Summary Designing Surveys: A Guide to Decisions and Procedures

ISBN-10 1412997348 ISBN-13 9781412997348
314 Flashcards & Notes
1 Students
  • This summary

  • +380.000 other summaries

  • A unique study tool

  • A rehearsal system for this summary

  • Studycoaching with videos

Remember faster, study better. Scientifically proven.

This is the summary of the book "Designing Surveys: A Guide to Decisions and Procedures". The author(s) of the book is/are . The ISBN of the book is 9781412997348 or 1412997348. This summary is written by students who study efficient with the Study Tool of Study Smart With Chris.

Summary - Designing Surveys: A Guide to Decisions and Procedures

  • 1 Survey Practice

  • What is a survey?
    Surveys collect information by interviewing a sample of respondents from a well-defined population.
  • What is the identifying characteristic of a survey interview?
    De use of a fixed questionnaire with prespecified questions.
  • What is a strength of observation versus surveys? What are two weaknesses of observation versus surveys?
    Strength: not subject to reporting bias. 
    Weakness: can't measure mental states, not efficient for measuring infrequent behaviors. 
  • What is a strength of depth interviews versus surveys? What are two weaknesses of depth interviews versus surveys?
    Strength: can probe freely and go into depth. 
    Weakness: expensive, poor population coverage.
  • What are two strengths of focus groups versus surveys? What are two weaknesses of focus groups versus surveys?
    Strength: can probe freely and go into depth, can see social dynamics.
    Weakness: expensive, poor population coverage.
  • What is a strength of panels versus surveys? What are two weaknesses of panels versus surveys?
    Strength: shows changes over time.
    Weakness: expensive, a limited number of people will participate.
  • What is a strength of experiments versus surveys? What is a weakness of experiments versus surveys?
    Strength: strong test of causation.
    Weakness: difficult to do outside lab. 
  • Survey research is often a combination of disciplines, true or false?
    True: surveys are usually a team effort of many people with diverse skills.
  • What influences survey practice over time? 
    Cultural norms. Social norms. Societal changes (can occur at different rates in different parts of the population). Technical development. 
  • What are two ethical concepts central to the treatment of respondents?
    1. Informed consent.
    2. Confidentiality. 
  • What is the main concern in designing and conducting a survey?
    To achieve the research of other date collection objectives within available resources. 
  • 2 Survey Error

  • To what refers the term error?
    Deviations of obtained survey results from those that are true reflections of the population.
  • What are three sources of error?
    1. Sampling error. 
    2. Sample bias. 
    3. Nonsampling error. 
  • What is sampling error?
    Sampling error refers to the fact that samples don't always reflect a population's true characteristics. Answers will be affected by chance variantion in the composition of the sample. 
  • By what is the level of sampling error controlled?
    By the sample size: as samples get larger, the distribution of possible sample outcomes gets tighter around the true population figure. Larger samples have less chance of producing results that are uncharacteristic of the population as a whole. 
  • What is sample bias?
    Sample bias refers to the possibility that members of a sample differ from the larger population in some systematic fashion. 
  • What are the three general ways in which sample bias can arise?
    1. Coverage bias: will occur if some segment of the population is improperly excluded from consideration in the sample or is not available trough the method employed in research. 
    2. Selection bias: will occur if some population groups are given disproportionately high or low chances of selection. 
    3. Nonresponse bias: will occur if failure to respond is disproportionate across groups. 
  • By what is the level of sample bias controlled?
    Not by sample size. Rather by defining the population of interest prior to drawing the sample; attempting to maximize population coverage, selecting a sample that fairly represents the entire population, and obtaining data from as much of the selected sample as possible. 
  • What is nonsampling error?
    Nonsampling error consists of all error sources unrelated to the sampling of respondents. 
  • What are three sources of nonsampling error?
    1. Interviewer error: related to the administration of the survey.
    2. Response error: related to the accuracy of response as given. 
    3. Coding error: related tot the accuracy of response as recorded. 
    All of these may result in either random or systematic errors in the data.
  • What is are three examples of interviewer error? How can these be controlled?
    1. Cheating; the interviewer fails to administer the questionnaire, or portions of the questionnaire, and simply fabricates the data. Controlled trough some form of validation. 
    2. Question administration error: the interviewer does something other than read the question as intended. 
    3. Probing error. Examples 2 and 3 are controlled through interviewer training. Also, data collection may be spread across multiple interviewers. 
  • For what three reasons may response error occur? How are these controlled?
    1. Comprehension error: if the respondent does not understand a question correctly or if different respondents understand it in different ways. Controlled through careful question design as well as presenting the questionnaire to ensure that respondents understand the questions as intended.
    2. Knowledge error: if respondents do not know the answer to the question, or cannot recall it accurately, but still provide an answer to the best of their ability. Controlled by getting the right respondent, screening for knowledge if appropriate, writing questions that are realistic about what people might know, and matching the time frame of the question to a plausible recall period.
    3. Reporting error: if respondents hesitate to provide accurate answers. Controlled in a variety of ways, including professionalism in the survey to convince respondents that you care whether they answer but not what they answer, or using self-report methods so the respondent does not have to present himself to an interviewer. 
  • How can nonsampling error be controlled overall?
    Open-ended questions may be assigned to at least two coders to identify and resolve inconsistencies. Questionnaires are edited to verify that the answers appear to be correctly recorded an to check for interitem consistency where applicable. Respondents may be recontacted to resolve apparent errors or inconsistencies. Doing a good job of training and supervising interviewers, using good questionnaire design, exercising good control over coding and data entry. 
  • What is variance?
    In general, variance refers to random differences between repeated samples and/or measures, quantified as the average squared difference between the statistic being estimated in any given trial and the mean of that statistic across all trials. 
  • When does bias occur?
    When the estimate in any given trial is more likely to be in one direction than the other. 
  • What is the mean square error (MSE)?
    The two types of error taken together: variance and bias. Calculated as variance plus bias squared (the average squared error due to random variation plus the average squared error due to systematic error). 
  • What is total survey error?
    When MSE accounts for variance and bias accross al the survey stages.
Read the full summary
This summary. +380.000 other summaries. A unique study tool. A rehearsal system for this summary. Studycoaching with videos.

Latest added flashcards

What is imputation?
The substitution of constructed values for items that are not answered or whoe answers are inconsistent with other responses in the same interview. 
What are the three principal forms of postsurvey statistical adjustments?
1. Weighting to adjust for unequal selection probabilities in the sample design.
2. Weighting to adjust for differential unit nonresponse accros population subgroups.
3. Imputation for item nonresponse. 
What are two general methods for assessing exporue to nonresponse bias?
1. Compare the survey data with external data.
2. Examine internal variation in the data collection.
How can duplication and clustering occur in counting frames?
Duplication and clustering usually result from a mismatch between the counting units and population units. 
How can ineligibility occur in counting frames?
Ineligibility results from some of the counted elements not meeting population criteria.
How can omisson occur in counting frames?
Omission results from underestimating the population size (or the population subgroups). 
How is sampling without lists done?
1. Estimate the size of the population.
2. Select a sample of numbers between 1 and N, where N is the population size.
3. Count the population and gather data from the appropriately numbered members. 
What are three basic ways to cope with clustering?
1. Gather data from all population elements in the selected clusters. This method provides a fair chance of selection for every member of the population. Unfortunately, it also produces a sample that contains related cluster members. Because of this problem, taking entire clusters is a good idea only when clusters are relatively small and relatively few in number.
2. Sample population members within clusters at some fixed rate. 
3. Compensate for clustering by randomly selecting one population member from each cluster and weighting that observation by the size of the cluster. 
What is the problem with clustering?
Clusterd members are underrepresented in the sample. 
What are three ways to correct for duplicate listings?
1. To cross-check the list, identify duplicated, and remove them. This will be done if the list is computerized. Computerized cross-checking usually doesn't remove all of the duplications in a list, but it should reduce them to a level where they cause negligible sample bias. 
2. To draw the sample and check only the selected elements to determine how many times they are duplicated in the total population list. Sample members that are listed k times would be retained at the rate of 1/k. 
3. To ask selected population members how many times they appear in the lit. Observations are weighted by the inverse of their number of times in the list (1/k).