 Hi, everyone. While we're waiting for a few more people, I'm going to type in a question in the chat box. And if you could respond, that'd be great. So I just typed in a question to see how many of you may be thinking about doing a capstone. I will just wait a little bit for some more people to come in. For those of you who just joined in, welcome. I just typed in a question in the chat box you can respond to or you can respond to the poll. We're just trying to have a bit of a sense of how many of you may be interested in doing a capstone instead of a thesis. So far, oh, no. Oh, no. I'm going to type in another question, a real cake, and just see where you all are. And if you can respond to that, what time zone are you in? OK, so as you are typing in about whether you're thinking about capstone or what time zone you are in, we'll get started. And I will get started with the screen share in just a minute, and you can keep typing. And it's really good to see you all. Wow, we have quite a bit of Eastern Standard Time. Nice. Korea. PST, Pacific. Wow, we have a lot of Eastern Standard Time. UTC plus A. CST, OK. Great. Can you all see my slides? Yes, we can. Great. All right. Some are in plus three in Kuwait. OK. Since we don't have anybody who's doing a capstone so far, I won't feel too bad on focusing on thesis in this session a little bit more. But you can always change your mind, like we indicated last time. And the session is recorded, especially the front part. And then we'll also spend some time working together. So probably the first 30 minutes will be me talking only. But feel free to stop me or punching a question in the chat box if something is not clear to you right away. And then we'll spend at least 30 minutes, maybe a little more, work on a couple of steps of the research process. So you can go back and then work on your own and before you fill out the form next week by next week, right? So I just want to encourage you to fill out that form definitely by August 17th, because the faculty will meet a few days later to match your choices with our expertise areas. So you want to make sure you fill that form by August 17th as we indicated. OK. OK, so just want to recap last time some of you may not have attended or maybe this been a while. And so what's the thesis? What's a capstone? In this case, I'm going to blend them together. So you need to structure an effective argument and make a compelling and logical case, especially for thesis, right? And gathering and analyze evidence in a systematic way and it's propositional, right? It is we're not offering a crystal ball into the future, not focus on the future, especially for thesis. And then for capstone, you are applying knowledge and skills to address important planning issue or issues. Now, that's not to say thesis, you're not doing that. You're not, you're doing that too, but it's not required to you do that, right? So if you're doing a thesis that's fairly theoretical or fairly empirical in a sense of collecting data on past trends, that is acceptable, OK? So that's just the recap from what we talked about last time. So you notice all the things you need to do above on the slide is a process. That's the process that we're going to be talking about or sketch out today. And you will be doing in the next six months or so, right? Really, if you count the real time, it's more like six months instead of two semesters, right? It's a process that will lead to a product towards the end that you either have a thesis or a capstone report. And let me just borrow what Lance Freeman, our colleague, you know, your professor, Freeman, he used to say the process matters more than the product. And I would totally agree because you will see six months time goes by really fast. And some of you may have really ambitious topic and you realize you're running out of time. What do I do if my thesis is not perfect? If my capstone is like two steps away? Well, get it done. This whole six months is meant for you to get started of doing a research or doing a large project on your own. The product may be great, the product may be OK. As long as it's satisfactory, sometimes the product will have to be what it is because of time. And so the process, learning how to do the research process and to create a product that you are at least happy with or proud of is really all it matters. So you'll see this towards February and March next year that you will have some time constraints and you will have to let go some parts of your research. You'll have to let go some interviews that you've planned online and so on. And that is OK. So what is this process that you will have to go through? Even if you're doing a capstone, you are going through a very similar process. So the very first step of a research process is to choose a research problem. So I'm going to lay out this somewhat abstractly without giving an example. And with the next slide, we will have two examples. So bear with me if it seems all very abstract at this point. And I also hope you've read that short piece of reading that I distributed called the research process. So the first step, you really want to think about a research problem or frame it in a way as a research question. And then perhaps in your mind, think about a potential answer to that question or even potential answers multiple to that question. And one your potential answer is a descriptive statement such as the crime rate in the city of New York has increased during COVID. That is a descriptive statement is what we call a proposition. And so I can say, oh, maybe the research question in that case is how has crime rate changed during COVID? So that's a research question. But if I revise my research question to the following, I'm breaking my own words. I'm giving you an example. It's hard to talk abstractly. So what factors have influenced the changes in crime rates in New York City during COVID? That's also a research question. But it's a little bit different. That research question is a little more complex. It's not descriptive. It's not asking for the trend. It's asking for the trend plus what factors influence the trend. And in order for you to tentatively answer that question, you will have to have a hypothesis. It's a statement that includes at least two things. One is the crime rate. One is the factors. And basically, you are trying to look at the relationship between the two things. And that's called a hypothesis. And a hypothesis is a statement that predicts a relationship between two or more concepts. OK, so you're saying, how do I come up with a research question? I'm just interested in crime rates, or I'm just interested in COVID. Well, let's wait a little bit. And in a couple slides, we'll talk about how do you develop a research question. And we'll have exercises today to do that. Anyway, so that's the first step. You really have to come down to a research question, or at least a research problem, and think about the potential answers. The next step is, OK, if my question is, what is the trend of crime rate in during the COVID time in New York City? I will actually have to collect some data, right? So how do I collect the data? I'm going to go with the easy research question. I'm just going to go with the trend question. So you have to formulate how you collect the information. That step is called formulating the research design. What data are you going to need? Where are you going to get it? What kind of indicators you are going to use to measure crime rates? Crime rate is not a concrete concept. It could be different kinds of crimes and their rates. And so you have to think about how you measure that concept of crime. So that's research design. More details for next slides. Then you go out and actually gather the data from whatever source you find possible. Then you either code, or maybe the data are in really good shape. You don't need to code. You just analyze the data. And then you interpret the results and testing the hypothesis or just simply showing the data, the results, and write it all up. Now, this doesn't mean you only write up to the last stage. You really should be writing all along all these five different steps. Some scholars will say there are four steps and maybe six steps. But generally, these elements are in there. But it is not a linear process. So I had a student last year who wanted to look at what New York's, to what effect, New York's Safe to Travel program or road improvement projects have had on traffic for accident rates. She did lots of work in trying to look at different data. Some data are not available. Some data are available. So in the end, she actually had to come back and reformulate her research question. And so all I'm saying is these five steps are not linear. You don't go from one to five and never look back. You look back a lot. Hopefully not until your fifth step, then you're looking back to reformulate your research question. But sometimes you do. Data availability, the availability of the people you want to talk to, the availability of the communities you want to visit, all may affect your research question. And you may have to go back and reformulate your question. Then you might have to change your research design and collect some other kinds of data. It's all possible. So during the fall semester, every breakout group under each advisor, the job of the advisor is to work with you through this process to come up with a feasible and viable research plan or research proposal that includes at least the first and second step of this process. So you have a good research question. You know how to gather your data. And some people move fast. They have already gathered some data during the process. And that's all basically what we're looking for in the fall semester. And then you move on in the spring semester to actually collect the data through the analy analysis and the interpretation. By data here, I do not just mean numbers. I also mean text information. I also mean qualitative information. And I also mean visual information, all sorts of information that constitute data. So it's by no means just numbers. So let's take a look at this table. I hope you can see well enough. And this is basically a recap of the reading that you did that I distributed early on. And the whole point of this recap has to do with how you design your research is really important. It might very well affect your results, your interpretations, and any implications you want to draw from those results that you say that are relevant for planning. So design your research properly is really important. But it also means that there isn't just one way to design a research. Meaning there may be multiple answers to the same research question. And that is totally fine. And we also know that when we do research, we generally have the goal of saying, well, if we're only focusing on New York City, does this mean that it's representative of another city? You have to also think about that. So when I summarize this reading, it really speaks about the issue of rep, oh, I cannot pronounce the word, replicability. Meaning if you use this research design in one setting, you should be able to get a set of results that can be replicated in a similar setting somewhere else. And that really is a good test of how what we call robust your research design is. But for a lot of master's thesis, I think these research design questions can be a little bit less rigid, which means we want you to go through the process at least once. But if you are interested in doing doctoral studies or become a researcher, research design would require a lot more your attention. And you might see that some research design needs more real work. So let's just look at this particular set of approaches in the same talk. So the research question is essentially, what is the social effect of density on humans? You know, as planners, we worry about density a lot. And so as you can see, there are two teams, at least summarized in the reading. One team is behavioral scientists, meaning psychologists mostly. The other team are primarily planners and urban scholars who are community-based or even city-based researchers. You can see the first team essentially has a hypothesis that it is the same exactly as the second team. Density causes harmful effects on humans. Now you're saying what's harmful, right? And that's subject to our interpretation. You could also reframe the hypothesis in a different way. You could say, as density goes up, the negative impact on humans goes up. That's also a hypothesis. That's essentially the same as this one. So there are also different ways of formulating your hypothesis. But I want to draw your attention to how they measured the two concepts in the hypothesis. One is density. One is social effect, right? So you notice, for both teams, they use exactly the same variables, social effect and density. But move on to the second box, right? Vertically, you notice the behavioral scientists use individuals as the unit of analysis. So a unit of analysis is essentially the subject that you are studying for your research. So if you're studying people, so each individual is a unit of analysis. If you're studying a neighborhood, the neighborhood is a unit of analysis. But you say, can I study people within the neighborhood? Yes, you can. But this is about unit of analysis that corresponds to your research. So we'll talk about the complexity when we get there. And then the population is essentially the totality of all your units. So the first team used 121 students. The second team actually used the unit of analysis very differently. It's a neighborhood, it's a community area, and then 75 areas in Chicago. Totally cool, right? Totally valid. And then data sources are also very different across two teams. The first one is primary. Meaning whenever we say primary data, at least in research design, we mean that you as the researcher went out and collected the data on your own. You talk to people, you interview people, you email the people, or you put the people in an experimental setting. In this case, it was experimental. They put students in a room and then they surveyed them. The second team used secondary sources. That means data that are already collected by other people. So anytime you use census data, Pluto data, Pluto data are New York City data, right? Or you used statistical yearbooks in different countries. Those are data that are already collected by other people. You're simply using them to do additional analysis. That's called secondary data in general. And so here focused on measure for density is totally different. For the first team is how many people are in one room. And second team, because they focus on neighborhoods, they focus on residents per acre. That's really kind of commonly, right? We measure density by how many people in that area. And then measurement for negative effects. The first team used attitude towards a hypothetical stranger. That say, if we add another person to the 121 student class, do you feel annoyed? Do you feel angry? Or do you feel just fine? That's basically the question that the researchers ask the students. So the measure is your attitude towards yet another person into this already large room. And then for the community researchers, they use delinquency rate, essentially crime rates in different ways. And the admission to mental hospitals basically measure people's mental health. And then the data collection, as I already alluded to for the first teams to interviews of number of students who were put in that room and being, you know, or having experienced the whole process. And then the second team used fact book, essentially secondary data sources, right? Now you read the chapter, so I'm not going to ask you, but you've read the chapter. Essentially the first team find, basically density has negative effects. We've said, you put more people in the room, people are gonna start feeling more negative towards the potential new strangers. So, and then researchers tried it on different classes and in different settings, they got the same results. The second team, however, in their original research design, did not find a definitive relationship between density and negative social effects, meaning you have community areas that have higher population density per acre, and those with lower, and the rates of delinquency and admission to mental hospitals vary all over the place. There's no relationship. Well, you say, okay, well, I just reject the hypothesis, right? Yes, you can do that. If you're doing the research and doing a thesis, that's your result, you have to report that. Maybe it's a research or design problem, but maybe it's just, there is no relationship between the two concepts. But the researchers obviously are aware of other research that's going on that indicates a negative effect. So they went about revising their research design and you also read about that, you read that chapter, the way that they, one way they modify the research is to change the measure for density. So they changed it to measuring density by how many people are in the individual housing unit. So that's a big change, right? In terms of measurement, and actually a little bit closer to the first team. And so basically, and then they also say, how many residents are in each bedroom or room, that's even the more refined density measure, meaning we find more micro in scale. So more closer to individuals rather than aggregated in a neighborhood. You know what they find? They find similar results as the first team. That is the more people who are living in the one room, there will be more issues related to delinquency rates and admission to mental hospital. There you go, you say, oh my God, really is very different, the results. So that example, oh, actually that's like two examples, really kind of give you a good sense, I hope you should read that chapter if you haven't, that research can be done in many different ways through different designs, how you choose the measures, how you choose the data sources and are all very, very important. Of course, how you collect data, there are lots and lots of different issues. And I'm just going through the basic skeleton of things today. And as you walk through the fall semester with your advisor in the breakout group, you will work on your own, right? So the first thing today that we wanna do a little exercise in about five minutes is that I want you to work with each other in small groups to look at choosing a research problem or a research question. Then you say, okay, all I know is I'm interested in looking at transportation planning. How do I really narrow it down? It's not so easy, but I wanna encourage you to start thinking about what aspects of transportation planning you might be interested. So that will be our exercise, right? So in the first step of research design process, you want to develop this overarching umbrella question that addresses your topic. So in the example we used is the effect of density, a social effect of density on humans. And then you want, so once, so it will take you quite, sometimes it may take you a couple of weeks to come down to a research question, and that's okay, all right? And, but once you have that, you want to identify the key terms in that research question, or you can go the other way around. Maybe you identify the key terms that help you formulate research question. That's okay too. So in this case, we notice density in the effect of humans, right? Humans can help you understand what your unit of analysis might be. But you notice from the comparison of the two projects, it could be individual, it could be residents in the total neighborhood. So humans can be measured in different ways. Density and effect, all can, right? We already know that. So generally a research question is a question that you do not know the answer to, but it's kind of interesting. And like you asked me, well, didn't you just say you want to come up with a potential answer? Yeah, you can potentially answer, but that's just a guessing. And your final research could either prove that potential answer or disapprove. We call nullify, right? You've done that in planning methods. It's called either prove the hypothesis or nullify the hypothesis or disapprove. So a good research question is important in the real world. So especially for planners, we're not looking at just theoretical questions and most of our research is applied research. And you also want to think about your research question as one that can be investigated systematically and answered empirically, okay? So what does that mean answered empirically? Meaning you can have some evidence to show whether the question should be answered the one way or another. So we're not doing, some of you may have had philosophy background or may have had even philosophy major taken philosophy courses. Sometimes we use logic a lot in philosophy. And it's about, you know, coming up with the logical steps to design certain things or to design a thinking process. Generally in social sciences, that's not a approach unless you're doing a theoretical thesis. And if you do, if you do have that, let us know and we'll also help you with that. A theoretical thesis requires a lot more argumentation, requires a lot more logics and that's usually not common. I want to reiterate that our thesis requirement as well as have some requirement requires all of you to not to base your thesis or capstone entirely on literature review, meaning entirely on other people's work. We require you to have new empirical or other kinds of evidence or new analytical framework or your own analysis, right? So just purely literature review kinds of thesis is not acceptable in the program. Okay, so next step is to practice asking questions about a topic as a pre-research. So Laila, our moderator is going to... Oh, sorry, I forgot to introduce Laila. Laila is the programming director at GSEP and she helped us with the first session and she's helping us with the second session and she'll organize these breakout rooms for a group of three of you to get into your little exercise. So rise and urban or planning problem you really care for and then write one or two questions you have about the problem. The way we're gonna do it is I have a set of Google Slides, which I will share with you just now. And then I know there are three of you may all have thought about your research question. You all want to share the three, that's great too. Just read out more than one slide. That's totally cool. Or you're still thinking about it. Maybe one of you would be willing to write down your research question. All right, so I'm going to stop sharing. We are going to go to this Google slide that I will share the link with you in just a minute in the chat.