Before conducting a study, researchers calculate study power. Study power is how likely the study will find a link between exposure and illness, if there is one.

Suppose a study does not have enough study power. In this case, the study's results may show that there is no link between exposure and illness, even if the exposure to contaminant really does lead to negative health effects. A larger sample size will result in a higher study power. A larger risk of disease in the exposed group compared to the unexposed group also results in a higher study power. Researchers often want to make sure that their proposed study has at least 80% or 90% study power.

1. Sample size
   The total number of people in the study. Researchers base this estimate on the proposed study design, and the total number of people who were exposed to contaminants.
2. Estimate of disease rate for unexposed population
   Researchers base this estimate on what is currently known about the presence and patterns of that disease in human populations.
3. Estimate of disease rate for exposed population
   Similar to the previous item, researchers use current knowledge to estimate this rate. Researchers may also conduct a pilot study to gather data to make this estimate.
4. Level of significance
   Before conducting the study, researchers choose the level of significance, which is the p-value they would like to achieve in their study. Researchers commonly set the level of significance at 5% or less.

This section's study power calculator will allow you to adjust the last three factors in this list to see how changing these factors affect study power. It assumes that researchers want the level of significance to be 5%.

(This study power calculator assumes that the proposed study is using a particular type of statistical test that compares the percentages of people with a disease in two different populations. This test is called the two-sample test for binomial proportions.)