The Best First Research Paper: A Practical Tier List (Full Transcript)

[00:00:00] Speaker 1: What is the best first paper to publish? This is one of the most common questions I get from students, and the honest answer is, it depends. But after working with hundreds of researchers who have published their first paper, starting from zero or little background experience, I've noticed something interesting. There's some types of papers that consistently get published much faster and more easily than others. And while some can look really impressive, really complex, really deep, they almost never make it through peer review or get a really tough ride. And some, my absolute favorite ones, are what I call low-hanging fruit. Those easy, quick wins that can help you get your first publication, get those first points on the board. So what I'm going to do here is draw on my experience as a professor, having taught at Harvard, Oxford, and Cambridge, worked with hundreds of students to publish. I'm going to rank a dozen of the most common types of academic papers that I see early researchers do. And I'm going to put them on a scale from A, B, C, to D, with A being fantastic, really publishable, straightforward, easy to do, and D being, frankly, terrible choices that I generally recommend avoiding if you can. I'm also going to include that classic S, superior tier list of a few types of papers that I highly recommend and encourage because I've just seen such a good track record of researchers doing them as their first paper. And to do this, I'm going to rank these papers on two broad dimensions to help you think this through for yourself, publishability, how likely the journals are to accept this type of paper, and ease. So how easy, how fast, how realistic and feasible is it for a first-time beginning researcher to actually complete this? And remember, now's not the time to win a Nobel Prize. That's not the goal, and almost no grad student ever has won a Nobel Prize. Your supervisor would probably steal the credit anyway. Your goal is just to get some points on the board because early success builds confidence, confidence builds momentum, and momentum is what leads you on to bigger and better papers in your research agenda and pipeline. So let's go ahead and start ranking these. And a quick warning, some of the papers that people think are the best first papers are actually not ideal choices. This is also going to vary a bit by field. So if you are in a lab, you might not even have a choice, you might just be assigned a paper by your supervisor or an experiment, and that's just what you have to do. But if that's you, this is still going to help you wade through and think through the different tactics and strategies. All right, so first up, surveys. Surveys are one of the most common ways researchers will go out, especially in the social sciences, and collect new data. You design a questionnaire, you distribute it to a group of people, and you analyze their responses. It can be really powerful. It can shed light on new phenomena, spotlight missing data, reveal scale of a problem, maybe something no one has properly measured before. There's definitely value. The problem is it's a high variance project that can be expensive to run, it can be slow to connect enough responses. There's a lot of threats to validity, so that even after months of data collection, you can end up with a data set that's dead on arrival and journals and reviewers don't find very compelling. In other words, you could invest a lot of effort before you even know if the result is going to be publishable. This is why generally, I recommend to avoid primary data collection, meaning collecting entirely new data yourself unless it's absolutely necessary. So for a first paper, it's just often too much. So this is a surveys aren't bad, they can be very good. But as a first publication strategy, I'm going to put surveys and this type of primary data collection in the B tier. And this might be a little controversial, the narrative lit review. This is the kind of classic traditional way of reviewing a field and starting out. So you collect maybe from Google Scholar, a large body of literature, you summarize the main themes and you build a narrative around what we know and we don't know. And that's extremely valuable starting out. Supervisors often might ask you to do this, it's great for mapping out debates, for spotting gaps, paving the way for future research. And in fact, most dissertations do require some form of narrative literature review as part of the journey. But here's the problem. They're not easy to publish. Why? Well, many of the narrative reviews you see published, if you look more closely, they were actually invited reviews written by experts. So the editors invite somebody who's maybe been in the field for decades and want a perspective on what's going on. They really care about that opinion and the experts earned the right to synthesize literature in even a cherry picked kind of way. For a beginner, it's much harder. There's better ways to do it that are going to end up on this board that is more rigorous and more step by step. So narrative literature review, even though it's an important step in the research journey because of the lack of publishability and the relative difficulty as your first paper is not the strongest option. And for that reason, narrative lit reviews are going to end up in the C tier. Next up, we have commentaries. So commentaries are a short piece where you might introduce an idea, a hypothesis, an alternative interpretation rather than a full blown data analysis or set of new research. And so sometimes it's a short case study, a short case observation of an emerging or novel phenomenon. And these commentaries can actually be a really interesting publication strategy. So for example, what you can commonly do is a journal would publish a commentary that responds to a recently published paper. So you might take that study and extend the interpretation, challenge the conclusion, or use it as a basis to propose your new hypothesis. And editors often like this because it draws citations and impact to their journal and it keeps the conversation academically flowing and moving forward. In fact, early in my own career, if you look at my profile on Google Scholar, you'll see I use this strategy quite a bit. I'd have a new idea or hypothesis and I would seed it first, kind of plant it through a commentary, get it into the literature, plant a flag that this was my idea. And then later follow it up with a full blown empirical paper. So commentaries can play a useful part of an early research pipeline. Downside is they're shorter, sometimes harder to get cited. And not every journal accepts these unsolicited commentaries. But as a first publication strategy, it's pretty reasonable. So for that reason, I'm putting commentaries into the B tier. Now we're moving right up the list and we have our first entry into the A tier. And that is qualitative analysis, but particularly content analysis or discourse analysis. What I like about this type of paper is you don't have to collect brand new data. You're often taking existing text or existing content and analyzing it. So it could be policy documents, media coverage, transcripts, social media discussions, parliamentary debates, organizational reports. And you take that material and you systematically code it and interpret it to identify patterns, themes, or narratives. And that's powerful. It's powerful because you're generating new insight from data that already exists. So compared to something like running a new survey where you have to design a questionnaire, recruit participants, wait for responses, qualitative content analysis is often much faster and lower risk. That's not to say it doesn't require methodological rigor and strong theory and a conceptual backbone. You need that. So you do need a clear coding framework, transparent methods, and careful interpretation. But as a first publication strategy, especially in fields that value qualitative work, I think this is a really solid option. So it lands firmly in the A tier. Closely linked to this is secondary quantitative analysis. And here's where you're going to take an existing data set, an existing survey or administrative data set that's been collected and analyze it to answer new questions. And I really like this strategy. If you look in my own publications, this is one I used extensively early on. And it de-risks the whole enterprise because the data has already been collected, validated, probably already been used in published research. So again, this takes some of the big risk variables away from primary data collection. So instead of months, I mean the time, instead of months collecting new data, you can focus on what really matters, which is asking a clever question and analyzing it well. Another big advantage here is that you can sometimes link data sets together in imaginative ways. So the big risk is maybe that in these existing data sets, you can't test your hypothesis and that's where the linking comes in. So for example, if you combine survey data with policy data sets or economic indicators or geographic information, that can help you again answer questions that previous researchers couldn't because they were just maybe limiting themselves to that limited survey. Again, the big risk is that these data sets don't contain exactly the measures you want, perhaps in the precise way that would be ideal. And you might have to adapt your research questions to the art of the feasible to the variables that do exist. But overall, taken as a whole, first publication strategy, secondary data analysis is one of the safest and most effective approaches. So that's why this one sits right here near the top firmly in the A tier. All right, next up, we have theory papers. And not everyone's going to agree with me, but I'm going to be blunt here. For our first publication, theory papers are incredibly difficult. A theory paper is where you're not just analyzing data, you're trying to make a new conceptual or theoretical contribution, you might be proposing a new framework, a new model, a new way of understanding a phenomenon. And in principle, that does sound great. But in practice, it can be a very crowded space. To make a genuine contribution, you usually need deep familiarity with the literature, you need a strong understanding of what frameworks already exist, a sense of where the intellectual debates in the field are and are moving. And that kind of judgment usually comes with experience and maturity and a certain depth in the field. What I often see is early researchers thinking they have a brilliant theoretical idea, which they might, but it's not yet grounded enough in the literature to really land with reviewers. And so remember, when you write a theory paper, you're competing directly in the marketplace of ideas without the support necessarily of hard empirical evidence. And I prefer to go the other way around, make a series of empirical contributions, then suture those up and use that to justify and launch a new theory. At any rate, even if you are asked to review theory as part of your PhD process or research process, it can be very valuable. But so intellectually, it's great. But in terms of publishability and ease for a first paper, they score low. And for that reason, I'm going to put theory papers in the D tier. Next up, we have another type of qualitative paper, semi-structured interviews. And this is really bread and butter qualitative research. You interview participants, usually using a semi-structured interview guide. You then analyze the transcripts to identify themes, clusters, patterns, or narratives. And when it's done really well, it can produce very rich, insightful research that's kind of thick. And it tells us more mechanisms and how things work. But there's a deep challenge. Interviews bring back many of those same risks that a survey and other forms of primary data collection has. You've got to go recruit participants, schedule interviews, conduct them, transcribe them, and then analyze them. And what I often see happening is researchers who are just starting out, they really struggle to chase small, unrepresentative samples. They get maybe 10 or 20 interviews, which can make it really hard to publish and convince researchers that the findings are valid and reliable. So it's not to say that this approach is bad. Far from it. But it can be timing and high risk, high variance. And so for that reason, I would place qualitative interview studies in the B tier. Next up, we've got randomized controlled trials, which include also, we've got broadly experiments or other field trials. And methodologically, there's no doubt these are incredibly powerful studies that can be highly publishable. They're often considered the gold standard for causal inference, and journals love strong experimental evidence. So publishability right at the top. But here's the catch. Running an experiment in the real world can be extremely difficult. You often need funding, institutional approval, you've got to get the participants, the infrastructure, months, years of implementation, and you might not even get publishable findings. In other words, you can be dumping all your eggs into one basket. Although it has this methodological strength offset by the time cost and complexity, making them difficult as a first publication strategy. And for that reason, I'm putting these trials and experiments into the C tier. There's another type of scholarly output as well that people sometimes aim for earlier in their career, the academic book, which is a fantastic intellectual contribution. But just recognize it's not really a first paper, it's usually written after a researcher's already built a body of work and is consolidating it, consolidating those years of papers, ideas, and evidence into a larger narrative. You can see that of the three books I've written. They were often consolidating five to 10 years of my own research and hundreds of papers being stacked into them. So trying to write a book right at the start of your research career is just often too big of a project, and it can really slow down your progress towards publishing articles. That might have been the thing to do three to four decades ago, but the world of research has changed and evolved. So my recommendation is to write papers first, build your expertise, get those quick wins, and later form the book. So for that reason, I'm putting the academic book into the D tier. All right, now we're getting close to the top of the list. The next method and the last entry into our A tier is the scoping review. Scoping review is similar to the lit review, but there's one key difference. Instead of summarizing papers in a narrative way that you've hunted around on Google Scholar perhaps, you're going to use a structured, reproducible search strategy of big databases and repositories of articles. So you're going to have to define your keywords. You're going to set up a search strategy. You're going to apply clear inclusion exclusion criteria and document all the steps that you took so that another researcher can do exactly what you did and come up with the exact same result. And that structure is what makes a scoping review much more rigorous and publishable than a traditional narrative lit review. Also makes it a whole lot easier to execute. Instead of wondering which papers should I include? You follow a step by step guided process to systematically identify the literature and do the analysis. So if you're going to write a lit review, I strongly recommend using one of the more reproducible approaches to lit reviews. And for that reason, scoping reviews land up firmly in the A tier. All right, we finally reached the S tier and the two approaches that I most strongly recommend if you want the best balance of publishability and feasibility. The first is a natural experiment. Natural experiments are type of secondary data analysis again, so you're working with existing data. But the key and the real magic is in the research design. So instead of simply analyzing correlations, what natural experiments are trying to do is simulate the logic of a randomized controlled trial, but using real world events. So this might take advantage, for example, of a change in a policy or a regulation. Take advantage of a shock that affects one group, but not another. Maybe differences in somebody who's eligible or gets exposed to an event. And researchers can take advantage of these naturally occurring situations to compare before and after, or sometimes as the parlance goes with natural experiments, treated versus untreated or experimental versus control group. And the reason journals love these is because these can get very close to experimental evidence, which is incredibly powerful for answering causal questions. And in fact, a few economists actually won the Nobel Prize in recent years for pioneering these kinds of methods. So natural experiments give you something very valuable. You get the experimental level credibility without all the risks of running an expensive or complicated experimental setup yourself. And so that combination makes them extremely attractive to journals. It works well with quantitative methods. If done right, it can also work with qualitative methods. And so for that reason, natural experiments land firmly in the S tier. So you may be wondering, Professor Stuckler, well, what's going to be able to beat that? Natural experiment sounds pretty good. Well, we are going right to the top of the evidence hierarchy, and that is systematic reviews. Systematic review is a structured way of synthesizing all the existing research on a question. So instead of just selecting papers and summarizing them, you follow a rigorous, reproducible process. Much like with scoping reviews, you're going to define your search keywords, search multiple databases, and systematically evaluate the evidence. In contrast to scoping reviews, these are often pre-registered, meaning the research plan is decided and declared in advance, which further increases transparency, credibility, and ultimately, publishability. And because these synthesize the entire body of evidence, that's why they are at the top of the pyramid, considered the highest level of evidence in many fields. In medicine, for example, clinical guidelines and policy decisions, real world life and death decisions are frequently based on these systematic reviews. What I love about them is that even as a beginning researcher, you can produce evidence right where you are right now that is going to sit at the very top of this evidence hierarchy. You're not just publishing a paper, you're synthesizing knowledge in a way that can actually shape the very foundations of your field, how your field understands an issue and become a reference point. And if you have the right guidance, systematic reviews, they may seem daunting, but because of this rigorous structure, they're actually very step-by-step to execute. And I've got a full playlist that you can check out that'll take you 100% free right on my channel. It'll take you start to finish, and we've had many students do this in under three months with just five to 10 hours a week, start to finish, no experience to publishing a review. And so for that reason, the publishability, power of them, and their ease, they are absolutely top of the list creme de la creme right here in the S tier. So that's my ranking of the best first research papers to publish. And of course, caveats is going to vary a bit depending on your field. So I'm really curious to hear your perspective. Do you agree? Do you disagree? Is there a type of paper that you would put higher or lower in the list? Let me know in the comments. I love to hear about your experiences publishing your first paper. And look, if you're looking for more structured support, mentorship, feedback, and a clear roadmap to getting your first paper done, click the link below. See how we work. See if it resonates with you and see if FastTrack might be a good fit for you. Thanks for watching.