A survey was conducted to investigate views on gun control. One thousand homes in a county were phoned, but responses were only obtained from 200 households. Two different questions were asked on alternate households:

(a) Should more controls be placed on gun ownership to prevent random gun violence by unstable people?

(b) Should fewer controls be placed on gun ownership to help people defend themselves from violent crime?

Which of the following is correct?

• The fact that only 200 of 1000 households responded is an example of a random sample.
• The fact that households were only contacted by phone could lead to nonresponse bias because not all households have phones.
• People often have strong opinions on gun control. This could lead to response bias.
• The researcher was puzzled by the dramatic difference in responses to questions (a) and (b). This is an example of potential bias caused by nonresponse bias.