Question: How do you conduct experiments?

Keywords: ,

  1. The first thing to do is to figure out a hypothesis, which is basically the idea you want to test in the experiment. It is important that this idea is “falsifiable”, which means it must be an idea that can be proven false somehow. The next thing is design a way you can test this hypothesis, and figure out what you actually want to measure (say, cell growth for example) and how you can accurately do that. It is really important to include control experiments that help you make sure the effect you measure is real and that the measurement you use is accurate. It’s also really important to do the experiment a number of times (we call them replicates) to make sure any effect you see is not just some random event that happened by chance. Another important part of any experiment is to keep accurate records of how everything was done (a bit like a complicated recipe) so you can go back later and repeat it, or try to figure out what may have gone wrong if it doesn’t work!


  2. Almost all of my experiments involve malaria, and we can grow it in the lab!

    Malaria parasites live inside red blood cells, and we use left over red blood cells from the blood bank that have been donated, but can’t be used for blood transfusions. We add the red blood cells and the malaria parasites to a petri dish, along with media broth that has all the nutrients that malaria needs to grow.

    I am making a vaccine to malaria, so some of my experiments involve looking at the immune response to malaria. This involves putting cell from the immune system, like T cells or phagocytes, in with the malaria cells and measuring what they do.

    Knowing how the immune system reacts to malaria will help design me a better vaccine, I want to know which bits of the immune system are the best at killing malaria!


  3. My experiments are all done on human subjects. A typical experiment will involve measuring something from the subject. These could be measurements from the body like strength or the electrical activity in their nerves and muscles, or it could a series of carefully designed questions that the subject answers. Then I change something about the subjects body, this could be blocking a nerve, making some muscles tired, or applying some illusion to trick the brain. Then repeat the measurements.

    What I change in the subject’s body and what I choose to measure both depend upon the question I am trying to answer and the specific hypothesis that I am testing.


  4. If you have an idea of what problem it is you want to solve – say, in my case, breast cancer- you have to come up with a hypothesis. That is, what is the idea you want to test? What is the question you want to answer? So my question is: Does this gene cause breast cancer? Then you design a set of experiments so that you can test this hypothesis. You need to include controls that will allow you to show that it is only this gene that is having this effect when you test it. Then you need to repeat these experiments to show that you consistently get this answer, that it’s not just a one off, or in case you mixed up your samples! One way of testing whether a gene has an effect is to create a “knock out.” That is,we remove the gene from the genome. Then we see what effect it has on the function of the organism. Does the organism grow slower? Is it missing anyhting? So basically, experiments are smart ways of using clues to find things out. Each clue will lead you to another clue and bit by bit you build up a picture of a how gene works,or how a disease affects the body.