Many national polls are based on random-digit dialing (RDD). In this method, a computer randomly generates a phone number in the hopes of reaching an individual at his/her residence. In registration-based sampling (RBS), voter files are used to obtain a random sample. Researchers at Pew Research conducted an opinion poll using both RDD and RBS in order to compare the results. 

(a) Voter files are typically used for election surveys (to find likely voters). What type of bias would you expect would exist in using a voter file to obtain a random sample of all adult Americans aged 18 years or older?
(b) Do you think the type of bias that exists when using RBS also exists for RDD? Which method do you think would suffer more from this type of bias? Explain.
(c) Which method of sampling do you think has the lower response rate?
(d) Which method of sampling do you think oversampled individuals who identified with or leaned Republican?