A quality control manager wants to determine the percentage of defective light bulbs in the production line. Which of the following sampling methods would most likely lead to a valid inference?

Random sampling, where light bulbs are chosen randomly from the production line for testing, would most likely lead to a valid inference about the percentage of defective light bulbs in the production line. This method helps to ensure a representative sample and reduces the risk of bias in the results.