Demography in Focus: Using Focus Groups in Demographic Research

1
Demography in FocusUsing Focus Groups in
Demographic Research

• PRC Research Seminar Series, November 18, 2003
• Cindy Buckley
• cbuckley_at_mail.la.utexas.edu

2
Central Goals

• Overview of focus group usage in academic and
  applied research
• Discuss methodological and ethical issues in
  focus group design, implementation, and analysis,
  highlighting cross cultural research
• Provide examples of focus group research in
  academic, policy, and private sector work
• Gain access to research and teaching materials on
  focus groups

3
Origins and Background

• 1960s gains popularity in survey construction
• 1970s gains in marketing research and consumer
  studies as well as academic research
• 1980s increasingly independent method of data
  collection
• 1990s mixed methods

4
Focus on Population

• First JSTOR article, 1979
• 80-89 emphasis on fertility
• 90- present expanding to migration, health care,
  aging, educational choices, and other topics
• Major journals Family Planning Perspectives,
  International Family Planning
• 21 indexed since 2000

5
Areas of Growth

• Academics increasing grant activity
  incorporating qualitative research, more focus
  group work in DHS, especially in cross national
  research
• Umberson, Rose, Williams, Pullum
• Business/Private sector marketing effectiveness,
  candidate appeal, market identification
• Evaluation Research stakeholder perceptions,
  changing norms and attitudes, Rapid Assessment
  Procedures

6
Practicality

• Cost effective small scale research
• Excellent part time job opportunities
• Lots of options out there
• Professional organizations and certification
  opportunities are accessible
• Monster.Com 1040 job listings for Focus Group
  analysts and moderators
• NYT classified- 72 postings last week

7

• Presbyterian Church (U.S.A.).  Research Associate
  in the Research Office of Presbyterian Church
  (U.S.A.) to direct evaluation and social research
  using focus groups, mail surveys, and secondary
  analysis.  Requires a Ph.D. in one of the social
  sciences knowledge of or willingness to learn
  about the Presbyterian Church and experience in
  sampling, questionnaire construction, social
  science research methods, and report writing.
   Salary 45,000 minimum. Excellent fringe
  benefits.  Review of applications begins January
  5, 2003and continues until position is filled.
   Please submit vita (or resume) to Research
  Services, Presbyterian Church (U.S.A.), 100
  Witherspoon St.,Louisville, KY 40202-1396, or
  research_at_ctr.pcusa.org.

8
How Useful Are They?

• Gerald Zaltman, How Consumers Think
• Unconscious thoughts are the most accurate
  predictors of what people will actually do in the
  space of 5 to 10 minutes in a focus group you
  cant possibly get at one persons unconscious
  thinking

• David Morgan, Focus Groups
• The steady increase in the use of focus groups
  over the past decade clearly demonstrates that
  sociologists and other social scientists have
  found then to be a useful and practical tool for
  gathering qualitative data.

9
What is a focus group?

• Goal A directed open discussion organized to
  clarify reactions, interpretations, and attitudes
• Elements Protocol, Moderator, Respondents,
  Observers (overt or covert) Record (tape, video,
  reconstruction)
• Use Gain background, clarify models, develop
  measures, test reactions

10
The Rights Kind of Questions

• Exploratory/ Mechanistic
• Public attitudes and norms
• View reaction
• Generate first level quick results
• Obtain shareholder views
• Check measures

11
Examples

• Family orientation among young adults in Central
  Asia
• Attitudes towards and tolerance of racism
• What Good Health means to elders
• Minority access to social institutions
• Reactions to public health campaigns
• Questions benefiting from interactivity

12
(No Transcript)
13
Comparative Orientation

• Maximizing comparability to increase reliability
  and consistency across groups
• Attempting to systematize moderator effects and
  context bias
• Increase the standardization in coding and
  interpretation
• Provide a valid comparison of views and attitudes
  across groups, researchers, moderators

14
Protocol Structure

• Ethics statement and consent forms, full
  disclosure
• Rules of the game- moderator trust
• Ice breaker - build rapport
• Initial questions, less directed less sensitive
• Middle questions, more directed, specific
• Final questions, directed, specific, sensitive
  issues
• Closing, feedback and confirmation
• Repeat ethics guideline
• example undergraduate exercise

15
Protocol Forms

• Multiple question forms
• open what do you think about-----
• directed how does your age group---
• consensus name the 1 most important---
• ranking from this list, rank in order----
• scaling relative to other aspects how
  important is----
• Group tasks can assist in multiple measure
  development
• Follow-ups or probes are critical, and are best
  standardized

16
Perceptual, Cognitive, Emotional

17
Protocol Examples

• Personalistic
• University of California San Francisco,
• Study of Gay and Bi-Sexual Men
• http//goodquestions.ucsf.edu/section3/3c6_men.htm
  l
• Are there differences in the way you meet men for
  causal sex versus for longer relationships?
• Whats the best sex youve had? Describe the
  situation

18
Issues with Personalistic Topics?

• UCSF study young women protocol
• Sex in Coercive Situations    Have you ever
  had sex with someone when you didnt want to?    
  Tell me about it. Tell me about the person.   
       Have you ever pressured someone into
  having sex with you?     Give me some
  examples. Were you successful?

19
Experiential and Attitudinal Questions

• Minnesota Dept. of Health and Human Services,
  long term care
• General Attitude
• If a consumer insists he/she wants to live
  independently but there is some danger to living
  independently, should the government do anything?
  
• Specific Experience
• How did you find out about the long term care
  services your family member/ you would receive?
• http//www.dhs.state.mn.us/agingint/ltctaskforce/l
  tcmtg5.17FocusGrps.htm

20
Moderators

• Potentially large influence, should be
  accessible to the respondents
• Extremely familiar with protocol and time
  schedule
• Clear in terms of precise goals and preliminary
  coding scheme to be used
• Should stick with protocol primarily to maximize
  comparability
• Training and linguistic skill is critical

21
Respondents

• 5 to 8 ideal, up to 10 possible
• Target the audience to minimize distance
• Anonymity is vital for findings and ethics
• Familiarity with topic should be similar
• Stigmatizing behavior can be an important
  sampling aspect
• Take care with vulnerable populations

22
Observers

• Covert versus overt
• Sending questions in
• live versus video
• Open to answer questions
• Concerns similar to moderators

23
Records

• Link of conversation with any questionnaires
• Transcription versus reconstruction
• Use of video cameras
• Collection of moderators notes and observer
  evaluations, best if standardized
• Respondent evaluations
• full packet sampling, protocol, training,
  reconstruction, notes, final evaluations, summary

24
Human Subjects

• Approval is important
• need justification
• recruitment strategy
• protocol
• consent forms
• payment explanation
• confidentiality discussion

25
Submit all your forms

• ????? ????????
• ?? ?????????? ??????? ??????????? ? ????
  ?????-??????. ???? ??????? ???? ????????? ?? ??,
  ??? ?? ??????????? ? ??????????? ? ????? ?
  ???????? ?????????? ????????? ?? ???? "????? ?
  ?????????? ??????? ??????????????? ? ???????? ?
  ??????????? ????". ???? ?? ????? ??????, ????? ??
  ?? ?????? ??????????? ? ????????, ?????? ????????
  ???. ?? ?????? ?????????? ???? ??????? ? ????
  ???????? ? ????? ?????.
• ___________________________________________
• ??? ?????????
• ??????? ????????? ____________       ????
• ??????? ??????????___________ ????
• ??????? ?????????????__________ ????

26
Cover all bases

• Sampling
• Protocol
• Moderator training
• Confidentiality
• Storage of consent forms
• Likely result of the research
• ANY ethical concerns

27
Evaluation and Coding

• Comparison of Groups or Individuals
• Interactivity issues
• Respondents individual or embedded
• Depends on researcher and on question type
• Interpretation
• Mechanism
• Measurement
• Scaling

28
Approaches

• Textual analysis
• Spatial analyses- networks
• Decision tree modeling- scenarios
• Conjoint analyses- comparison based on
  demographic attributes
• Perceptual mapping scaling the responses and
  organizing into categories

29
Examples

• DHS studies
• University of Michigan PSC archive
• Journal articles and UT Sociology researchers
• USAID and World Bank- RAP approaches
• Marketing agencies and political interest groups

30
Contacts

• http//www.focusgroupmoderators.com/
• Useful set of articles from a market research
  firm
• http//www.decisionanalyst.com/index.asp
• Home page of decision analyst, a major marketing
  firm, downloadable software and papers
• http//qrca.org
• Home page of the Qualitative Research Consultants
  Association

31
Projects using focus groups

• Intercultural Project An assessment of study
  abroad experiences in Britain
• http//www.lancs.ac.uk/users/interculture/deliver1
  2.htm
• California firm that will pay you!
• http//www.onsurvey.ca/focusgroupstudies.html

32
New Areas

• Expanding out of fertility oriented work
• Moving onto the web and virtual groups
• Use of computers- digitalization