how do you know what experiment to do and when, when you are researching a particular topic?

Question: how do you know what experiment to do and when, when you are researching a particular topic?

Asked by harsh07 to Darren, Deuan, Duncan, Lori-An, Michelle on 17 Jun 2010 in Categories: Your Research.
- Darren Nesbeth answered on 16 Jun 2010:
  
  First of all you think of a good question.
  
  Then design an experiment that will answer you question.
  
  Obviously it has to be an experiment you can afford to do, and is safe.
  
  So ‘what would happen if Mars collided with Earth’ wouldn’t be any good as it would be very expensive (and beyond current technology) to get Mars to smash into Earth, and also very dangerous.
  
  How many txt messages can I send in 2 minutes? There’s a cheap and safe experiment right there!
- Lori-An Etherington answered on 17 Jun 2010:
  
  Research starts with wanting to know the answer to a question. Then you think about the best way to answer that question and what experiments might have to be done. Sometimes you might not know which experiments to do first (especially when you are just starting out) so you talk about it with colleagues and read books and journals to help you. Sometimes when you start experimenting you might not get the results you expected so you have to rethink what you are going to do next.
  Hope this helps

Question: how do you know what experiment to do and when, when you are researching a particular topic?