Four surveys are described below. which survey methodology would lead to the least biased conclusion

The answer is 3) From calls made to one hundred randomly generated telephone numbers, people replied to a question about television shows they watch

This is because the people questioned were completely chosen at random and were asked about something they might or might know much about. This is easily the least biased option.

However, other surveys were not biased, e.g. asking heart surgeons about children's diet or asking ex-employees of The World of Baseball Bookstore about the importance of baseball.

These are not a random selection and might not actually provide accurate results.