.jpg)
or
orResearch a method of studying something. It is generally different from what is done in school, since often the outcome of the research is unknown. There is no Google to check if you are correct or not. There are no answers in the back of the textbook. You are generating new knowledge.
There are many ways to go about doing research. I will break it into 3 categories
Theory: Theory ideally predicts behavior. Models are simplifications that attempt to predict behavior. They tend to be general. Simulations are less simplified than models and similarly predict behavior. A simulation takes input parameters and then gives specific output parameters. They use many models. Models typically focus on specific phenomena, which, when taken together, can be used in simulations. Simulations generally can't be done by hand, whereas theories can often be implemented by hand.
Experiment: Experimentalists check the sanity of the theorists. They perform and devise experiments, taking into account many of the things that theorists ignore, to see whether theories have any basis in reality. Experimentalists must have a good understanding of theory in order to do this and also help to develop theories, themselves.
Engineering: Engineering uses elements of experiments and theory to create new things, which can then be studied. Engineering can allow experimentalists to do more experiments than they could do before. For example, without scanning or transmission electron microscopes, there would be much less nanotechnology research because there would be no good way to image things that small. Furthermore, really powerful computers must be engineered for theorists to do their computations and come up with new theories.
The lines between these approaches can be fluid or rigid, depending on what you are studying. For example, I engineered a high-temperature stage for some future experiments. I use theory to understand my experiments, although I have yet to formally come up with my own theories. An electrical engineer studying antennas might simulate an antenna radiation pattern, build the antenna, and then experiment on it.
Depending on the type of research you do (not necessarily what you study), your day can be quite different. When doing experiments in the lab, it is important for other people to be present for safety reasons, so a somewhat regular schedule is required. Some theorists have been known to do their best work at 3 AM, which would be difficult to do safely for an experimentalist in a smaller lab (although data analysis can generally be done at any time). While I rage over signal noise, bad statistics, imperfect mixing, thermal expansion, or irreproducible growth conditions, a theorist might rage over slow computations, programming bugs, or a program not compiling.
What are your experiences? What are your biggest frustrations? What does your typical day look like?
Partly for my own selfish reasons, I would separate theory and simulation into their own headings. Albeit, simulation does heavily rely on theory but, computations are all that I do (i.e. http://en.wikipedia.org/wiki/Computational_fluid_dynamics).
ReplyDeleteMaybe it's different between fields (fluid mechanics versus particle physics or scientists versus engineers)? From what I've seen, people studying fluids (within mechanical/aerospace engineering) are separated largely as computationalists or experimentalists. Engineering, even at the PhD level, is partly an applied science. Perhaps make comment that the exact versus applied science distinction behaves as a spectrum rather than separate categories. And, as you go from BS to PhD level within Engineering you move more towards the exact science end rather than applied end of the spectrum.
As a computational person, code compilation (/debugging), software licenses not cooperating, and DoD supercomputers malfunctioning can be big frustrations. The biggest time sink in the day may be moving large data sets over the network (order of 10-100TB). My computations can be on the order of days weeks but others in our group have simulations lasting on the order of months.
I completely agree that there is a definite spectrum between applied and more exact science (although I hesitate to say exact because things are really so messy, at least in my field). It is common for scientists to just do computations, just as there are some experimentalists who stay as far from computations as possible. I was having difficulty separating simulations and theory. Would you have any suggestions on how to make a clear distinction?
ReplyDeleteIt's tricky. I 'm a little inclined to just introduce both computation and experiment at the same time as theory. That is to say, theory is coupled with simulations and experiments. Experiments are used to develop/validate the models but also to interrogate the behavior of the physical system. Computations are an generally an implimentaton of models (from theory or otherwise) in an attempt to bridge the gaps where experiment might not capture. I would say that theory is the most fundamental mathematical or physical understanding of a system with perhaps two fundamental components. For me, there is the primary mathematical model---the Navier-Stokes equations ---that govern fluid flow but these are used for the experimental and computational people alike. On the other hand, in a supersonic flow, for example, we also have an understanding of the physics that tell us when and how shock wave systems will appear.
ReplyDelete