[DGthe3]

Results of scientific studies depend on what assumptions you make and what variables you include. The more variables you include, the better. But the choices made for particular assumptions can make a huge impact. Thats where bias comes in. If you want to prove one thing you chose one end of the scale, to prove the oposite you pick the other end. Also, there is data selection methods. Again, you can pick something optimistic or conservative. Anyone trying to sell their technology will make it appear as good as possible. Someone trying to discredit a claim will use worst case senarios.