Include two to three groups to represent each perspective

Include two to three groups or individuals to represent each
perspective or comparison group
You will need two to three focus group discussions to represent a particular
perspective. The number of interviews or discussions required depends on the level
of representation and types of comparisons you desire from the data. Here, we use
the example of focus group discussions. The same guidance applies when using
semistructured interviews, direct observation or other qualitative data collection
  Review the information needs and required comparisons stated in your analysis plan
to determine the overall number of groups or individuals needed.
If you plan to simply represent the survey population as a whole, conduct two or
three focus group discussions.
If you plan to compare the perspective of communities where a particular
intervention was highly successful with the perspective of communities where
the same intervention was much less successful, plan to conduct two (or three)

focus group discussions in the more successful communities and two (or three)
focus group discussions in the less successful communities.
If you plan to compare the perspectives of men and women regarding the current
obstacles for education for girls in the community, hold two or three focus
groups with women and two or three focus groups with men. Men and women
are likely to havevarying viewpoints on this topic and to collect data from only
one or the other would not represent the full range of current obstacles.
  Itis important to disaggregate the groups based on the characteristics that are likely
to influence their opinions or perspectives on the key issues or topics to be discussed.
If socioeconomic status could potentially influence participants’ perspectives on the
availability of water in the community or the barriers to the education of girls in the
community, then hold separate focus groups with participants from higher and lower
socioeconomic backgrounds. If you do not separate these groups, the data would not
showwhich obstacles were specific to which socioeconomic group. Consider which
characteristics or factors (e.g., gender, age, socioeconomic status, type of livelihood), if
any, are relevant for your discussion topics when deciding whether or how to
disaggregate the participants.
Consider which participants should represent a certain perspective or are most likely
to give reliable information. For example, if you are interested in understanding
more about care-seeking practices for children under age 5, conduct focus group
discussions with, or interview, mothers and caretakers of children under age 5. If
you are interested in local agricultural practices, hold the discussions or interviews
with persons involved in agriculture.
Determine the appropriate method for selecting focus group or interview
participants. Common methods include asking community leaders to help select
participants and asking individuals with the desired characteristic (e.g., mothers or
caretakers of children under age 5) to help identify additional participants. It is
important that the method chosen does not only yield participants from the same
family or social group (unless your methodology specifies it).
Ensure the exercise does not exclude marginalized groups. Consult your team to
determine which groups are likely to be marginalized in your target areas. Ensure
that members of these groups are included in the discussions or interviews or, if
more appropriate, hold separate discussions or interviews with members of the
marginalized groups only. Explain the reasons you would like to include the
perspective of these groups to community leaders so that they will not feel
threatened by their participation and possibly assist in locating these households.
  Include a description of your selection methodology in your report. Be specific
about how and why you choose sites and participants. Include any possible biases in
your selection method in the limitations section of the report. Be honest and remember
that many surveys encounter one type of limitation or another.

Md. Kaysar Kabir

Click Here to Leave a Comment Below 0 comments