Yesterday, I conducted a very rough pilot interview for mainly one purpose: to get a sense of how long it would take to have a respondent answer all the questions in one sitting. It was very rough because the respondent did not match my target population well. He would have been disqualified for two reasons. First, I know him (too) well. Second, he had lived in the target country before and had family/friend living there. Because of these two reasons, I already knew about 1/3 of his answers well and about another 1/4 could not be answered, thus the interview probably went faster than with someone who fit the target population description better.
Anyway, I learned that the fastest a one-sitting interview would probably take is 45 minutes. I would estimate that it would take at least 90 minutes with a fully qualified participant. My pilot study from last summer, showed me that about half of the participants elaborated or expanded upon their answers, thus adding 50-100% more to the allotted time. So a very rough estimate would be 45-90 minutes for uncooperative or reserved interviewees and 90-180 minutes for cooperative and/or talkative interviewees.
Fortunately for me, I do not plan to ask all the interview questions to the respondents in one sitting. The only way this would happen is if the respondent prefers to get it all done in one sitting and that I have at least 3 hours of my own free time. I do not expect this to happen often if at all.
I prefer to do a hybrid of synchronous (web-conferencing) and asynchronous (email) interviews. My pilot respondent yesterday told me he would have preferred to answer all by email, which would help me in terms of not transcribing any audio recordings. But I do like to engage in a little spontaneous conversation about the topic, especially when talking about concepts that can lead to ambiguous answers. From this interview, I learned that I should give my participants the option to do more web-conferencing or more emailing.
However, I'm not confident enough yet, so I plan to conduct a few more interviews using both formats so I can find my level of confidence.