 Reading research articles can be really hard if you are new to the field. In another video I talked about the importance of understanding the concepts and understanding the theories that link those concepts when you read an article. In this video I'll take a look at how you can use the structure of a research article to guide your reading and maybe the structure will help you to be more efficient so that you don't have to read every part of the article to understand it and also more effective so that you would understand the ideas on a deeper level than just skimming through the article would give you. Articles are not read like novels, so novels would be read from the beginning to the end and you would process progress sequentially. Articles are not read that way, so if you are trying to read it from the beginning to the end, you are doing it wrong. When I read an article, depending of course what's the purpose of the reading, I might spend 15 minutes on the article to get the main point and then move to the next article. Or if I'm evaluating the article, if it's done right or not, then I might spend three or four hours on the article. If I just want to know what the article is about, I might spend two minutes on it. So how much time you read the article and which part of the article you read depends on what is the purpose of you reading that article. This is a typical structure of a research article. So we have an introduction which tells what is the article about, why it might be interesting, why it might be important, and typically you also find the research question here. Then we have a literature review or some kind of theory part which typically contains citations to previous research, but often also expresses new ideas. Like we have the article by Sapienza that I use as an example, and in that article the authors claim that early international expansion increases the risk of failure, and it also increases the probability of growth, and the theory part of the article explains the authors reasoning on why they think that is the case. The theory part contains the explanation of the causal mechanism. Then we have data and methods, and I will not go in detail what that includes because it depends a lot on the type of article, but typically you would have an explanation of what data were used, what companies, what people, was it a quantitative study, how the data were analyzed, and so on. Then we have the results section where we explain how we think that the data either support or refute some idea that we present in the article. The discussion section then starts from takes the results and then explains what do those results mean. So why would somebody care about those results? Do they have some policy implications? Do they have some implications for research, and so on. And then we have conclusions, which is typically sort just focuses on the interesting and importance of the article, and typically one page or so just telling what to do next. A former colleague of mine, Mikri Askelainen, told me that this structure follows what he calls master's thesis wave, and this also applies to master's thesis, of course, as the name implies. The idea is that in the introduction, you start from a very high level of abstraction, you focus on big picture, you have high level concepts, and then when you go to the theory and the literature review, then you start to explain the causal mechanism, and then you go into the nitty-gritty detail on what exactly happens and why, the same with methods, there's lots of detail, and then when you start discussing the results and the particle with a good conclusion, you increase the level of abstraction as well. So why is understanding this useful? It is useful because sometimes understanding the big picture is enough. So if you just don't understand the big picture of what the article says, then typically reading the introduction, discussing the conclusions might be sufficient if the article is well written. Now, let's take a look at how we actually read articles. So reading an article, how you do it is not from beginning to the end, but instead you start, you are kind of like asking the article questions. So you are looking at answers to questions from the article, and there are some steps. The first step that you need to do is the first question that you need to ask is what is the topic and what is the research question? Typically the answer is found in the introduction. Once you have that covered, so you read the introduction, you might need to read it many times over if it's poorly written, but typically one read is enough to understand this topic. So this question, so you don't move on reading before you understand what is the topic and what is the research question. Reading beyond that, the introduction, if you don't understand those two things would be a big waste of time. The next thing is identifying the key concepts and the theory that links those concepts. I have another video about the concepts and theories, but ideally if there is this kind of boxes and arrows, diagrams, the boxes are the concepts, the arrows are the causal relationships that the article proposes or tests, then you would have to understand what are the boxes, so how do we define fungibility, for example? What does it mean that there is manager experience? What is the initialization of internalization process and so on? So you look for the diagram, you look for the boxes and explanations for the boxes. So you identify the concepts and you identify the propositions that explain the causal relationships between those concepts. Okay, once you've understood the theory, the next question that you ask is what are the data? So what research design and data are used? I'm not talking about that you would go and look like, so what flavor of a grounded theory is being applied or what kind of regression analysis or generalized linear model or something's applied, but just more on a higher level, like is it a study of people or is it a study of companies? And if it's people, what kind of people? Was it students? Was it people who are working startups? Was it executives and large companies? Just understand where do the data come from? It does not take much understanding of research methods to understand what the data are. Then you need to understand what is the research design? Was it interviews? Was it survey? Was it archival data? Or was it perhaps experiment? And this is the level of understanding that you need to have to understand the main idea of the article. Then if you want to evaluate if the article is, if the claims are valid and how strong the claims are, then you need to go and look deeper. But normally just looking at what's the research design? What is the sample is sufficient? Then main results and what it means, typically you might skim through the results section because that basically provides the evidence behind the results and then the results are typically stated again in the discussion section. So the discussion is the next main focus. So you would focus less on the data methods and results part and more on the discussion because it explains the results unless you want to evaluate how well the article is done. Okay, now there's another thing beyond the macro structure, the big structure of the article that you need to understand. If you understand this, it may make you a better reader of academic research articles. The idea is understanding the microstructure. So an article is a set of paragraphs. So the paragraph is the basic unit of composition. This is an important idea for those also who write master's thesis and want to have a higher grade, understand that one idea, one paragraph, if you have two ideas, then you split it into two paragraphs. If we understand this, then we understand that once we have read the paragraph and if we get the main idea, then we can proceed to the next paragraph. The paragraph itself has a structure as well. Let's take a look at this paragraph, the introduction section of Sapiens' article. It is four pages, 15 paragraphs. That's quite a long introduction because it also explains the theory as well. So how do we go about making sense of this all? Well, we do make a sense of it all by understanding the structure of the paragraph itself. Now, this is something that people don't usually cover on research methods courses, but I find this a very useful thing to understand. So in a well-written paragraph, there are three kinds of sentences. You have the opening sentence here, which gives the topic. So this is the first sentence of the paragraph. It tells that the only internalization of startups challenges traditional theories of internalization. If you know those traditional theories, good. If you don't know, well, you understand that that startups are something that prior theory does not explain really well. And then there is the conclusion sentence, which says that studying these startups might give us new insights that existing theory, existing empirical evidence does not. So this is called the concluding sentence. And anything between is supporting sentences. So we have the topic sentence, which tells what this paragraph is about. And then we have the concluding sentence, which give us the conclusion of that paragraph. Everything else is supporting the conclusion starting from the topic. So why is this useful to know? Well, here is my pro tip. If you want to quickly understand a part of an article, don't read the supporting sentences. There are some exceptions, but focus on the topic sentence, focus on the concluding sentence, and then don't read anything between here. So between here is simply just an explanation of some prior internalization theories. But the topic sentence already said that they don't explain startups really well. And if we don't want to go into detail and understanding which specific theory does not explain this well, then we don't actually need to read any of that stuff. So just read the first sentence and the last sentence and see if you can get the main point of the paragraph from just those two sentences. There's a caveat. This applies only to well written articles. So some articles that you read might be really difficult to read because they are really poorly written. And this is something that I often like to complain as a reviewer that many articles just have loads and loads of text to express a simple idea. So a simple idea should be expressed in a short form and one idea, one paragraph, instead of paragraphs that might be two pages long, which I've also seen, that's horrible writing. Then the blame is on the author, not on the reader. So this article is well written, so we can apply this rule of thumb of reading the topic sentence and the concluding sentence to understand what the introduction is about. Let's apply it. So paragraph one tells us that early internationalization of startups challenges prior theory on international expansion, therefore studying startups is useful because it can open up new avenues for theory. So they're basically saying that startups are something that existing theory does not explain, therefore we should study them. And that's the point of the first paragraph. Once we understand the point, the main idea, one idea per paragraph, we move on to the second paragraph. So the second paragraph has only three sentences. And because it's only three sentences, we don't save a lot of time and effort by omitting the middle paragraph. And these short paragraphs you probably should read in full. And this is particularly important here because they tell that they have a framework where they apply the concept of capabilities and the concept of dynamic capabilities and they give definitions. So if we have a marker at hand, now it would be a good time to underline these definitions and write to the margin that in this part of the article, you will find definitions for some of these key concepts. So this is about definitions and it's about giving the key concepts. Then the third paragraph, they say that they build on the idea of imprinting. They explain that the idea of imprinting is that what you do early on has long lasting consequences. And then they say that early internalization has been proposed that it affects later growth. But then again, this has not been explained in detail. So the idea here is that they apply the concept of imprinting to extend a previously presented idea. This is the point of the third paragraph. Then the fourth paragraph, this is a bit longer one and in here we have lots of citations. So this is more of a supporting evidence for this paragraph. So maybe not worth reading. And the paragraph says that the early part of the internalization process provides an interesting context in which to study the development of capabilities. And then they say that they integrate various prior theories. So what those prior theories are, if we want to focus more on this paper, we might go to the detail. But just it's enough if we want to get the big idea that this paper integrates various different theoretical perspectives. We don't necessarily need to understand what those are to understand the main idea. Now, another interesting and important thing to note is when there is a change of topic within a section. So this introductory section has been focusing on general terms. So they talk generally why studying startups is important and they study why or what studying startups enables them to do. But they have not presented any specific claims yet. They have presented some concepts but no specific claims. And now they go and they start explaining their specific claims. So when you identify this kind of switch of theme within a section, then it's a good idea to stop and think about have I understood the previous paragraphs within this section, this introduction well enough that I can proceed further. Should I go back and perhaps revisit one of these first four paragraphs? We'll continue. And then again three sentence paragraph, they say that they advanced two main claims. And the first claim extending out your prior research out there is a co-author in this article. They claim that early internalization leads to learning advances. And therefore, internalization expansion early on can help you to grow much quicker than what you would otherwise do. So that's the first main claim. And then the second claim that they do is that they point out that in previous study by Autio, it did not look at the possibility that early internalization expansion might actually hurt the company, make it less likely to survive because international expansion takes resources which the young company might not have. And then they address the timing of how the timing of internalization expansion affects the capabilities and resources and how that affects survival. So that's their second idea. So they think that it enables companies to grow, but it also affects survival. If you expand internationally as a young company. Paragraph seven is talks about the relationship within growth and survival. So that is the topic sentence. And then there's some research cited here. We don't need to understand the specifics, but we look at the conclusion. The research demands of growth are challenges that if survived may make a firm stronger, yet many will be unable or unwilling to develop the dynamic capabilities required. So if you try to expand internationally early on, that will help you to grow, but it also increases the risk of failure. That's the main point of this paragraph. Paragraph eight and nine are basically supporting paragraphs for paragraph seven. And they just say that relationship between growth and survival is complex. So it's not studied much within this context. And I would read all these two paragraphs together because this is a single sentence paragraph. And those are typically best understood within the context of the preceding or the next paragraph. So that is paragraphs eight and nine. And what we will note now is that they have advanced, they also now switch to a different theme again. So in these paragraphs between the five and nine, they argued that early international expansion has a positive effect on growth because of imprinting and developing dynamic capabilities. But it has a negative effect on survival because not all firms are able to do so and therefore they fail. And now they move on to another topic. So they decreased the level of abstraction and they look at these two mechanisms. So would there be other factors that affect the magnitude of these mechanisms? And they say that we posit that aged internationalization, it's a mandatory experience and research fungibility moderate the impact of internationalization or probably the survival growth. Okay, there is a term moderate. How do we know what that means? That term has a specific meaning. I know based on my training that that meaning is that you have one variable that affects the magnitude of the effect of another variable on a third one. So that's moderation. So it's kind of like it strengthens or weakens. That's the meaning of the term moderates. If you don't understand that term, then you need to go to the research methods course because it's kind of assumed that you would understand this kind of terminology when you read the articles. There are a few terms that you need to know and this is one of them because if you give moderate to Google, then Google gives you a definition that relates to moderation within the context of, let's say, a forum posting when someone is moderating if a posting is allowed to go through or not. And this is the three centers paragraph. Then we should probably read it. The idea is that resources and capabilities determine survival and then this expands the arguments from prior research. Paragraph 11 is a short one, just two sentences. We read it fully. So the idea is that the younger the firm, when it expands internationally, the stronger the effect on growth and the stronger the effect on survival. And this is the conclusion that they have. So the younger, if you expand internationally early on, then the imprinting effect is larger. You have more capabilities that you develop, you grow faster. But on the other hand, if you start early on, you're international expansion, you have less resources, therefore the risk of failure is higher. The next paragraph is the next claim. So we can now see that there's one paragraph or one of these moderators. So the first one was age at international expansion. And the second claim is that increasing managers or having managers with experience from prior international expansions will make the effects on growth stronger and effect on survival weaker. So the negative effect on survival is weaker. The positive effect of growth is stronger. So that's what they claim. And they explain here in the file sentence that the mechanism for that is that the managers bring routines. So routines are ways of doing things within an organization and the routines are the mechanism. If you read the supporting sentences, you would see that the concept of routines come from Nelson's and Winter's book from 1984 or 82, but you don't necessarily need to know that to understand what is the main point. So managers bring routines and routines help you to develop capabilities. Paragraph 13 finally is about resources and resource fungibility. So they explain fungibility again. It was explained before and it's explained here because that might be something that the reader might not know. So the idea of fungibility as they explained is that you can take a resource and repurpose it for other purposes. So for example, if you hire people with general skills, then those are fungible because you can also use them in the markets where you operate currently. But if you hire people with specific skills like you want to expand to China and you hire people whose main language is Chinese but they don't know much Finnish, then they would not be of much use to the company if the expansion to China fails and the company just to stay in Finland. So that's an example of non-fungible resource. And they say that the fungibility, they use it because they think that it's independence of company size. And now we have another paragraph that says that these three moderators are basically independent. So they can be studied together. Fungibility is related to age, somehow they explain that, but they can be studied as independence. So that's the main point of paragraph, which you get from reading the first sentence and the last sentence. And then there's another change of topic. So now we can see that the previous group of paragraphs talked about these, what they call moderators. And before we proceed, it would be important that we understand the argument behind each moderator, each of those three what they are, what is the argument. Once we understand that, we can proceed to the final paragraph which is simply saying that here is the complete model, how nice of them to provide this kind of diagram which summarizes the claims of the paper. And then they say that in the next section, in the theory section, they develop these propositions further and explain the reasoning, explain the causal logic behind each of these arrows. So that is how you would read the introduction part of this paper using this technique of focusing only on the topic sentence and the conclusion of sentence, except if it's a short paragraph in which case you read it fully. And to summarize how to read and understand articles that present theory, or how do you understand an article that presents theory and data? That's also an article that presents theory. You first focus on identifying the central concepts that are typically a few. Not all are boxes. Some of these concepts might be the arrows, like dynamic capabilities and imprinting. They are part of the arrows. They are the mechanisms. Then you find the definitions. And you use the structure of the article to do that. So the definitions are almost always found in either the introduction or the theory section. And the definition of a concept, the concept is almost always mentioned in the introduction sentence. So you can just go through the introduction sentences or the topic sentences and look where do I find that concept being discussed. Then I'll probably find the definition there. So understanding the structure will make you more efficient and more effective reader.