Week 10

1
Week 10

• Data Collection Field Work

2
Objectives

• Describe the field work process
• Discuss the possible errors arising from field
  work
• Explain the management of field work

3
The nature of field work
4
Overview of Non-sampling Errors
Noncoverage Nonresponse
Nonobservation
Not-at-homes
Refusals
Nonsampling Biases
Data Collection Errors Office Processing Errors
Observation
5
Model of Bias in the Interview
Interviewer
Respondent
Background Characteristics Age Education
Socioeconomic Status Race Religion Sex
Etc.
Background Characteristics Age Education
Socioeconomic Status Race Religion Sex
Etc.
Psychological Factors Perceptions
Attitudes Expectations Motives
Psychological Factors Perceptions
Attitudes Expectations Motives
Behavioral Factors Errors in Asking
Questions Errors in Probing Errors in
Motivating Errors in Recording Responses
Behavioral Factors Responses to Questions
(Adequate-Inadequate) (Accurate-Inaccurate)
6
Methods for Handling Non-coverage Errors
Definition
Failure to include some units or entire sections
of the defined survey population in the sampling
frame.
Methods for Handling
1. Improve basic sampling frame using other
sources.
2. Select sample in such a way as to reduce
incidence such as by ignoring ineligibles
on a list.
3. Adjust the results by appropriately weighting
the sub-sample results.
7
Methods for Handling Non-response Errors
Definition
Failure to obtain information from some elements
of the population that were selected for the
sample.
Designated respondent is not home when the
interviewer calls.
Not-At-Homes
1. Have interviewers make advance appointments.
2. Call back at another time preferably at a
different time of day.
3. Attempt to contact the designated respondent
using another approach, i.e., use a
modified callback.
8
Methods for Handling Non-response Errors
Refusals
Respondent refuses to cooperate in the survey.
1. Attempt to convince respondents of the value
of the research and the importance of
their participation.
2. Provide advance notice that the survey is
coming.
3. Guarantee anonymity.
4. Provide an incentive for participating.
5. Hide the identification of the sponsor by
using an independent research organization.
6. Try and get a foot-in-the-door by getting
the respondent to comply with some small
task before getting the survey.
7. Use personalized cover letters.
8. Use a follow-up contact at a more convenient
time.
9. Avoid interesting, but not vital questions.
10. Adjust the results to account for the
non-response.
9
Methods for Handling Data Collection Errors
Definition
When the individual participates in the study, he
or she refuses to answer specific questions or
provides incorrect answers to them.
1. Match the background characteristics of the
interviewer and respondent as closely as
possible.
2. Make sure interviewer instructions are clear
and written down.
3. Conduct practice training sessions with
interviewers.
4. Examine the interviewers understanding of
the studys purposes and procedures.
5. Have interviewers complete the questionnaire
and examine the replies they secure to see
if there is any relationship.
6. Verify a sample of each interviewers
interviews.
10
Methods for Handling Office Processing Errors
Definition
Errors that arise when coding, tabulating, or
analyzing the data.
1. Use field edit to detect the most glaring
omissions and inaccuracies in the data.
2. Use a second edit in the office to decide how
data collection instruments containing
incomplete answers, obviously wrong answers, and
answers that reflect a lack of interest are
to be handled.
3. Use closed questions to simplify the coding,
but when open-ended questions need to be
used, specify the appropriate codes that will be
allowed before collecting the data.
4. When open-ended questions are being coded,
and multiple coders are being used, divide
the task by questions and not by data collection
forms.
5. Have each coder code a sample of the others
work to ensure a consistent set of coding
criteria is being employed.
6. Follow established conventions when coding
the data for computer analysis (e.g., use
numeric codes and not letters of the alphabet).
7. Prepare a codebook that lists the codes for
each variable and the categories included in
each code.
8. Use appropriate methods to analyze the data.
11
Budgeting and scheduling of field work

• Office wages and salaries
• Materials and supplies
• Telephone charges
• Travel expenses
• Field supervisors compensation
• Interviewers compensation
• Reproduction of questionnaires and other forms