Final answer:
The correct answer is option D. Decrease.
Step-by-step explanation:
When dealing with attribute sampling plans, one of the key factors is the confidence level. The confidence level indicates the degree of certainty we have that our sample statistics will reflect the true population parameters. If all other factors remain constant, decreasing the confidence level from 95% to 90% would reduce the required sample size. This happens because a lower confidence level means we are willing to accept a higher risk that the sample statistic may not represent the population parameter, so we do not need as large a sample to achieve this less stringent requirement.
Therefore, the correct answer to the question is D. Decrease. This is because the z-score (which corresponds to the confidence level in the calculation of sample size) is smaller at a 90% confidence level than at a 95% confidence level, leading to a smaller sample size being required to achieve the reduced confidence level.