You are planning to conduct a case-control study that would examine a possible association between a genetic characteristic—called T1—and the incidence of lung cancer. About 10% of the population possesses T1. In this study, you have two laboratory methods that potentially could be used to measure a person’s T1 status from a sample of his/her DNA. Neither is perfect.

Method A will correctly categorize everyone who truly is positive for T1, but is expected to misclassify 5% of truly negative persons as being T1-positive. Conversely, method B will correctly classify everyone who is truly T1-negative, but will misclassify 5% of T1-positive persons as T1-negative. Which is the better choice of tests to minimize bias in the odds ratio relating T1 status and lung cancer?

a. Method A

b. Method B

c. Neither is a better choice than the other.