Question: how do you make best experiment than other sienciests?

Leah Fitzsimmons answered on 10 Mar 2014:

First of all, there is no perfect experiment! The key is designing an experiment so that you can be as confident as possible in the answer that you get back.
For example; if I wanted to know whether a drug can kill cancer cells I would need to think about.
1) How much of the drug to use – ideally that same amount that you would use on a person
2) What to measure – you could measure whether the cells stop growing, or look for signs of them dying
3) Controls – you always need a positive and a negative control. In this case you could treat some cells with a different drug that we know will kill them (the positive control) and you could just add water to a second batch of cells, which should not kill them (the negative control)
3) How many times you need to repeat it to be sure about the result (I usually take 3 separate measurements for anything I look at and then I do the whole experiment three times – 9 measurements in all)
4) If any other explanation could explain your measurements – did the cells definitely die because of the drug or because of something else you did?
5) What other experiments you could do in different ways to support your theory and make it more convincing.
But probably most important is to always have an open mind – just because you expect one thing to happen doesn’t mean it will. When something unexpected happens it might lead you to find out something really important.