A factory produces light bulbs. The probability that any single bulb is defective is 0.02. A random sample of 200 bulbs is tested. For the normal approximation to the binomial, are the conditions met?
- AYes: np = 200(0.02) = 4 ≥ 10 is satisfied.
- BNo: p = 0.02 is too small for any approximation.
- CNo: np = 4 < 10, so the conditions are not satisfied.Correct
- DYes: n > 30 is sufficient for normal approximation.
Why C is right
The conditions for the normal approximation to the binomial are np ≥ 10 AND n(1-p) ≥ 10. Here: np = 200(0.02) = 4 < 10. Although n(1-p) = 200(0.98) = 196 ≥ 10, both conditions must be met. Since np = 4 < 10, the normal approximation should NOT be used. A Poisson approximation would be more appropriate here.
Why the other 3 choices miss
- A
- np = 4 does NOT satisfy the condition np ≥ 10. The condition is ≥ 10, not ≥ 1.
- B
- Small values of p are problematic for normal approximation, but Poisson approximation works well for small p and large n. The problem is np < 10, not that no approximation exists.
- D
- n > 30 is the condition for the Central Limit Theorem approximating sample means, not for the binomial-to-normal approximation. The binomial approximation requires np ≥ 10 and n(1-p) ≥ 10.