1. Don't put words in people's mouths. In fact, talk as little as possible. A pause that's a bit longer than a pause would be comfortable in everyday conversation can work wonders in provoking further insights from a respondent. It signals that you're waiting for them to say more, and gives them time and space to think more deeply. (See? Awkward pauses aren't a reflection on your social skills. It's a research technique.)
Yes, you have to say things sometimes, but you don't have to say a lot. Often just saying "Hmm" will provoke an interviewee to expand on an idea or offer up new information. "Can you say a little more about that?", repeating an interesting word mentioned by the respondent in a questioning tone, briefly re-stating what the respondent just said to make sure you're understanding can also all be interview gold.
2. Try to avoid positive or negative feedback. Listening to lots of interviews about people's experiences with illicit drugs gave me a clearer idea of why this can be a good guiding concept. If an interviewer says "That's wonderful" in response to someone's description of a recent attempt to quit using a substance, the respondent may react by emphasizing quitting attempts and downplaying current use, for example. There's no getting around the feedback loop between an interviewer and a respondent, and that loop can be an important part of analysis, but avoiding positive or negative feedback as much as possible can often produce information that is less aligned with the interviewer's biases.
3. Expect to be surprised. But don't expect anything else. A difficult part of the notion that interviewers should avoid positive or negative feedback is that not getting much feedback can make some people shut down. Also lots of people use expressions with a positive inflection like "That's great" as conversational support.
For me the word "interesting" can be a good compromise - partly just because it's a word I say too much anyway. What's not interesting? But "interesting" can also be a way to let people know that you want to hear more without expressing a judgment about the content of what they're saying, and without imposing "common sense" expectations about what people are saying or how they're interpreting it. If I say someone's attempt to quit using a substance is "great," I might not learn that, for example, quitting made the respondent realize all the things they enjoy about the substance and that they never want to quit using it again.
4. Respondents are the experts on their own experiences. You are just there to listen and learn. It can be helpful to say something at the beginning like "I might ask some questions that seem really dumb or obvious. That's just part of the interview process. You're the expert, and I want to try to understand how you see things, in your own words." You don't necessarily have to ‘lead' an interview. Sometimes what seems like a tangent in the moment ends up being the most meaningful part of an interview.
But yeah, you do have to guide people, and manage how much you need to cover in a limited amount of time. In an interview where the interviewer and respondent can see each other, body language can work really well for this. Although introverted or shy people may not always be great with body language, it can be more manageable when you understand your role as primarily listening, and the spotlight is not on you. One visual cue that I use in interviews is hunching down and making myself visually lower than the respondent. They're the expert. When I need to jump in and it's hard to find a good pause, I sit up straighter, maybe shuffle around some papers. This usually produces a pause and I can jump in without it feeling like a jarring interruption.
5. Recruiting complete strangers can totally work. When I've needed to "get out and meet people" for a recruiting effort, I have found that handing out flyers in person has worked best for me. (A) I don't have to give a spiel. (B) The people I'm encountering don't have to waste time trying to figure out whether/how to engage with the possibly crazy person's spiel, but can just read the flyer if they feel like it (assuming literacy or language differences aren't an issue) and then let me know if they're interested. If you're a smile-y sort of person, smiling can help, or just generally looking approachable. You don't necessarily have to be the life of the party who has mastered circulating and networking though. People may find it easier to approach you on their own terms.
Rachelle Annechino (22-3-2012) Interviewing for Introverts (post) Ethnography Matters