The last two methods listed are likely to be biased.
The third one, sampling 30 boards in a row an hour after production starts, does not take a very large sample. Additionally, if something in the machinery is defective only later in the day, the data will be misleading. Similarly, if the machinery is only faulty until it "warms up," the data will be misleading.
The last method, the voluntary survey, is likely to only get responses from people that feel very strongly one way or the other about their product. Choosing only 5 responses from this is also not the best choice, as that is an extremely small sample.
