Science vs Pseudo-Science

For something to be scientific it needs to be 2 things

  • logically consistent
    • if it rains, the floor gets wet
  • needs to be falsifiable
    • theory: all swans are white
    • how to falsify: find a non-white swan
    • if I can’t find a white swan
      • “I cannot falsify the theory yet”
      • but the theory is not proven

Errors

Type I - false positive (alpha)

  • the prisoner is found guilty, but is innocent
  • standard is 5% as a scientific community

Type II - false negative (beta)

  • the prisoner is not found guilty, but has committed the crime
  • = power of a test
    • not relevant, but for future studies
  • no definite decision, no yes/no
    • just a rather/rather not