Chris has been hired to assess a new version of a college entrance exam. He randomly assigns 100 high school juniors to take the new exam and 100 high school juniors to take the old exam. So that the participants were unaware of the two versions, the new exam was administered in the school gym while the old exam was administered in the school auditorium. The students taking the exam in the gym complained about the smell, the temperature and the uncomfortable seats. The students taking the exam in the auditorium made no complaints. Chris calculated a statistically significant difference between the two versions of the exam (t(198)= 3.1, p< 0.005) and concluded that the new exam was not a valid substitute for the old exam. There is a problem with validity. Which validity is weak in this example? a. external validity
b. construct validity
c. statistical validity
d. internal validity