Guide to entering human sciences and humanities metadata

• Individual
• Organization
• Family
• Family: Household family
• Household
• Housing Unit
• Event/Process
• Geographic Unit
• Time Unit
• Text Unit
• Group
• Object
• Other

Individual

This variable reports election returns at the constituency level.           

Population residing in metropolitan France (excluding Corsica): between 18 and 79 years old; -in main residence; -reading enough French to answer a self-administrated questionnaire.

• Longitudinal
• Longitudinal: Cohort/Event-based
• Longitudinal: Trend/Repeated cross-section
• Longitudinal: Panel
• Longitudinal: Panel: Continuous
• Longitudinal: Panel: Interval
• Time Series
• TimeSeries: Continuous
• TimeSeries: Discrete
• Cross-section
• Cross-section ad-hoc follow-up
• Other

Time Series

Panel survey

Questionnaire Administration, Survey Research Center (SRC), University of Michigan.

Describe the research project, describe the population and sample, suggest methods and language for approaching the subjects, explain questions and key terms of the survey instrument.

Monthly

• Total universe/Complete enumeration
• Probability
• Probability: Simple random
• Probability: Systematic random
• Probability: Stratified
• Probability: Stratified: Proportional
• Probability: Stratified: Disproportional
• Probability: Cluster
• Probability: Cluster: Simple random
• Probability: Cluster: Stratified random
• Probability: Multistage
• Non-probability
• Non-probability: Availability
• Non-probability: Purposive
• Non-probability: Quota
• Non-probability: Respondent-assisted
• Mixed probability and non-probability

Total universe/Complete enumeration

Samples sufficient to produce approximately 2,000 families with completed interviews were drawn in each state. Families containing one or more Medicaid or uninsured persons were oversampled.

• Actual: 385
• Formula: n0=Z2pq/e2=(1.96)2(.5)(.5)/(.05)2=385 individuals

Before adjustment, following categories appear under-estimated in the sample of survey respondents :- Single person households; - Paris basin housings; - Households with individuals over 60 years old, or without individual under 25 years old; - Under graduated persons.

• Interview
• Face-to-face interview
• Face-to-face interview: CAPI
• Face-to-face interview: PAPI
• Telephone interview
• Telephone interview: CATI
• E-mail interview
• Web-based interview
• Self-administered questionnaire
• Fixed form self-administered questionnaire
• Fixed form self-administered questionnaire: E-mail
• Fixed form self-administered questionnaire: Paper
• Fixed form self-administered questionnaire: SMS/MMS
• Fixed form self-administered questionnaire: Web-based
• Interactive self-administered questionnaire
• Interactive self-administered questionnaire: CASI
• Interactive self-administered questionnaire: CASI: VCASI
• Interactive self-administered questionnaire: CASI: ACASI
• Interactive self-administered questionnaire: CASI: TACASI
• Interactive self-administered questionnaire: CAWI
• Focus group
• Face-to-face focus group
• Telephone focus group
• Online focus group
• Self-administered writings and/or diaries
• Self-administered writings and/or diaries: E-mail
• Self-administered writings and/or diaries: Paper
• Self-administered writings and/or diaries: Web-based
• Observation
• Field observation
• Participant field observation
• Non-participant field observation
• Laboratory observation
• Participant laboratory observation
• Non-participant laboratory observation
• Computer-based observation
• Experiment
• Laboratory experiment
• Field/Intervention experiment
• Web-based experiment
• Recording
• Content coding
• Transcription
• Compilation/Synthesis
• Summary
• Aggregation
• Simulation
• Measurements and tests
• Educational measurements and tests
• Physical measurements and tests
• Psychological measurements and tests
• Other

Telephone interview

Mail questionnaires

Structured

There were 1,194 respondents who answered questions in face-to-face interviews lasting approximately 75 minutes each.

To minimize the number of unresolved cases and reduce the potential nonresponse bias, four follow-up contacts were made with agencies that had not responded by various stages of the data collection process.

Ten percent of data entry forms were reentered to check for accuracy.

The pilot sample selection procedure underestimated the impact that the small sample size and the primary unit selection process could have on the dispersion of the weights. To limit this dispersion, it was proposed to set a standardized sampling weight for the entire sample. Individual weights are adjusted for non-response using the homogeneous response group method, then by calibration on five margins of the “Enquête annuelle du recensement 2014” (sex, age, nationality, diploma, ZEAT). The weights of only the survey respondents are adjusted again to the same margins to correct for non-response. Their sum is equal to the sample size of respondents. More details about the adjustment procedure and the use of weightings in the weighting documentation.

Checks for undocumented codes were performed, and data were subsequently revised in consultation with the principal investigator.

Response rate: 84.9% of 2.770 asked to respond the survey.

To assist NES analysts, the PC SUDAAN program was used to compute sampling errors for a wide-ranging example set of proportions estimated from the 1996 NES Pre-election Survey dataset. For each estimate, sampling errors were computed for the total sample and for twenty demographic and political affiliation subclasses of the 1996 NES Pre-election Survey sample. The results of these sampling error computations were then summarized and translated into the general usage sampling error table provided in Table 11. The mean value of deft, the square root of the design effect, was found to be 1.346. The design effect was primarily due to weighting effects (Kish, 1965) and did not vary significantly by subclass size. Therefore the generalized variance table is produced by multiplying the simple random sampling standard error for each proportion and sample size by the average deft for the set of sampling error computations.

These data files were obtained from the United States House of Representatives, who received them from the Census Bureau accompanied by the following caveats: "The numbers contained herein are not official 1990 decennial Census counts. The numbers represent estimates of the population based on a statistical adjustment method applied to the official 1990 Census figures using a sample survey intended to measure overcount or undercount in the Census results. On July 15, 1991, the Secretary of Commerce decided not to adjust the official 1990 decennial Census counts (see 56 Fed. Reg. 33582, July 22, 1991). In reaching his decision, the Secretary determined that there was not sufficient evidence that the adjustment method accurately distributed the population across and within states. The numbers contained in these tapes, which had to be produced prior to the Secretary's decision, are now known to be biased. Moreover, the tapes do not satisfy standards for the publication of Federal statistics, as established in Statistical Policy Directive No. 2, 1978, Office of Federal Statistical Policy and Standards. Accordingly, the Department of Commerce deems that these numbers cannot be used for any purpose that legally requires use of data from the decennial Census and assumes no responsibility for the accuracy of the data for any purpose whatsoever. The Department will provide no assistance in interpretation or use of these numbers."

Example #1:

• Type: Description
• Subject: Data
• Text: The variables in this study are identical to earlier waves.

Example #2:

• Type: Information
• Subject: Files
• Text: Data on employment and income refer to the preceding year, although demographic data refer to the time of the survey.