Structured Survey Interviewing: Telephone and In-Person Surveys

1
Structured Survey Interviewing Telephone and
In-Person Surveys

• Behice Ece Ilhan
• Chihchien Chen

2
Outlines

• Interviewer factor
• Interviewing techniques
• Theories and methods of telephone surveys
• Does Conversational Interviewing Reduce Survey
  Measurement Error?
• Future research directions

3
Interviewer Factor

• Who is an interviewer?

• Interviewer vs. researcher
• Interviewer as gatekeeper
• Interactive nature of interviews ( Cannell et al
  1981)

Why is interviewer a factor in survey research?

• Response rates
• The accuracy of reporting
• Consistency or precision of measurement
• Interviewer expectations

• Role expectations
• Attitude structure expectations
• Probability expectations

4
Interviewer Factor

• Role of interviewer

• To locate and enlist cooperation of selected
  respondents
• To train and motivate respondents
• To ask questions, record answers, and probe
  incomplete answers

• Why are these skills particularly important for
  telephone surveys?

5
Interviewer Factor

• Standardization to control for the interviewer
  factor
• What is standardized interviews?
• Standardize what?

• Five aspects of interviewer behavior that the
  researcher tries to standardize
• Presenting the study
• Asking questions
• Probing
• Recording the answers
• Interpersonal relations

• Tailoring vs. standardization
• Which one? When?
• Implementation mode
• Socially desirable questions/ knowledge questions
• Interviewer profile

6
Interviewer Related Error

• When do we have interviewer related error?

• Interviewer related error vs. interviewer bias

• How does each affect the quality of data?

7
Interviewer performance

• Training
• How do we rate interviewer performance?
  (telephone vs. face-to-face vs. mail surveys)
• Why is it important?

In Class Activity As a researcher, pretend that
you want to hire an interviewer to conduct your
research. Explain him/her or the group about the
expected roles and performances from the
interviewers. Mention about standardization,
probing, performance measures, tailoring etc.
8
Selecting Interviewers

• The criteria to select

• Experience
• Race
• Religion
• Ethnicity

• Age
• Education
• Part time / Full time

• In Class Activity
• According to the literature, which of these
  criteria are independent of the type of the
  study? In other words, which of them dont cause
  interaction between subject matter and the
  demographic characteristics of the respondents
  and the interviewers?

9
In Class Activity Role Playing

• Presenting a study
• Pretend that you are an interviewer. By the
  researcher, you are assigned to call people about
  a study about counterfeiting products survey. How
  will you present the study? Lets write a script
  together.
• The Study seeks to identify differences in
  consumers attitudes toward counterfeited
  products in fashion, medicine, food etc. (across
  race, gender, education, and age)

Possible informants are elderly, professionals,
young people, housewives etc. Try to tailor the
presentation for different profiles of
informants. What type of informant do you think
will be more challenging? Support your answers
with data from literature. (e.g. difficult to
engage old people)
10
Presenting the Study

• In class activity

11
Interviewing Techniques

• Historical perspectives
• Current perspectives
• Question answering process (diagram in
  cognitive processes)
• Cognitive and motivational difficulties in
  answering questions
• demands placed on informants Deviation from
  process due to some situational cues (or
  sometimes due to personal traits) such as social
  desirability bias, acquiescence bias,etc.
• Response Errors
• Underreporting /Overreporting due to elapsed
  time, salience of events, perceptions of social
  desirability (e.g. hospitalization)
• Standard Interviewing Techniques

12
Interviewing Techniques

• Improving interviewing techniques
• Question length
• What are some findings about question length?
• How do you interpret these findings?
• Response Modeling
• Introduction
• Reinforcement and Feedback
• Commitment
• Pace
• Rapport
• Experiments with the new techniques of
  commitment, instructions, and feecdback

13
Theories Methods of Telephone Surveys (Groves,
1990)

• What is the goal of survey theory?
• What are the factors affecting the quality of
  response?
• Three kinds of sampling frames
• Telephone directories
• Computerized files based on directories
• Area code-prefix frames

14
Theories Methods of Telephone Surveys (Groves,
1990)

• How can social exchange theory be adapted to
  explain nonresponse error?
• How can communication theory be adapted to
  explain nonresponse error?
• Why do people accept or reject a survey request?
• What are the factors affecting the nonresponse
  rate?

15
Theories Methods of Telephone Surveys (Groves,
1990)

• How is communication theory adapted to explain
  measurement error?
• Methods of decreasing the measurement error

16
Does Conversational Interviewing Reduce Survey
Measurement Error? (Schober Conrad, 1997)

• Do we need a standardized procedure? What are the
  pros and cons of a standardized survey?

17
Does Conversational Interviewing Reduce Survey
Measurement Error? (Schober Conrad, 1997)

• Alternatives for implementing flexible
  interviewing
• Type of interaction
• (un)scripted definition
• Why or why not use flexible interviewing? Does a
  flexible interviewing initiate other effects?
• When to adopt a flexible interviewing? It is not
  a panacea.

18
Does Conversational Interviewing Reduce Survey
Measurement Error? (Schober Conrad, 1997)

• If we want to measure social desirable questions,
  telescoping errors, closed-ended questions, and
  threatening and knowledge questions, can the
  flexible interviewing itself change respondents
  answers? If so, how can that happen?

19
Theories Methods of Telephone Surveys (Groves,
1990)

• CATI (computer-Assisted telephone interviewing)

20
Brainstorming

• Since nowadays many people only have cell phones
  as their primary phone contact, how does the
  fewer number of viable landlines affect the
  traditional telephone survey? What are the
  reasons causing survey errors in the situation?
  How would survey errors appear, and how can we
  reduce the survey errors? Would that change
  things?

21
Theories Methods of Telephone Surveys (Groves,
1990)

• How does social exchange theory be adopted to
  explain nonresponse error?
• How does communication theory be adopted to
  explain nonresponse error?
• Why do people accept or reject a survey request?

22
What to take home?

• 1.
• 2.
• 3.
• 4.
• 5. Prof. Shavitt
• This is an interactive class activity, which
  aims to summarize 5 important points of what we
  have learned in the session.