When I first got into science, I thought that it was about coming up with brilliant and evolutionary ideas and the rest would come together by itself. After ten years in research, however, while being by no means a great scientist, I did learn a few things about the making of one, which I would like to share with you. It's an immodest thing to do. But anonymous modesty never serves anyone. So here I go.
First of all, if you have not read Richard Hamming's lecture, "You and Your Research", read it now. If you haven't reevaluate your research after reading it, read it again. Hamming's lecture should really be writings on the wall for all aspiring scientists, for it shows us the most important step to great research: how to pick the great problems to work on. Avoid mediocrity.
But picking the right problem is just the first step. The really great scientists are also great problem solvers. Rome is not built in one day. Great science does not happen by itself even if you have a great idea. There is only subtle difference between a genius and a mad person. That difference is the methodical implementation of one's ideas.
Once you have an idea to solve a great problem, you set out to see if it is right. You carry out an experiment. If you are really onto something big, often you will need to design a new experiment that nobody has done before, you will build some new apparatuses, write some new computer programs, collect some hard-to-collect data. (Incidentally, having to do something difficult and unusual also hints that you may be onto something significant.) These new elements bring in extra uncertainty. The experiments may at first disagree with your hypothesis. You have to determine whether your hypothesis is wrong or your experiment has flaws.
The ability to verify the integrity of your experiment is the paramount skill you need to have as a great scientist. You have to be able to design experiments where you know the outcome to test your new apparatus, devise simple models with analytical results to test your computer program.
Once your experiment agrees with your hypothesis, the real test of greatness begins. You must overcome the initial excitement, and with a cool head, consider if there is an alternative hypothesis that also agrees with your experiment. How can you design additional experiment to distinguish between different hypotheses? Don't draw conclusions too quickly.
First of all, if you have not read Richard Hamming's lecture, "You and Your Research", read it now. If you haven't reevaluate your research after reading it, read it again. Hamming's lecture should really be writings on the wall for all aspiring scientists, for it shows us the most important step to great research: how to pick the great problems to work on. Avoid mediocrity.
But picking the right problem is just the first step. The really great scientists are also great problem solvers. Rome is not built in one day. Great science does not happen by itself even if you have a great idea. There is only subtle difference between a genius and a mad person. That difference is the methodical implementation of one's ideas.
Once you have an idea to solve a great problem, you set out to see if it is right. You carry out an experiment. If you are really onto something big, often you will need to design a new experiment that nobody has done before, you will build some new apparatuses, write some new computer programs, collect some hard-to-collect data. (Incidentally, having to do something difficult and unusual also hints that you may be onto something significant.) These new elements bring in extra uncertainty. The experiments may at first disagree with your hypothesis. You have to determine whether your hypothesis is wrong or your experiment has flaws.
The ability to verify the integrity of your experiment is the paramount skill you need to have as a great scientist. You have to be able to design experiments where you know the outcome to test your new apparatus, devise simple models with analytical results to test your computer program.
Once your experiment agrees with your hypothesis, the real test of greatness begins. You must overcome the initial excitement, and with a cool head, consider if there is an alternative hypothesis that also agrees with your experiment. How can you design additional experiment to distinguish between different hypotheses? Don't draw conclusions too quickly.
No comments:
Post a Comment