A common problem with most research designs is bias, which could be introduced win several ways one if which is the favored selection of certain variables rather than being randomized. In the scenario above, the sampling time frame introduces a bias into the design as the tune span of pesticide use does not allay with that of the exposure history of the respondent. Therefore an entire lifetime evaluation of one variable and just 20 years evaluation of the other introduces a sampling bias into the research design. Similarly, evaluation the type of pesticide usage over certain time period may need to be considered as the effect of very old pesticides may differ from those used very recently.
