$A(n),$ selection device shows a clear link between test performance and job

performance.

a$)$ reliable

b$)$ invalid

c$)$ valid

d$)$ unreliable