Government 1010 Final: Gov 1010 Study Guide
26 Jun 2018
School
Department
Course
Professor
https://quizlet.com/243911549/gov-1010-
midterm-guide-flash-cards/
Exam starts at 1:25pm (60 minutes long)
Identification Questions
● 20%
● Four out of five terms (each 5%)
● For each term write a short paragraph defining the term and identifying its importance for
survey research
● Terms will be from the readings
Survey evaluation (case study)
● Two short surveys will be provided
● Content may include media report, press release, questionnaire, and/or methodology
description
● “What’s the sampling error”
● “Do you think there is coverage error”
● “Variables to…… (didn’t catch the end of this)”
Go through index and notes?h
Slides
● Survey Modes and Design
○ What is a survey?
■ Method of data collection
■ Designed to measure things such as attitudes or behaviors
■ Can be descriptive or analytic
● Numbers driven or description based
○ Elements of a Survey
■ Procedures that permit replication
■ Well-defined population
■ Standardized Measures
○ Traditional Approaches
■ Face to face
● Sampled at household level by interviewers, it is good at
encouraging cooperation and identifying populations, but it
generally very expensive
■ Telephone
● Randomly generated telephone sample, interviewer administered.
It is centrally supervised, which is good, but decline in response
rates obviously
■ Mail
● Household addresses, self administered by paper. Convenient for
respondents but difficult to route complex questions
■ Internet
● Email addresses, web questionnaire. Inexpensive but it is tough to
validate responses
○ With emerging technologies there are new ways to interview
■ Skype, texting, apps, smartphones etc…
○ Interviewer Involvement
■ Can be fully involved, all the way down to not involved depending on the
situation
○ Human Contact
■ Can vary from high to low
○ Channels or Communication
■ Vary from simply viewing to actually getting a response
○ Control
■ Can be respondent controlled, interviewer controlled, or computer
mediated controlled
○ Privacy as well varies from anonymous to not private and so is the level of
technology used.
○ Mixed Mode Surveys
■ People can implement different types of modes into their survey to bolster
results (in person interviewing and email surveys for examples)
○ Questions need to be clear so there is no ambiguity when answering the question
itself
○ Need to consider item nonresponse and unit nonresponse, bias of questions, the
level of education of the people we are interviewing etc...
○ Primary and Recency Effects
■ Respondents answer first thing they see (primacy) or respondents answer
last thing they heard (recency)
■ Rotate response choices to reduce this error
○ Want to reduce all biases and possible errors so be careful when designing your
survey in general
● Survey Implementation
○ Design of Surveys
■ Select mode → design instrument or design specifications → specify
procedures or develop sampling frame → recruit, hire and train staff or
select sample → collect data, process data, analyze data
○ Most of the time we focus on maximizing response rates
■ However, study suggests that non-response rate is not correlated with
nonresponse bias
■ Tailor your design approach!
○ Conditions to Optimize Response
■ Government, academic, universities
■ Intrinsic interest to respondent to help out those groups
○ Fixed Design vs Adaptive
■ Making field adjustments as we gather knowledge about our respondent
■ Now people are thinking of using these new adaptive surveys
■ Surveys mostly become part of a larger mixed-method study
● Writing a Research Proposal
○ Overview
○ Discuss your target population and why you choose them
○ Field protocols
○ Discuss the survey and the ethical considerations involved
○ Consider the human subjects in your research
○ Protect your respondents
■ Using rules of your organization make their responses anonymous etc…
● Survey Non Response
○ Between getting our sample and collecting data from respondents, the possible
error here is nonresponse error
○ Unit nonresponse → missing people
○ Item nonresponse → missing answers
○ We discuss primarily unit nonresponse as item nonresponse is natural and
generally accounted for within surveys themselves
Document Summary
Four out of five terms (each 5%) For each term write a short paragraph defining the term and identifying its importance for survey research. Content may include media report, press release, questionnaire, and/or methodology description. Do you think there is coverage error . Variables to (didn"t catch the end of this) . Designed to measure things such as attitudes or behaviors. Sampled at household level by interviewers, it is good at encouraging cooperation and identifying populations, but it generally very expensive. It is centrally supervised, which is good, but decline in response rates obviously. Convenient for respondents but difficult to route complex questions. Inexpensive but it is tough to validate responses. With emerging technologies there are new ways to interview. Can be fully involved, all the way down to not involved depending on the situation. Vary from simply viewing to actually getting a response. Can be respondent controlled, interviewer controlled, or computer mediated controlled.
