Side projects are a funny thing in graduate school. Grad students seem to commonly get sucked into a series of side projects, yet their “main projects” don’t just magically go away so there is no need to worry about them. So far in my research “career” of about 7 years, it seems that a mixture of side and main projects is important for research success, but that mixture should not be maxed out to the point of burnout. In this reflection, I will talk some about my experiences managing side projects, and how this balance leads to a greater chance of success because of chance of failure of both side and main projects. It is uncomfortable to talk about failure. Many people try to avoid it. However, it is important that researchers and future researchers hear about the near inevitability of failure in research. Sometimes, research projects fail. Academics also know the pain of rejection in the publication and grant application process. Acceptance rates are small for grant applications. Tenure track position applications are even worse. Many will apply for 100+ positions and only get 1 interview. Aspiring academics and researchers should be prepared to face that harsh reality. Thankfully, some academics are starting to talk more about failure in research by creating a “shadow CV” or “CV of failures” which is a list of failed projects, rejected grants and publications, etc. But here I will only talk about the first part, when research projects fail. Sometimes, main projects fail and side projects succeed. Other times, side projects fail and main projects succeed. (And maybe there are a handful of instances where they all fail or all succeed). It is hard to know from the outset which projects will succeed and which will fail, even when taking your advisor’s advice and guidance into consideration. I have had a decently wide array of success and failure in research projects, which I discuss below. When I worked at Oak Ridge National Laboratory (ORNL) in 2017 and 2018, one of my side projects turned in to my main “success” (in the typical academic sense, referring to a publication). While my main project of about 6 months of full-time work, assisting in synthesis and characterization of PdSe2, was successful and led to a publication, I was not given authorship for silly political reasons, so I have little to show for it. One of my side projects, which was coding an in situ monitoring system for pulsed laser deposition synthesis of 2D materials, gave me not one but two co-authored publications (and an Outstanding Scholarly Output team award) for work that I did in the last three months of the 16 months I worked there full-time. In my 2018 stint specifically, I was in charge of building a scanning photocurrent microscope. While we did achieve some success, further refinement (particularly of the beam spot size) was necessary before publication quality results could result. I am unsure what happened to that project, as I did not receive a reply to email in September 2019 asking about it. Last I knew, an importance piece was ordered and set to arrive after my internship ended. I had two other side projects that I did not have sufficient time to complete or make significant process on, one of ended in a publication (though my contribution was negligible so I was unsurprisingly not on it). One final side project was writing a review paper on perovskite solar cells, which was an extraordinarily daunting task for an undergraduate with no experience in perovskite solar cells. Though I spent countless hours (including multiple sleepless nights over my Christmas break) reading hundreds of papers, and writing a forty-page literature review, I was unable to tun the review to be publication quality. The feedback I received from my mentors was less than enthusiastic (and were uninterested to help me finish), and looking back at the paper now, I can see their critiques easily. Going from not knowing anything about a field to writing a review paper on it in six months is…ambitious to say the least. Not smart and not a good use of time is probably a better way to describe it. However, that review paper has served as the primary motivation for my class project this semester for my Materials Design Studio course, as I am applying machine learning to a perovskite solar cell database using insights from I gained from that review paper process. The class project *might* turn into a publishable result, but it is far too soon to say (I am not even sure I will or should pursue it). Before my ORNL experience, I worked with Micah Green in the TAMU chemical engineering department. Early on, I hopped on to a PhD student’s project and helped him finish out the last steps of the project, earning me authorship on a publication. Then, I tried to expand this work, extending a nanomaterial ink coating method for 3D printing to produce conductive printed parts, writing my first undergraduate thesis on the topic. My lab partner and I were having some technical difficulties getting the conductivity we wanted, my partner graduated, and I was becoming more interested in fundamental science and physics rather than engineering, so I switched lab groups. The project was dropped by Dr. Green to focus on other projects, such as one of my earlier side projects. The side project was to create an array of carbon-based pixels that could be heated by applying voltage that could create a thermal image, and this project was carried on by a lab mate into a publication. Finally, that leaves my current research group with Xiaofeng Qian in the materials science department (at least for the part of my research experience I am going to discuss). I have had one main project since I started halfway through my 4th year of undergrad (January 2019). I wrote the first stage of the project for my second undergraduate thesis in spring 2020. I had a full draft for publication sometime that year. Long story short (which I may tell later), it was just recently accepted for publication, thankfully. I have started work in another primary project that is even more condensed matter physics-heavy, which I am excited about. One side project is currently being head by a lab mate. Another side project I hope to write up for publication this summer. A final side project, which I proposed to work on for the NSF GRFP in 2020, was scooped in March of this year. Although they did not do calculations on the specific materials I was interested in exploring (along with their stability), and we can do some calculations that the paper did not cover, it was the same general class of materials, pentagonal 2D materials. My advisor thinks that this is sufficient to neglect work on that project and focus on other more impactful and interesting projects. Reflections While my research experience is a bit of a graveyard of dead side and main projects, there are the gold treasures in there that were able to be published, and for that I am very thankful. My key takeaway is that even with equally good scientific basis for a project, you just do not know what projects will fail and what will succeed. Some perfectly legitimate projects may end up crashing and burning horrifically, while some stretch ideas might accidentally succeed through l̶u̶c̶k̶ divine providence. Sometimes, research projects might actually require decades of development in a neighboring field in order to overcome certain obstacles. Sometimes, you may get scooped. My takeaway: do not become so attached to a single project that your research career lives or dies with that one success. Get your hand involved in multiple projects, collaborating with lab mates, and your chance of publication will be maximized. On the other hand, there is life-work balance and issues with overcommittal. You don’t want to get involved to so many projects so that you don’t make progress on any. I like to have at least a solid 2 projects going so that I can work on one when another hits a brick wall. I can’t say I have settled into an optimal routine, as I transitioned in to graduate school during the pandemic, which contributed substantial mental stress and threw off my productivity quite a bit, which led to a runaway reaction of increasing activation energy to work on manuscript revisions that I had been staring at for way too long. Now that that publication has been accepted, I can hopefully reestablish a routine for the last months of graduate school. While graduate school can be high-stakes and high-pressure, constantly feeling like everyone is telling you that their project is most important, and your advisor may be inconsistent with trying to instill all six of your projects as being top-priority work, it is important to take charge of your life and your projects in a healthy way and find a system of productivity that works for you, and you can work with your advisor to come to a happy equilibrium.
0 Comments
Philosophy of science is an important topic for philosophers, science communicators, scientists, and educators at all levels. I have discovered this as I have been reading various items in the philosophy of science over the past year, and I also audited a philosophy of science course in fall 2021. Philosophy of science attempts to answer various questions that science itself does not aim to, nor can, answer. For example, philosophers of science ask questions like, 1) what distinguishes science from non-science or pseudoscience (called the demarcation problem), 2) are scientific theories (approximately) true, 3) what makes a good scientific theory or explanation, 4) why and how do scientific theories change over time, or 5) what is the nature of scientific laws? Besides these key questions there are plenty of other interesting questions worth exploring, such as whether or not chemistry and/or biology reduces to fundamental physics, what explains the fine-tuning of the fundamental constants, or for my own research, what are the philosophical implications or assumptions behind my own research’s use of density functional theory, which makes some substantial (inaccurate) assumptions about reality, such as independent particles, that all relevant information in the wave function is in the electron density, and more. Just as a case study, I will reflect on my journey on the question of whether scientific theories are (approximately) true, which is the debate between scientific realism or antirealism. There is a spectrum of views on trial here within the realism and antirealism camps, as seen in the diagram below. Another related question (that is also grouped under realism vs antirealism) is whether we are justified in believing those scientific theories as true. The average person is a scientific realist: our best scientific theories are (at least approximately) correct, and we should believe they accurately describe reality. In the words of my philosophy of science professor, “Every single one of you was a realist before you walked through that door this semester.” In philosophy, however, naïve realism won’t fly. You better be able to give a good reason (or, preferably, an argument) to believe anything. Before reading any philosophy of science, I was a comfortable naïve realist. Of course, scientific theories are true! Why wouldn’t you think so?
I’m glad you asked. One reason is to look at the history of scientific theories: they’ve all been wrong in the past! All non-contemporary theories were false theories. Even worse, they were very empirically successful theories and led to many discoveries and predictions, resulting in great technological advancements. And yet, strictly speaking, they (e.g. Newtonian mechanics of phlogiston theory of combustion) are false. They do not accurately describe reality. This argument from the history of science is called pessimistic meta-induction (it is doing induction over the history of scientific theories, so a form of meta-induction). At the same time, how could a scientific theory make such accurate predictions unless it were, in fact, true? Or, to contend with the historical graveyard of scientific theories, approximately true, since at least Newtonian mechanics is correct to an approximation (for velocities much smaller than the speed of light). This argument is called the no miracles argument, since it would be a miracle if scientific theories would be so predictively accurate and yet false. Yet, why should we think that there is a necessary connection between predictive accuracy and truth? Perhaps scientific theories do not aim for truth in the first place, but just aim solely for predictive accuracy in order to get things like technological outcomes. This argument thus assumes that predictive accuracy as a way to judge between theories confers justification on that theory (i.e. is an epistemic theoretical virtue), which is exactly the question at hand. In other words, the no miracles argument begs the question. After coming to this realization, I knew I had my work cut out for me to defend scientific realism. It seems like, rather than appealing to the history of science to defend realism, we would need to appeal to a priori (from reason) facts to justify realism and avoid begging the question. In the end, I think realism is defensible because of a priori truths that support the epistemic (i.e. justification-conferring) nature of the theoretical virtues, such as consistency, empirical accuracy, and explanatory power. In order to evaluate each virtue individually, we need to consider a general principle for justification (which here is going to be to find its associated a priori principle) and think through each virtue and whether there is something unique about it that would make it different than the rest (i.e. find a symmetry-breaker between the two). Consistency is supported by the truth of the law of non-contradiction, so theories that do not contradict themselves are more likely to be true. Theories that are greater supported by evidence (i.e. greater empirical accuracy) are more likely than true because of probability theory, as in P(T|E) > P(T), P() is probability, and T is theory, E is evidence. Finally, explanatory power is an epistemic theoretical virtue because of a principle of sufficient reason, which says that contingent (scientific) truths have an explanation, and this truth is a necessary a priori truth. Arguing the points above is precisely the point of one of the essays that I wrote for my philosophy of science course last semester. It was difficult yet exciting to attempt to justify the very nature and truthfulness of science that I had so naively believed for so long. Although I had already been exposed to the self-defeating nature of the view called scientism, which says that science is the only (or best) way of knowing anything, this new challenge to science was deeper. Every scientist needs to reckon with this. There are serious and knowledgeable scientists and philosophers at both ends of the spectrum pictured above. There are profound questions lurking at the foundations of science that cannot be answered by stomping your feet with the anti-philosophical boot of Stephen Hawking or Richard Dawkins. Bill Nye, while originally disparaging of philosophy, began to appreciate it after reading some of it (because he was lambasted for some wild comments he made on it). There is no experiment that can currently answer the question of the correct interpretation of quantum mechanics, and there is no experiment that answer the question of the nature of scientific laws or the truth value of scientific theories. Philosophers of science have made great progress in exploring arguments for and against the relevant positions, and scientists would do well to read some of the latest philosophy of science, and not just the 20th century work of Thomas Kuhn and Karl Popper. My Thought Process To attempt to be more reflective than informative, I will discuss my process of trying to answer the question of ‘why should we think that scientific theories are true?’ First, why am I asking this question? One of events that most helped me think more deeply was a graduate seminar in ethics I audited in spring 2021. It was me (who has only taken one previous philosophy course and audited one other) with six philosophy PhD students (who typically had a BA and/or MA in philosophy). The professor led this course in a way that challenged my critical thinking well beyond my previous experience, and I loved it. Anytime we gave a positive argument for something, the professor would ask us, “But why should we think that?” Further, he did not let us agree with him and got ‘upset’ when we agreed with him too easily. I remember one time in particular when he challenged one student’s claim, she conceded, “I think you’re right.” He replied, “No, no, no, never say that.” These experiences (and others like it) in philosophy have shaped my ability to think more rigorously, including about why trust science as getting at truth in the first place. One final note is that I have seen in Idea Puzzle that philosophy of science can be applied to any scientific research by forming a coherent research framework that can be used in any scientific discipline. It helps you think through questions that are helpful (but not easy) to help get a grasp on the full picture, motivation, and potential issues with your research project. This could be applied to a specific paper or thesis or a research program. Since TAMU does not have institutional access, I could only try a few things. The questions are difficult to answer but rewarding to think through. Overall, philosophy of science has greatly increased my understanding of the nature of science, and it has forced me to think through deep questions on the implications and presuppositions of my research, and it also enables more clear thinking through my research methodology. It will also ensure that I do not end up saying embarrassingly incorrect things about the implications of my research, which scientists are prone to do whenever they make philosophical claims, such as those about free will, God’s existence, or ethics. In this list of 22 different work-related values, there is one that sticks out more than any of the rest, by far: I want to Work on the Frontiers of Knowledge. There is almost nothing more exciting to me than expanding the human knowledge base. Many of the application essays I have written include statements like this, with good reason. The excitement of discovery of something new is beyond thrilling. Even just a tiny breakthrough in the circle of knowledge, as seen in the illustrated guide to a PhD, is worth much celebration.
I am by nature very curious, and this has only increased over time, as I have discovered how much more that I have (get) to learn. I have many, many questions, and I want answers to them. I believe knowledge is intrinsically valuable; that is, I think knowledge has value for its own sake as opposed to only having instrumental value, or value as what one can do with that knowledge. Of course, knowledge also has a very high instrumental value, such as the ability to Help Society (another work value I am concerned about and will discuss shortly). The consideration of this one value, Work on the Frontiers of Knowledge, already has me directly cemented into a career path: research. I started carrying out ‘official’ research the first semester of my freshman year (though I had already presented some research at an ASME conference in Puerto Rico a few years prior). I have continued every semester and summer executing research projects within a total of 5 different research groups (including research internships), and I know I want to keep doing it for the rest of my life. Regarding Helping Society as a work value, my goals are much more long-term focused. I am not too worried about seeing the fruit of my labors within one or two decades, or even my lifetime. I aim to discover new physics, or materials with new physics, that will set the stage for technology decades down the road. I am quite interested in the advancement of solar technology, hence my interest in optical properties of materials. My work on nonlinear photocurrents can potentially be realized as shift current photovoltaics. If successful, these materials could lead the next generation of solar cells, which has substantial sustainability and environmental benefits as a renewable energy source, particularly in places where conventional electricity is difficult or possible to get. Outside of my research, I am also passionate about helping society, especially through sacrificial giving of my income towards those who need it for their necessities rather than my luxuries. I am an "effective altruist," and I have pledged to (and actually) donate at least 10% of my income to effective charities. I have personally committed to giving a (time-averaged) 50% of my income to effective charities. Overall, I want to Work on the Frontiers of Knowledge, using that new knowledge to Help Society. I am excited about making new discoveries and how those innovations can be used to make new technology, especially renewable energy. |
Archives
May 2022
Categories |