Frequently, focus groups complement a survey -- but sometimes you must choose between them. The actual situation can be very complex, but here are a few general guidelines.

Surveys and Focus Groups Use a survey when: Use a focus group when You need quantitative estimates You need in-depth insights into why people behave as they do, or into the subtleties of a process You need confidence intervals or estimates You can live without (many) quantitative estimates You can get a high response rate (or you can expect respondents and non respondents to be similar) Statistical representativeness is not a necessity You have a clear idea of the questions you want to ask and how to ask them. You want the flexibility to pursue issues, reasons, behaviors, concerns that are "discovered" during your inquiry. You can afford the cost and time of questionnaire design and testing, sample design, and processing of results. Time and/or budget is limited.

The advantages of conducting in-person surveys include:

• Good for surveying people with poor reading and writing skills
• Interviewers can probe for in-depth answers
• Best method for hard-to-reach populations
• Interviews can use visual questions, e.g. pictures for children
• Interviewer can build rapport with respondents
• Interviewer can collect data in addition to the questionnaire

There are, of course, disadvantages as well:

• Slow completion time
• Costly
• Requires trained interviewers
• Some locations may be hard to reach
• Requires clusters of respondents
• Interviewer bias can contaminate results
• Respondents may refuse personal contact

The advantages of telephone surveys include:

• Faster than mail or in-person surveys
• Higher response rates than mail surveys
• Good for open-ended questions
• Less costly than in-person surveys
• Permits some probing

The disadvantages include:

• Omits those who do not have phones
• Requires well-trained interviewers
• Interviewer bias can contaminate results
• More costly than mail surveys

The advantages of mail surveys include:

• Can be completed at respondents' convenience
• Information can be looked up
• Privacy in completing survey
• Relatively inexpensive
• Large numbers can be done in a short time
• Doesn't require trained interviewers

The disadvantages include:

• Response rates are lower than phone and in-person surveys
• Takes longer to complete than telephone surveys
• Requires accurate addresses
• Requires high quality questionnaires:
  • need to be self contained
  • need to be complete
  • no chance for probing questions

Certainly. Like all other ways in which surveys can be administered, there are both advantages and disadvantages:

The major advantages are that it is quick , can be done at respondents' convenience, and software is available that allows fax to be used as scanner, eliminating the data entry process.

The disadvantages are that it requires accurate fax numbers and respondents must have fax machines.

Non-response is the single biggest hurdle for surveys, especially for mail surveys. It is hard to get high response rates. Within some offices in the Department of Education, some researchers will not use data unless the response rate is at least 70%. Other people and/or organizations have different standards. As a general guide, however, anything below 50% is too low to be reliable and anything above 85% is marvelous. If 70% is okay, we can, generally speaking, view 60% as barely acceptable.

Getting good response rates is the goal that you, every direct marketer, and every evaluator in the country share. First, write a good letter to accompany the questionnaire. If possible, personalize the cover letter. Include return postage (a commemorative stamp elicits a better response than a more boring stamp) and keep your package neat and attractive. Try using colored paper for the cover of the questionnaire(s). Let the respondent know there is a deadline. Use reminders. Resend a complete package, (your first reminder) with a different cover letter, if you have not received a response. Make your second reminder a postcard or letter. Follow-up by telephone. Offer to complete the questionnaire over the phone with the respondent.

It frequently turns our that things take longer and cost more than expected. Fortunately, there are techniques that may be employed to reduce the cost and time that may yield a high degree of accuracy. If you can't afford to conduct follow-up for a large group do the following:

• Split your sample into two subsets, one large and one small
• Follow up intensively on the small group
• Compare your respondents and responses from the large, low response group and the small, high response group
• If they are similar, combine results.
• If they are not similar seek advice from a statistician to see if adjustments can be made.