PHE’s Work

I am not going to talk about “hard-to-reach” population in this post. I am interested in these studies on these hard-to-reach populations or hidden populations.

http://www.laria.gov.uk/content/features/68/feat1.htm

(This URL links to a seminar held by LARIA on 15^th. Nov, 2001. I never heard of it before and have no idea which category it should go in, therefore I record it here. This groupd charge about 30 pounds for each its publication. For me, these are hard-to-reach documents. )

It’s said from the seminar that “Most research with hard to reach groups has to be qualitative, as these groups do not usually figure strongly in quantitative surveys.” That’s true.

What’re requirements for a good quantitative research?

• A population with known mean and variation; or
• A population with unknown mean and variation but a representative and large enough sample can be obtained.

Instead of defining “hard-to-reach” population, identifying characteristics of the target population is more important. One of key characteristics of a hidden population is “No sampling frame existed, and thus the size of the membership and group boundary is unknown.” The other two are acknowledgment of belonging to the group is threatening, because membership involves being the object of hate or scorn and sometimes fear of prosecution; and members are distrustful of nonmembers, do whatever they can to avoid revealing their identities, and are likely to refuse to cooperate with outsiders or to give unreliable answers to questions about themselves and their networks (Heckathorn 1997; Benoit et al. 2005).

Is it the destiny? All qualitative studies targeted on hidden populations are limited in terms of making generalizations about the entire population since a random sample of research participants is too hard to achieve. There is no a “work-in-all” strategy or method. It’s the same in quantitative research but not so exclusive.

(When MRC said “a study should be generable”. How do they evaluate generalization? To which extension? In which way? I so should think about this. )