Assume a database with 10,000 elements of data is queried, through a statistical interface, for the average data value. The response is determined using the random-sample query method, with a subset of 500 elements being used. Assuming you had access to all data elements, how could you determine what the least accurate response could be?
closed as off-topic by bobthechemist, Kuba, RunnyKine, Yves Klett, Sjoerd C. de Vries May 13 '14 at 20:39
This question appears to be off-topic. The users who voted to close gave this specific reason:
You can take a sample of random selections and determine the probability of an inaccurate result.
In the example below a 'response' (e.g. mean) of less than -0.0767 or greater than 0.0767 has a small 0.27% chance of occurring.
See: 68–95–99.7 rule