Carefully controlled experiments that test equally rigourously derived hypotheses should be used to examine the effects of each postulated cause. In a laboratory situation it is possible to eliminate, or adjust for the effects of factors that are either not of interest, such as diet or family interaction, or which are considered to confound the impact of some putative cause. The effect of the causal agent can then be assessed by exposing half of the group of subjects to it and preventing exposure in the other half - the control group while both groups are kept in standard conditions. Such a design allows us to be confident that the consequent development of the effects of exposure only in those exposed, are due to the causal agent and not something else.

However, in most situations, these kinds of rigourously controlled experiments are impractical on large numbers of people who live their day-to-day lives in the real world. More important, they are often unethical. So we must adopt other methods.

Different types of evidence (which ultimately underpin the EBP standards of evidence) that can be used to determine the extent that a cause and effect are linked are found in two main sets of guidelines developed to address these questions:

Koch’s Postulates and Hill’s Criteria. Both are very similar, with Hill’s Criteria being derived largely from the earlier work of Koch.