Why do qualitative fieldwork

1
Why do qualitative fieldwork?
Qualitative Basis of Survey ResearchScott
McIntosh, PhD

• Qualitative methods provide an opportunity for
  in-depth exploration of a topic
• Quantitative methods (like survey research) can
  rarely collect enough information to provide an
  adequate analysis of context
• The language and concepts of the participants is
  not generally known to researchers
• Qualitative methods allow the researcher to
  collection information unbound by methodological
  rigor

2
Increasing specificity with transition from
qualitative to quantitative methods
Qualitative methods, unstructured
Quantitative methods, structured,
systematic POParticipant Obs., KIIKey
Informant Interviewing, FGFocus Groups
3
What are the dominant features of qualitative
methodologies?

• The approaches are frequently open-ended
• Go into research situation without preconceived
  ideas and concepts
• Labor- and time-intensive
• Qualitative approaches generally take a lot more
  time than do quantitative approaches
• Emic in orientation
• Strive to learn more about the participants own
  perspectives

4
Qualitative Methodologies

• Watching behavior
• Participant observation (PO)
• Engaging in dialogue
• Key informant interviewing (KII)
• Observation and dialogue
• Focus groups (FG)

5
Participant observation

• Long considered the cornerstone of effective
  field work
• Involves long-term immersion in the lives of the
  people being studied
• Generally unsystematic, unstructured
• Data consist of notes and observation
• Researcher needs to be aware of own biases and
  culture

6
What does PO provide/not provide to the
researcher?

• PO provides
• Deep level of information
• Exposure to culture and lifestyle
• Insight into problems and correlations
• Opportunities for forming hypotheses and study
  questions

• PO does not provide
• Systematic approach to collecting information
• Quantification of things
• Opportunity for statistical analyses
• and it takes a long time to complete!
• CAUTION! Observers who participate, can
  substantially AFFECT the people they are
  observing!

7
PO from Dominican Republic...

• Resided in village for a week
• Somewhat understood conditions of life
• Lived with daily routines
• Lots of interaction on many levels
• Lent insights into attitudes that could be
  quantified in survey

8
Key informant interviewing

• In-depth, semi-structured interviewing
• Generally open-ended questions
• Key informants are those individuals in the
  culture who are capable and willing to share
  information
• Return again and again to key informants for more
  information

9
What does key informant interviewing provide/not
provide the researcher?

• KII provides
• Opportunity for semi-structured inquiry
• Interaction between researchers ideas and
  participants ideas
• Can nail down language issues
• Cultural representation
• Can generate components for survey questions

• KII does not provide
• True quantification
• Clear categories for response
• Unbiased questions and responses

10
Key informant interviewing in Dominican Republic

• Interviewed Comido owner/operator
• Provided information on tobacco products
• Variety sold
• Buying patterns in community
• Informed Survey Approach (e.g., number of smokers
  per household, teen smoking patterns, community
  patterns)
• Informed Treatment modalities (e.g., where people
  congregate and smoke, teen and elderly smoking,
  how people generally quit smoking)
• Etiologic concepts explanatory information
  (e.g., smoking rates across communities)
• Language (terms, slang)
• Themes emerged (Tobacco vs. Cigarettes)

11
Other examples of KII...

• Community Health Improvement Clerkship project
• Potential examples???

12
Focus Groups

• Group discussion
• 8-12 participants
• Semi-structured, moderated discussion
• Some activities - rankings, listings, vignettes
• Data generated is both substantive from
  participants, and observations (interactions)
• Generally video/audio taped and transcribed

13
What do focus groups provide/ not provide the
researcher

• FG provide
• Test setting
• Moderated semi-structured conversation
• Ability to extract key ideas and explanations
• Details for incorporation into surveys

• FG do not provide
• Quantification
• Representation
• Responses in natural settings
• Somewhat unreal public event
• Unbiased participants

14

• EXAMPLE SMOKING CESSATION IN ADOLESCENTS FOCUS
  GROUP RESULTS
• GOAL To examine results of focus group study on
  adolescent cessation as relevant to current
  survey development and treatment implications.
• Study consisted of 12 focus groups with 74
  smokers/ex-smokers ages 14-18 to examine
  adolescent reactions to various types of smoking
  cessation interventions and resources.

15
Goals and Methods

• Adolescent smokers and ex-smokers were recruited
  through flyers placed in schools, social service
  agencies, and community locations, email, and
  newspaper and radio public service spots.
• Focus Groups were 2 hours in length and conducted
  by a professional moderator. Adolescents were
  paid 25 for participation.

16

• Participants were ages 14-18 and were one of the
  following
• Ex-smoker
• Current Smoker Not Motivated to Quit
• Current Smoker Motivated to Quit
• Feedback from adolescents was sought for the
  following areas
• General comments about smoking - warm up
• Communication with physicians/health care
  providers
• Specific Resources
• Quit Goody Packs to be handed out at physician
  offices
• Written materials for TEEN CESSATION
• Quitline (Telephone help-line)
• Website (With CESSATION messages)

17
?
?
?
?
18
OVERALL REACTIONS

• Many teens felt that they were addicted, and
  wanted their difficulties in quitting to be taken
  seriously.
• I dont smoke because its cool. I smoke
  because Im addicted.
• If I ever see my little brother pick up a
  cigarette, Ill kill him!
• Smoking prevention messages were not viewed as
  helpful.

19
PHYSICIAN/HEALTH CARE PROVIDER INTERACTIONS

• Many adolescents reported a lack of trust in
  their doctors.
• Adolescents were not always truthful to their
  doctor.
• Some teens indicated that they lie to avoid
  discussing smoking with their doctor, or because
  they are concerned that their responses will not
  be treated confidentially.
• He always asked me and Id say no because I
  wasnt sure if he was going to tell my mom.

20

• Teens said that doctors didnt ask them if they
  wanted to quit.
• Teens reported that few doctors asked if they had
  a desire to quit or offered tools to help them
  quit.
• They gave me a pamphlet about the dangers of
  smoking. It didnt say anything about how to
  quit.
• Teens noticed if/what their doctors said.

21
WEBSITE Intervention

• Some used internet rarely or never high end
  users expected interesting graphics, interactive
  content, and quick information. Visuals were
  critical.
• It is just too much to read through and not
  enough pictures. If it was a little cartoony
  pictures and fun stuff.
• Few have used it for health information or
  thought they would spontaneously go there to quit
  smoking (link to promotion)
• Teens liked Scare Me button,. interactive
  quiz on smoking facts, Quit Booklet on line .
• Later focus groups showed preference for chat
  feature as a way to ask questions, get assistance
  with quitting.

22
How can you use qualitative methods to do better
survey research?

• PO can provide understanding of the context of an
  event
• Can provide insight, ideas, hypotheses
• Can help provide legitimacy
• Key informant interviewing
• Can further provide details on concepts,
  language, meaningful questions
• Insightful, meaningful relationship - can ask
  anything
• Focus groups
• Lots of detail regarding issues related to study
  questions, interaction effects among
  participants, testing, reactions...

23
Specific applications of qualitative methods to
survey research

• Provides depth of topics to be explored in a
  survey
• Provides logic for questioning
• Language, phraseology
• Insight into interpretation
• Provides case studies to exemplify quantitative
  findings
• Opportunity to validate quantitative findings
• Opportunity to pretest questionnaire

24
Pre-Testing

• Testing the flow of the questionnaire is such a
  matter of intuitive judgment that it is hard to
  describe or codify (Converse)
• Pretest for
• Typographical errors
• Smoiing Research Program
• Content errors
• Are you familiar with some of the top reasons why
  gluttony promotes cardiovascular diseases and
  some of the preventions?

25
Pre-Testing

• Pretest for
• Order of the Questions
• Skip Patterns
• If the skip pattern is at all ambiguous,
  visually, an interviewer may take the wrong
  route and then confident that this is the
  correct way, never carefully read that
  instruction again
• Phases of Pretesting
• Get feedback on structure and flow
• Get feedback on content
• Get real-life process to test all procedures

26
Integrating methods

• ALWAYS consider how and which qualitative methods
  can inform your research
• PO, KII, FG, Surveys
• Qualitative methods ALWAYS provide greater depth
  and insight than do quantitative methods
• In addition to informing your survey, qualitative
  findings are important alone
• Could add time onto your study
• Could use REA (Rapid Ethnographic Assessment)
  approach if necessary. Or RAPs Rapid
  Assessment Procedures.

27
Rapid Assessment Protocol Example
Notes to the fieldworker This worksheet will
help you record information gathered through
observations, site visits, and conversations with
people about beliefs and behaviors around data
collection and analysis of health information
28
Guidelines for asking questions

• Use closed-ended questions to get answers that
  can be categorized.
• Judiciously use open-ended questions.
• Make the question clear (define terms)
• Avoid double barrel questions
• When you last got drunk, did you feel angry with
  yourself and worry that you were adversely
  affecting your health?

29
Guidelines for asking questions

• Respondents must be able to answer the question
• Respondents must be willing to answer the
  question
• Questions should be relevant
• Short items work best
• Avoid being negative
• Avoid biased items and terms