• Question: how do you test your theory

    Asked by Eadaoin to Gavin, Karen, Mark, Michel, Roisin on 9 Nov 2016.
    • Photo: Michel Destrade

      Michel Destrade answered on 9 Nov 2016:


      With great difficulty! For example I’m interested in how sound travels in soft things like the brain. I make some calculations (using pen and paper and a computer) based on the laws of physics about how fast it should travel. Then I work with engineers to find out if I was right. We design an experiment using a medical device and a pig brain. We make the measurements and see how the theory fits. If it doesn’t I have to go back to the calculations and figure out if I made a mistake or if I simplified the problem too much and need to refine it. Back and forth like this until everything is consistent and explains the physics of brain matter. Can go on for years!

    • Photo: Karen

      Karen answered on 10 Nov 2016:


      In my field of science theories are tested by controls. This mean if you want to test if something is abnormal you must first establish what is normal. So you take a defined number of normal people (i.e. In the age group you require, who are disease free etc) test them, establish a reference range ( the average result +/- 2 std deviations) and use this as your control/ reference for diagnosing disease and abnormal results.

    • Photo: Roisin Jones

      Roisin Jones answered on 10 Nov 2016:


      In general in chemistry, you come up with theories by looking at the work done by previous scientists in the area that you’re interested in! Then you propose an experiment to try and prove your theory, usually either by running a reaction with chemicals, or running a computer simulation that takes into account chemical properties (this is sometimes less reliable because we don’t know all the factors at play in a reaction and so it might give less accurate results), and you analyse the results to see if they prove your theory. If they do, great! If they don’t, then it’s back to the drawing board to figure out why not!

    • Photo: Mark Kennedy

      Mark Kennedy answered on 10 Nov 2016:


      In astrophysics, you have to apply for telescope time to test out any theory. It’s very hard to get time on the best telescopes, as everyone wants to test their theories. For example for every 1 second that XMM-Newton, which is the best X-ray telescope we have, observes an object, 6 seconds will have been requested by observers. Which means 6 people don’t get to do their experiment.

Comments