 --> 00:18:04.449 and analyzing data I will probably create new categories that comes directly from the data. For instance, 00:18:04.449 --> 00:18:11.900 I can observe that the children they, some children thet tend to avoid either conflict or 00:18:11.900 --> 00:18:15.300 cooperation. They avoid interaction with other children. Well, NOTE Treffsikkerhet: 85% (HØY) 00:18:15.300 --> 00:18:22.850 I can create a third category, which is avoidance. And then start coding this behavior as avoidance. 00:18:22.850 --> 00:18:34.699 Well, the problem of course, with this kind of qualitative approach is that my cording, so 00:18:34.699 --> 00:18:42.400 the way I attribute a piece of data to a category can be somehow subjective. So we need 00:18:42.400 --> 00:18:45.400 somehow to improve the validity of NOTE Treffsikkerhet: 82% (HØY) 00:18:45.400 --> 00:18:50.900 of this coding. And there are different strategies that you can find. This is why you have the 00:18:50.900 --> 00:19:01.700 optional reading about intercode reliability. Intercode reliability means exactly that I 00:19:01.700 --> 00:19:08.900 verified that my coding is not subjective or arbitrary, but it is really represent 00:19:08.900 --> 00:19:15.650 objectively or as objectively as possible the data. Here you can find two NOTE Treffsikkerhet: 78% (HØY) 00:19:15.650 --> 00:19:23.100 possible strategies to improve the validity of coding. The first one is called 00:19:23.100 --> 00:19:30.200 independent coding in which you ask two people to code the same data, the same piece of data, 00:19:30.200 --> 00:19:40.100 independently and afterwards you compare their cooling. The more their coding is similar. The 00:19:40.100 --> 00:19:45.350 more there is an agreement and it means that the categories that I do you are using somehow NOTE Treffsikkerhet: 78% (HØY) 00:19:45.350 --> 00:19:55.450 are really representing the data. The second strategy is called intercode 00:19:55.450 --> 00:20:05.900 agreement. So it means that the two coders they could the data together and every time they disagree 00:20:05.900 --> 00:20:15.500 on a coding, they discuss this disagreement until they find the best solution to code this data. So, NOTE Treffsikkerhet: 78% (HØY) 00:20:15.500 --> 00:20:23.100 in both cases, it is about different people coding the same data, but the difference is in the 00:20:23.100 --> 00:20:35.100 way they define the agreement on single coding. Now, another strategy of data analysis is also 00:20:35.100 --> 00:20:45.550 based on categorisation and coding process, but is a form of mixed methods. So NOTE Treffsikkerhet: 77% (HØY) 00:20:45.550 --> 00:20:54.900 while you can, once you have categorized and coded some piece of data you can 00:20:54.900 --> 00:21:03.850 use them in different ways. You can use them in completely qualitative way. So, you can try to 00:21:03.850 --> 00:21:14.000 interpret the relationship between the categories or you can also transform them into a matrix of 00:21:14.000 --> 00:21:15.350 frequencies. So, NOTE Treffsikkerhet: 89% (HØY) 00:21:15.350 --> 00:21:25.600 my research question, can be "What is the meaning of the data I have?" So the categorisation 00:21:25.600 --> 00:21:31.800 I do helps me to understand the meaning but you can also have different kind of 00:21:31.800 --> 00:21:38.800 research question. Like for instance, how many time children cooperate or conflict 00:21:38.800 --> 00:21:45.400 in classroom? So as you may see, how many is a quantitative question. So in order to NOTE Treffsikkerhet: 87% (HØY) 00:21:45.400 --> 00:21:53.400 answer this question I need to convert my qualitative data in frequencies, that allow me 00:21:53.400 --> 00:21:57.900 to perform some quantitative analysis. NOTE Treffsikkerhet: 68% (MEDIUM) 00:21:58.400 --> 00:22:10.300 Another strategy of data analysis is called interpretive analysis or hermeneutical analysis. So as 00:22:10.300 --> 00:22:19.500 we already discussed, the process of qualitative research is about making visible. Here 00:22:19.500 --> 00:22:27.050 we want to focus on the way the participant is NOTE Treffsikkerhet: 86% (HØY) 00:22:27.050 --> 00:22:36.300 making meaning of a specific phenomenon. But here we also want to get the first person 00:22:36.300 --> 00:22:45.200 perspective. So we want to understand the person, the participant, experience from her home 00:22:45.200 --> 00:22:54.700 perspective. So, it's not our interpretation overlapping. And in this way, in this case, we use 00:22:54.700 --> 00:22:57.000 a different approach. NOTE Treffsikkerhet: 77% (HØY) 00:22:57.400 --> 00:23:07.100 So, the idea is that when we collect the data, or when, for instance, when we interview the 00:23:07.100 --> 00:23:17.200 participant, what we observe is the participant which is acting or 00:23:17.200 --> 00:23:22.900 making meaning of a situation from her own perspective. NOTE Treffsikkerhet: 91% (HØY) 00:23:23.500 --> 00:23:34.800 So when I record this data, I am already performing an interpretation on the interpretation of the 00:23:34.800 --> 00:23:45.600 situation by the participant, because when I collect the data I am actually making a cut on the 00:23:45.600 --> 00:23:52.950 experience. I'm imposing already my perspective on the experience of the participant. NOTE Treffsikkerhet: 73% (MEDIUM) 00:23:52.950 --> 00:24:08.300 So this is already somehow an act of interpretation. Then when I write down my 00:24:08.300 --> 00:24:18.200 interpretation, what I'm doing is I'm creating an account, which is of course my perspective on 00:24:18.200 --> 00:24:23.150 the participant perspective and this account becomes NOTE Treffsikkerhet: 86% (HØY) 00:24:23.150 --> 00:24:31.500 it self the object, the possible object, of a new interpretation. So, here I have two choices. I can go 00:24:31.500 --> 00:24:41.450 back to the participant and give to him my interpretation of his experience and get feedback. So 00:24:41.450 --> 00:24:48.500 I will have an interpretation of the interpretation of the interpretation, or I can share my 00:24:48.500 --> 00:24:53.150 interpretation, not with the participants, but with other researchers. NOTE Treffsikkerhet: 90% (HØY) 00:24:53.150 --> 00:25:01.050 So and this circle in which every interpretation becomes the starting point for a new interpretation 00:25:01.050 --> 00:25:13.100 is the way through which we increase our understanding of the phenomenon. Of course, this as you may 00:25:13.100 --> 00:25:21.449 may realize, this can be a very democratic process because we share our interpretation looking 00:25:21.449 --> 00:25:23.150 for a common NOTE Treffsikkerhet: 85% (HØY) 00:25:23.150 --> 00:25:31.800 creation of knowledge or can be more authoritarian, so to speak process. In which the researcher 00:25:31.800 --> 00:25:41.700 interpretation is somehow stronger than participant. So this implies a very, very careful 00:25:41.700 --> 00:25:45.500 reflexive process from the written by the research. NOTE Treffsikkerhet: 83% (HØY) 00:25:45.500 --> 00:25:52.550 Now, last example of a strategy of analysis. NOTE Treffsikkerhet: 91% (HØY) 00:25:52.550 --> 00:26:02.900 Is especially relevant when you have multimodal data. When you have different kinds of data that 00:26:02.900 --> 00:26:22.550 goes from the images to gestures to verbal data. For instance, you may have images which are NOTE Treffsikkerhet: 89% (HØY) 00:26:22.550 --> 00:26:32.700 by definition polysemic. So, while words and numbers and figures are themselves polysemous. So 00:26:32.700 --> 00:26:39.100 they are open to different interpretation because they have range of different meanings. Images are 00:26:39.100 --> 00:26:47.400 even more open in the interpretation. So the multi modal analysis is when you combine the 00:26:47.400 --> 00:26:50.300 different kind of data in order NOTE Treffsikkerhet: 91% (HØY) 00:26:50.300 --> 00:27:01.600 to get an understanding that feeds into each other. One example is analyzing 00:27:01.600 --> 00:27:08.100 photovoice. As you may remember in photovoice you have the participant producing images. You have 00:27:08.100 --> 00:27:14.800 the participant and producing captions for the images. But you have also, the process of 00:27:14.800 --> 00:27:20.150 interpretation of these images together between the participant and the NOTE Treffsikkerhet: 91% (HØY) 00:27:20.150 --> 00:27:26.200 researcher. So here you have different modalities of data that feed into each other to produce a 00:27:26.200 --> 00:27:34.000 richer interpretation and richer understanding of the phenomenon. So, these are just 00:27:34.000 --> 00:27:44.300 examples of possible analysis of strategies. Now, I want to go back to the reading today. You 00:27:44.300 --> 00:27:50.100 will find some of these issues discussed in the chapter about data analysis NOTE Treffsikkerhet: 91% (HØY) 00:27:50.100 --> 00:27:58.400 interpretation. What are the key issues, the key concept in this chapter? The first one is 00:27:58.400 --> 00:28:05.600 the chapter actually discuss the limits and advantages of qualitative and quantitative methods of 00:28:05.600 --> 00:28:14.300 micro and macro levels of analysis. But especially the potentialities of using them in 00:28:14.300 --> 00:28:20.150 combination. So there is no such of a position, like quantitative versus qualitative. NOTE Treffsikkerhet: 86% (HØY) 00:28:20.150 --> 00:28:27.000 The chapter clearly states that they are complementary perspectives on human 00:28:27.000 --> 00:28:35.700 phenomena. We must feel very free to use them in combination. The second very important claim in 00:28:35.700 --> 00:28:43.300 the chapter is that evidence is always theory driven. If you remember the methodological cycle, 00:28:43.300 --> 00:28:49.850 the theory is a tool, is a pair of glasses, that we use to capture NOTE Treffsikkerhet: 77% (HØY) 00:28:49.850 --> 00:28:58.850 some aspect of the phenomena and the theory is also leading us to choose different strategy of data 00:28:58.850 --> 00:29:06.900 collection, different methods and different strategies of data analysis. Then the chapter provides 00:29:06.900 --> 00:29:17.400 some interesting examples of mixed methods studies in different fields, but I think that maybe the 00:29:17.400 --> 00:29:20.100 most important concept in NOTE Treffsikkerhet: 88% (HØY) 00:29:20.100 --> 00:29:34.200 the chapter is the idea that when we apply quantitative analysis, we treat the results as if they 00:29:34.200 --> 00:29:43.800 were causal explanations. So when we find some relationship between variables we 00:29:43.800 --> 00:29:50.050 treat them as a causal explanation, but actually they just show that there is something going NOTE Treffsikkerhet: 91% (HØY) 00:29:50.050 --> 00:29:57.400 going on. But unless we understand the underlying process, you're not really understand any causal 00:29:57.400 --> 00:30:01.199 explanation, or any relationship. NOTE Treffsikkerhet: 81% (HØY) 00:30:01.199 --> 00:30:13.900 Well, that was all for today. I recommend you to read carefully the chapter and I will see you 00:30:13.900 --> 00:30:21.200 back on question and reflection time. Bye.