Do you ever feel that your investigative work is too inexact and uncertain to provide real scientific evidence?

Question: Do you ever feel that your investigative work is too inexact and uncertain to provide real scientific evidence?
- Keywords:
  evidence,
  inexact,
  research,
  uncertain,
  work
Asked by mayi001 to Wei on 15 Jun 2011.
- Wei Xun answered on 15 Jun 2011:
  
  Hi Mayi001:
  
  I think I have answered a similar question here: http://ias.im/43.53
  
  So the short answer is that no, since we report how accurate our results are. If I’m not at least 95% sure that something is true, then I decide that it’s not a significant finding which could happen by chance.
  
  Sometimes non-significant finding is just as important as significant ones. Think the game 20 questions: you are asking “is it an apple?” “No”, “is it a banana” “No”, after you’ve exhausted what it cannot be, you’re closer to what it must be. It’s a bit like the principle of investigative work a la Sherlock Holmes style:
  
  http://www.thescienceofdeduction.co.uk/

Question: Do you ever feel that your investigative work is too inexact and uncertain to provide real scientific evidence?