They are not limited to the questionnaires, and can freely talk about related matters that could support their answers. Focus group discussions (FGD) usually allow participants to talk about issues that may not usually arise from individual interviews. For example, there may be issues that the individual interviewee may forget during the course of the interview, which other participants in an FGD might bring up. Email survey questionnaires, meantime, may not be able to paint the real pictures since respondents would only be those who have email addresses, and not everyone has one.
3. Respondents may hesitate giving accurate answers depending on who or what group is administering the survey. They may feel more comfortable, or withheld information depending on their level of comfort with the one conducting the survey. For example, if the respondents have had previous unpleasant experience with the surveyor, the tendency is not to provide honest and accurate answers, or maybe even avoiding the survey altogether. On the other hand, if respondent has already established trust towards the person or the group conducting the survey, they would tend to be more engaging or pleasant in their answers.
Survey instrument design on the other hand may stem from the manner by which a survey is conducted. For example, absence of an email address for every potential respondent may result to inaccurate results. If the surveyor wants to measure the service quality of a supermarket and only sends the questionnaires to registered email addresses of frequent customers, there would be biased answers. Same goes if the survey is featured in a certain website only. Of course, only those who have accessed the website can participate in the survey. There will be no chance for people who are not internet savvy to share their experiences even if they are actually frequent