THE VULNERABLE RESPONDENT

1
THE VULNERABLE RESPONDENT

• Rapport, Meltdown, Types of Respondents,
  Sensitive Subjects

2
RAPPORT

• Introductory letters and phone calls
• Institutional support (the Rutgers name)
• Opportunity for respondent to help (a struggling
  student)
• First impressions
• Taking the time (not often possible in the class)
  to bond with the respondent
• Women interviewers usually do a better job of
  this than men interviewers but there is no
  intrinsic reason why both sexes cannot do this

3
MELTDOWN

• Crying
• Incremental nervousness (not always manifested in
  refusal to answer questions)
• Keeping your eye on the non-verbal cues
• Flashbacks
• Need for backup psychological help. Dont try to
  play psychiatrist except perhaps in first-aid
  situations

4
DELAYED REACTION MELTDOWN

• A seemingly normal interview may terminate
  without incident and the respondent may not
  experience a reaction for as much as a week
  afterward
• For particularly vulnerable topics/respondents, a
  routine call-back may be advisable.
• In most cases, simply providing the respondent
  with post-interview contact info is sufficient

5
INTERVIEWING EXPERIMENTAL SUBJECTS

• Importance of replicating conditions and wording
  of questions exactly
• Interview as soon after the experiment as
  possible
• Such interviews have a debriefing component as
  well as a data collection component. If there
  was any deception in the experiment, the true
  purpose must now be revealed.

6
INTERVIEWING RESPONDENTS

• Populations and samples
• The most common form of interview for many
  researchers, the only kind of interview
• Assessment of the preciousness of the case
• Large-N, Small-N
• Set-in-concrete protocols

7
INTERVIEWING INFORMANTS

• The informant doesnt represent an underlying
  population but is usually atypically
  well-informed about the population.
• Almost all questions can be open-ended.
• No need for consistency among informants.
• No need for all informants to be asked the same
  questions.

8
INTERVIEWING SPOKESPERSONS

• Most similar to journalism
• Battle of wills. The spokesperson is usually not
  interested in representing the population
  honestly but presenting a version of collective
  self that is being promoted by the collectivity

9
ILLEGAL ACTIVITIES AND CRIMINAL HISTORIES

• The wording of the question embodies tacit
  assumptions that will be clear to most
  interviewees.
• These embody both knowledge and expectations.
  (WHAT IS NORMAL? The social desirability effect)
  Especially a problem with frequency questions
• Try asking ever question first then
  currently question.
• The how many times approach can create
  confidence especially when using a checklist. But
  beware of assumptions built into the categories.

10
SEXUAL SUBJECT MATTER

• Great variance in comfort in talking about sexual
  matters to a stranger. Aim at the most prudish
  to avoid sampling bias
• Laumann study found surprisingly weak differences
  according to the gender of the interviewer/intervi
  ewee
• Masturbation the most difficult sexual subject
  matter to ask about

11
WHEN YOU THINK THE RESPONDENT IS LYING

• Confrontation is not advisable except with
  spokespersons.
• Probing for more detail can be a two-edged sword
• Suspicions must be marginally noted (especially
  when the interviewer is not the investigator).

12
WHO GETS TO SEE YOUR DATA?

• Importance of keeping all identifying information
  separate from the data
• Recent tendency to subpoena entire hard drives
• Use of secret concordance key
• For non-longitudinal research, it may often be
  best to destroy all identifying information soon
  after data collection and verification are
  completed. Not even a single copy for the PI is
  retained.