Sample size for focus groups

[Byrd Sellers, Johnita]

Hello All,

 

I was hoping that someone could point me in the right direction regarding the following:

 

I am involved in a qualitative research project.  To answer the exploratory phase of the research question, we would like to put together several different focus groups.  Assuming that each focus group contains about 10 – 12 people, I was unsure how many focus groups would be necessary to best address the geographic distribution for a national study.  In other words, how many focus groups are needed to provide a nationally representative sample?  Any advice/references that can be provided pertaining to the size of the focus groups and/or the ENTIRE sample size (which includes the number of total focus groups needed), I would greatly appreciate it.  Thanks in advance for your help.

 

Sincerely,

Johnita

 

[Muir Houston-2]

http://www.bmj.com/cgi/content/full/311/7000/299

 

http://sru.soc.surrey.ac.uk/SRU19.html

 

both of these and the cited references should provide you with what you need

 

Muir

---

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Byrd Sellers, Johnita
Sent: 29 April 2009 14:03
To: [hidden email]
Subject: Sample size for focus groups

Hello All,

 

I was hoping that someone could point me in the right direction regarding the following:

 

I am involved in a qualitative research project.  To answer the exploratory phase of the research question, we would like to put together several different focus groups.  Assuming that each focus group contains about 10 – 12 people, I was unsure how many focus groups would be necessary to best address the geographic distribution for a national study.  In other words, how many focus groups are needed to provide a nationally representative sample?  Any advice/references that can be provided pertaining to the size of the focus groups and/or the ENTIRE sample size (which includes the number of total focus groups needed), I would greatly appreciate it.  Thanks in advance for your help.

 

Sincerely,

Johnita

 

[Hector Maletta]

Johnita,

For a qualitative exploratory study you do not usually require a representative random sample of focus groups. The groups are usually selected in some fashion that covers the various regions and socioeconomic strata you intend to cover in your study.

However, if you want a representative national sample there are several ways to do it. One normally divides the country in several strata (e.g. states or groups of states or regions), then selects a sample of smaller areas (clusters) within each stratum (e.g. counties), then probably some minor unit inside the selected cluster (communities within counties), and finally uses some procedure to choose the members of the focus group within each site. There are standard formulas in sampling techniques textbooks to estimate the standard error of estimates based on the variability of key known variables across states, counties or communities. For this purpose each group is a sampling unit. To get a reasonable standard error with this kind of model, sample sizes are usually in the order of 800-1000 cases (groups in your case). But it all depends on the degree of variability between and within strata and clusters.

The size of the focus group does not matter much for sampling purposes, though sizes of 8-12 people are usual to get enough variation within groups.

I repeat: in an exploratory qualitative study you probably don’t need a sophisticated statistical model. On the other hand, if you incur the cost of a nationwide statistically representative sample, why doing a mere exploratory qualitative study? One does an exploratory study when the subject matter needs exploring; but perhaps the pertinent literature provides enough insight, prior knowledge and conceptual framework for you to design a proper, more advanced investigation.

Hope this helps.

Hector

 

---

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Byrd Sellers, Johnita
Sent: 29 April 2009 10:03
To: [hidden email]
Subject: Sample size for focus groups

 

Hello All,

 

I was hoping that someone could point me in the right direction regarding the following:

 

I am involved in a qualitative research project.  To answer the exploratory phase of the research question, we would like to put together several different focus groups.  Assuming that each focus group contains about 10 – 12 people, I was unsure how many focus groups would be necessary to best address the geographic distribution for a national study.  In other words, how many focus groups are needed to provide a nationally representative sample?  Any advice/references that can be provided pertaining to the size of the focus groups and/or the ENTIRE sample size (which includes the number of total focus groups needed), I would greatly appreciate it.  Thanks in advance for your help.

 

Sincerely,

Johnita

 

[Byrd Sellers, Johnita]

Thanks everyone for your help.  Your references and input have been MOST helpful!

 

Johnita

 

 

[Martin P. Holt-2]

Dear Johnita,

 

I note that you ask to cover the geographical distribution, and Hector has answered in this vein. How you go about splitting populations into focus groups might depend on what the research question is. Sticking with the geographical example, if your research question was about alcoholism, you might expect national differences (maybe not, it's just an example). Or if there was no reason to expect a geographical effect, choosing your population would depend on what the research question is about, before you form your focus groups. In summary, I would think that you need to target your focus groups on the research question, and as Hector says, not worry too much about demonstrating a sample size calculation. But in the process of targetting the research question, I would expect you to sense what you needed to do regarding the positioning of the focus groups and how many to use where, from the research question.

 

HTH,

Martin

----- Original Message -----

From: [hidden email]

To: [hidden email]

Sent: Wednesday, April 29, 2009 2:03 PM

Subject: Sample size for focus groups

Hello All,

 

I was hoping that someone could point me in the right direction regarding the following:

 

I am involved in a qualitative research project.  To answer the exploratory phase of the research question, we would like to put together several different focus groups.  Assuming that each focus group contains about 10 – 12 people, I was unsure how many focus groups would be necessary to best address the geographic distribution for a national study.  In other words, how many focus groups are needed to provide a nationally representative sample?  Any advice/references that can be provided pertaining to the size of the focus groups and/or the ENTIRE sample size (which includes the number of total focus groups needed), I would greatly appreciate it.  Thanks in advance for your help.

 

Sincerely,

Johnita

 

[Arthur Burke]

Johnita ... A number of national opinion polls use a sample size of approximately 1100.  That sample size yields the familiar error band of +- 3%.

 

Art

------------------------------------------------------------------

Art Burke
Northwest Regional Educational Laboratory
101 SW Main St, Suite 500
Portland, OR 97204-3213

Phone:  [phone]

Fax: [phone]

[hidden email] 

---

From: Byrd Sellers, Johnita [mailto:[hidden email]]
Sent: Wednesday, April 29, 2009 6:03 AM
To: [hidden email]
Subject: Sample size for focus groups

Hello All,

I was hoping that someone could point me in the right direction regarding the following:

I am involved in a qualitative research project.  To answer the exploratory phase of the research question, we would like to put together several different focus groups.  Assuming that each focus group contains about 10 – 12 people, I was unsure how many focus groups would be necessary to best address the geographic distribution for a national study.  In other words, how many focus groups are needed to provide a nationally representative sample?  Any advice/references that can be provided pertaining to the size of the focus groups and/or the ENTIRE sample size (which includes the number of total focus groups needed), I would greatly appreciate it.  Thanks in advance for your help.

Sincerely,

Johnita