But another way is to address the same question using many different experimental techniques. If many different types of experiment give the same conclusion we can be more confident of it!
This is a very good questions – because if you tell everyone you have made a discovery, but they think your data is not accurate you will look a bit silly!
I make sure my data is accurate by getting a second person to check it. For example, I will conduct and interview and then try and pick out what the people in the interview were talking about. When I have done this, I will ask someone else to pick out what they think the interviewee was talking about. Then we will compare – if we think the same then we can be happy that the data is accurate. If we don’t think the same, then it’s back to the drawing board!!
Another way we can look to see if our experiments are accurate is to test lots of different combinations of the factors we are interested in, to build up a clearer picture of how these factors interact and see where the boundaries are i.e when does the experiment start to flop! This allows us to create whats called a design space – we can then do further smaller repeat experiments inside the design space. This helps if you have experiments where it takes a long time to get repeats or if you are trying to improve the experiment
Comments