Hi,

I have found vignettes to be most useful in such circumstances. It will also be important to carefully consider the kinds of people who comprise your focus groups. Ensuring homogeneity amongst these groups will be of key importance. I have also found carefully setting out the purpose of the study, and the guidelines of the focus group as being of the utmost importance. People who participate in research like this want to be helpful, so just be as specific as you can be about what you are looking for from them. Good luck, sounds like a great study!

I've also followed Meagan's suggestion of using vignettes but another strategy we use is to give participants a packet/open ended survey with key FG questions that we ask participants to complete by writing their initial responses down and then discuss as a group.  The packet responses provides an opportunity to do some informal member checking when comparing the discussion to the written responses, and the FG discussion provides the opportunity for analyzing interaction of participants.

Good suggestions from both Meagen and Jodie. When I do what Jodie describes, I have an assistant tabulate a summary of the results on large sheets of paper during the instructions. then the participants discuss the overall pattern of results as the first question.

In addition, lot of it depends on your instructions and initial questions. I explicitly tell participants about the problem of suppressing out their ideas and feelings when the group is saying something else, and stress that the reason why we have people in a group is so that we can hear as many different opinions as possible. Then I match that with a first question that is both interesting to the participants and not too personal, so that they can begin sharing things with each other.

If you want to find out how big a problem not sharing "honest" feelings was, then a brief confidential (i.e., no names) questionnaire can give you data.

The other thing you have to expect is that if you interview a group of heterosexuals about heterosexual behavior, they are indeed likely to express support for heterosexual values. This is not a "distortion" caused by talking in a group; instead, this is what groups of people do. The idea that there is some kind of "context free" set of "true" beliefs that people somehow "have in their heads" is a myth. So, you have to accept the conclusion that if you ask people to talk in a group, they will indeed act like they are in a group. (And individual interviews that involve self-disclsoure to a researcher have their own "distortion" issues.)

So, think about the context that matters the most to you, and accept the limitations that come with whichever form of interviewing you do.

After an opinion has been stated and reinforced by another group member, I ask for "other points of view" or "other perspectives on this." It's just another way of letting people know that multiple perspectives are valued. I have found this to be especially important with groups of women, who may be less likely to openly disagree with each other. And I agree with my colleagues that the initial instructions are key, and can be reinforced regularly.

If you are looking for ways prejudice leads to "positive" behaviors, it seems to me that people will speak freely about how they "help those people out" if you phrase your questions "correctly." No?

These are all extremely helpful contributions so far! Thank you for your input, everyone.

maybe reconsider focus group as an instrument here. Looking at your research interest, I would consider the trade-offs to high you need to bring in with vignettes and written initial contributions. This somehow kills what makes focus group such a useful instrument: the vivid group discussion. I would strongly support David's thoughts on the contextuality of the phenomena to study. Groups do things like playing by the rules of social desirability.