Can you group people correctly into exposed and unexposed groups?
Exposure misclassification occurs when people who are exposed to a toxic chemical and people who are unexposed to a toxic chemical are categorized into the wrong group.
If the "exposed group" contains some people who were actually not exposed to the contaminant, and the "unexposed" group contains people who were exposed to the contaminated air, the study may not be able to show any difference in the number of people who got sick between the "exposed" and the "unexposed" groups.
Here's an example of how this might happen:
Imagine that the facility in a town released a lot of a toxic chemical into the air over the last 20 years. However, at that time no one sampled the air at different places in the town to find out where people may have been exposed to the contaminated air.
Scientists model how the chemical may have migrated in the air throughout the town to make their best estimate of who in the town was exposed to the chemical.
Scientists will have a hard time determining which people in the town were exposed to the chemical, and which were not. This is because there could have been a day when a large amount of toxic chemical was released, and the levels of toxic chemical in the air would have been high on that day.
Also, people move around in the town. People spend many hours of their day outside the home at places such as work and school. So just because a person's address was in an area that was thought to receive high exposure, it does not necessarily mean that this person received a high exposure.