A: We've used personal interviews, computer-assisted telephone interviews, self-administered surveys, and computer-based surveys. There are pros and cons to each.
One advantage of face-to-face interviewing, is that with a skilled interviewer, you get higher quality data: fewer errors, no items get missed, and you can get more in-depth answers to open-ended or semi-structured questions. This kind of interviewing also allows fairly complex skip patterns. For example, in studies of sexuality you can skip children over questions that aren't appropriate for their ages or levels of experience.

A: The main drawbacks are lack of anonymity and expense. Kids may be less inclined to be honest about their behaviors in a face-to-face interview. It's also not as efficient as other methods; it takes more time and money, you need to make the effort to train your interviewers well, and you have to enter and clean the data.

A: In Computer Assisted Telephone Interviews (CATI) the survey is programmed on a computer and the interviewer asks the questions as they appear on the computer screen. As the name indicates, the interview is given over the phone, rather than face-to-face, thus it's easier on the budget. Because it's computerized, you can change the direction of the interview based on participant response, skipping over inappropriate questions. The data are generally high quality ñ few missed items, for example. Data entry costs are negligible since answers are directly recorded on the computer as the respondent gives them. But there are drawbacks: it's not anonymous, so you may get fewer people admitting to high-risk behaviors.

A: Again, less expensive ñ you can mail them out, or pass them out at school or other groups. You have the advantage of anonymity and confidentiality with these kinds of questionnaires, so you may get more accurate data. The disadvantages are missing data (people may skip questions they don't want to answer or by accident), time spent entering and cleaning the data, and dealing with open-ended questions which may or may not result in the data that you are looking for.

A: We get very high quality data with this method, because kids seem more willing to report high-risk behaviors when they are answering questions on a computer. You display questions on the computer screen, and participants enter their responses directly. The advantages here are that it allows complex skip patterns, which has been helpful to us when we are studying sexuality, for example, and want to skip some questions based on the participant's age, or answers to previous items. You can also include audiovisuals. Also, because answers are input directly by the participant, data entry and cleaning are simplified.

A: The obvious disadvantage is that the participants need to have some level of computer skill and need to feel comfortable answering questions on a computer. You also need to consider the initial costs of purchasing the equipment, which can get quite expensive.