A researcher uses simulation to estimate a probability. She generates 10,000 random samples and observes that 2,340 of them satisfy a certain condition. What is the estimated probability of the condition, and how could the estimate be improved?
- AEstimated probability = 0.234; decrease the number of simulations to speed up the calculation
- BEstimated probability = 2,340; the number of simulations is irrelevant
- CEstimated probability = 0.234; increase the number of simulations to reduce the variability of the estimateCorrect
- DEstimated probability = 0.766; use the complement to get a more accurate result
Why C is right
The estimated probability is 2,340/10,000 = 0.234. Simulation estimates are subject to random variability. Increasing the number of simulations reduces this variability (the standard error of the proportion decreases as n increases), producing a more precise estimate.
Why the other 3 choices miss
- A
- Incorrect because decreasing simulations would increase variability, making the estimate less reliable.
- B
- Incorrect because 2,340 is the count, not the probability; the probability is a proportion between 0 and 1.
- D
- Incorrect because 0.766 is the probability of NOT satisfying the condition; using the complement does not improve accuracy.