|
Chapter 3 Collection of Data
Data collection includes all the processes involved in implementing
a planned design to acquire data. Common types of data collection include:
- Regulatory
data collections (e.g., the airline traffic data required by 14 CFR 241),
- Administrative
data collections (e.g., the border crossing data), and
- Surveys
(e.g., the Commodity Flow Survey).
In cases where BTS conducts or
sponsors the data collections, BTS has control over the collection
process. BTS also uses data from
external sources. In these cases, BTS
has little or no control over the data collection. External-source data vary in importance for
BTS use. Some data, such as the border
crossing data, BTS both disseminates and uses in further analyses. BTS uses other external-source data only
incidentally in analysis reports.
This chapter contains standards
for acquiring data from external sources (Section 3.1), maintaining the frame
(list of the target population) (Section 3.2), conducting data collection
operations (Section 3.3), and documenting the data collection process (Section
3.4). Except for the guideline on confidentiality
protection (Guideline 3.3.4), only Section 3.1 is required for incidentally
used external data.
3.1 Acquiring Data from
External Sources
Standard 3.1: Data that BTS
acquires from external sources must be evaluated and understood in order to
assess the quality for the intended BTS use.
Key Terms: confidentiality, external
source
Guideline 3.1.1: Obtaining External Data.
Obtain the highest quality version of the
external data that is available from the source.
- Verify that the data
set is the latest version, and that no corrected or revised data are available
for the current or previous time periods. Keep a backup copy of the data.
- Obtain the most
complete data documentation available for the corresponding time periods. Acquire any available documentation that can
be used to assess data quality.
- Evaluate data from
external sources for data quality before deciding whether the data are
appropriate for the intended BTS use.
The level of BTS’s evaluation effort should depend on the thoroughness
of the external source’s quality control and on the importance of the data for
the intended use.
Guideline 3.1.2: Confidential External Data
If the external data
are confidential or proprietary, written agreements to acquire the data must
stipulate the confidentiality requirements for protecting it.
Related Information
Bureau of Transportation Statistics. 2005. BTS Statistical Standards Manual,
Chapter 2 (Data Collection Planning and Design) and Chapter 4 (Processing of
Data). Washington,
DC.
Office of Management and Budget. 2002. Guidelines for Ensuring and Maximizing the
Quality, Objectivity, Utility, and Integrity of Information Disseminated by
Federal Agencies, Public Comments and OMB Response (Applicability of
Guidelines). Federal Register, Vol. 67,
No. 36, pp. 8453-8454. Washington,
DC.
February 22.
U.S. Department of Transportation (DOT). 2002. The Department of Transportation Information Dissemination Quality Guidelines, Appendix A, Section 1.3 (Applicability). Washington, DC. Available at http://dms.dot.gov/ombfinal092502.pdf as of January 19, 2005.
Approval Date: April 20, 2005
3.2 Frame Maintenance and
Updates
Standard 3.2: Frames (lists of potential data providers)
must be maintained, updated, evaluated, and archived to ensure that coverage is
as complete and current as possible.
Key Terms: administrative data collection, bias, coverage, frame, regulatory data collection, target population
Guideline 3.2.1: Maintaining Coverage
Frames must be
maintained and updated.
- Maintenance is the continuous revision of the frame based on
new information that becomes available during data collection. For regulatory or administrative data
collections, frame maintenance requires that changes related to reporting eligibility
are promptly reflected in the data collection system.
- Updates are systematic, comprehensive searches for frame
changes that canvass all available information. Updates can also include re-examination
of reporting categories using more recent information, such as reclassifying
airlines based on annual operating revenues.
Maintenance and updating actions include:
- Additions of new potential data providers,
- Revisions due to changes in ownership, name, or address.
- Changes in how data providers are classified (for reporting or sampling
purposes), and
- Deletions of data providers no longer in the target population.
Guideline 3.2.2: Coverage Evaluation
In addition to routine maintenance and
updates, periodically evaluate target population coverage of frames that are
used for recurring data collections.
- The
frequency of coverage evaluations depends on the relative stability of the
target population and on the frequency of data collection.
- Evaluate
coverage of administrative or regulatory data collections at least annually.
- If
the frame is properly maintained and updated, problems in coverage for
regulatory based systems can be avoided.
- Conduct
an evaluation of the potential bias if the frame’s coverage of the target
population falls below 85 percent (OMB 2005).
Guideline 3.2.3: Archiving
Frames are a critical component of data collection and
documentation. A
backup copy of the current frame must be created and archived prior to each
major frame update (or periodically, for continuously maintained frames).
- All
active and inactive data providers must be included on the archive file.
- Inactive
records may be periodically deleted from the current file, after the prior
file has been archived.
- During
a frame update, information on potential data providers should not be
deleted from the frame. Instead, a
status indicator field in the frame should designate whether the entry is
active/inactive or in-scope/out-of-scope.
- Whenever
the information contained in a frame is modified, record the effective date
of the change.
- Provide
a way of tracking changes in frame record identifiers over time.
Related Information
Bureau of Transportation Statistics. 2005. BTS Statistical Standards Manual, Section
2.2 (Target Population and Sample Design).
Washington, DC.
Federal Committee on Statistical
Methodology. 1990. Survey
Coverage, Statistical Policy Working Paper 17, Washington,
DC:
Office of Management and Budget. Available at http://www.fcsm.gov/working-papers/wp17.html as of November 5, 2004.
__________.
2001. Measuring and Reporting Sources of Error in Surveys, Chapter 5
(Coverage Error), Statistical Policy Working Paper 31, Washington
DC:
Office of Management and Budget. Available at http://www.fcsm.gov/01papers/SPWP31_final.pdf as of December 20, 2004.
Office of Management and Budget (OMB). 2005. Standards for Statistical Surveys (Proposed),
Section 2.1 (Developing Sampling Frames).
Washington, DC. July 14.
Approval Date: April 20, 2005
3.3 Data Collection
Operations
Standard 3.3: Design and administer data collection methods and instruments to balance among:
- The
maximization of data quality,
- The
control of measurement error and bias due to missing data, and
- The
minimization of respondent burden and cost.
In addition, if BTS promises
confidentiality of respondent’s data, then BTS must protect the privacy rights
of the respondents and data providers, and protect their data from unauthorized
disclosure.
Key Terms: confidentiality,
data collection, Information Collection
Request (ICR), key variable, measurement error, nonresponse bias,
response rates
Guideline 3.3.1: Quality Assurance
Develop protocols to
monitor data collection activities, with strategies to identify and correct
problems to ensure quality during data collection:
- Implement a process
control system during data collection to monitor data quality. The quality control system should be
integrated into the data collection process, and enable staff to identify
and resolve problems. The control
system should also provide data quality measurements for use as indicators
of data collection performance and data quality. Use a data
tracking process to ensure that data are not lost when transferred to
BTS.
- Use a
verification process in data entry to ensure entry errors remain below a
set limit based on data accuracy requirements. Include data verification rules in online
or other electronic data collection systems.
- Conduct
refresher training periodically for persons involved in interviewing,
observing, or providing data to maintain proper procedures and standards.
- Track on-going response rates
and item nonresponse for key variables.
Conduct an evaluation of potential item nonresponse bias if
response rates (defined in Section 4.3.1) fall below 70 percent for core
items (OMB 2005).
- Determine the core items to
obtain when a respondent is unwilling to complete the whole information
collection instrument. Target the
core items to meet the minimum standard for unit response and to analyze nonresponse
bias (Section 4.4).
Guideline 3.3.2: Encouraging Cooperation
To encourage data providers and respondents to participate, train data
collection staff on obtaining cooperation, building rapport, and converting refusals,
even for mandatory data collections. Response
rates and data quality can also be improved through means such as the use of
prenotification letters, multiple contacts, and reminder notices.
Guideline 3.3.3: Information Collection Request
Provide
respondents with an Information Collection
Request (ICR) when collecting information.
The ICR is usually placed on the information collection instrument. Follow the requirements for ICRs given in the
“Information Collection” section of the BTS Confidentiality
Procedures Manual.
Guideline 3.3.4: Protecting Confidential Data
In all phases of data collection,
confidential data must be protected from unauthorized access or release:
- Protect
identifying information of respondents as collected or on the sample frame
from unauthorized release or access.
- Ensure
that controls are in place to prevent unauthorized access to electronic
information collections (computer assisted interviewing, web based
collections, or other electronic filing methods).
- Ensure
that all data collection staff have received confidentiality training and
signed a non-disclosure form prior to collecting data.
- Use
secure means when handling and storing the data during collection to
protect against disclosure.
- Use
other means to protect confidential information as outlined in the Confidentiality Procedures Manual.
Related Information
49 U.S.C. 111, as amended by the Safe, Accountable,
Flexible, Efficient Transportation Equity Act: A Legacy for Users. P.L. 109-59.
Bureau of Transportation Statistics. 2004. Confidentiality
Procedures Manual. Washington,
DC.
__________.
2005. BTS Statistical Standards Manual, Section 2.3 (Data Collection Methods)
and Chapter 4 (Processing of Data). Washington,
DC.
Groves, R. 1989. Survey Errors and Survey Costs. New York, NY: Wiley,
Chapters 10 and 11.
Office of Management and Budget (OMB). 2005. Standards for Statistical Surveys (Proposed),
Section 2.3 (Data Collection Methodology).
Washington, DC.
July 14.
Privacy Act of 1974.
Confidential Information Protection and Statistical Efficiency Act (CIPSEA) of 2002, P.L. 107-347, Title V.
Approval Date: April 20, 2005
3.4 Documentation of Data
Collection Procedures
Standard 3.4: The data collection procedures should be
documented both for internal staff reference and for the public. Documentation should be thorough enough to
allow reproduction of the steps leading to the results.
Key Terms: external source,
frame
Guideline 3.4.1: Documentation of External Data Sources
All data that
BTS acquires from external sources must have adequate levels of documentation.
Documentation for external sources should include:
- The organization providing data,
- The exact name of the data source,
- If the data were obtained from a publication, the full
publication information and source for the data within the publication,
- If the data were acquired as a data file, how the file was
obtained, the date obtained, and the cost (if any),
- If the data were obtained from the web, the web address and the
date acquired,
- The best documentation available from the external data source
on the data collection design (including sampling, if used), the data
collection and processing procedures, any analysis or modeling performed,
and any evaluations of the data quality,
- Information on whether the external data are confidential or
proprietary, and if so, a copy of the written agreement used to obtain the
data,
- Any additional notes on the interpretation and use of the data,
- Any personal communications required to obtain the data,
information about the data source, or information about data quality, and
- Contact information for further questions.
Guideline 3.4.2: Frame Maintenance Documentation
Documentation
for maintaining and updating frames must be written and revised as
necessary. The documentation must
include:
- The
frequency of routine maintenance and major updates,
- Sources
of information used for maintenance and updates,
- Procedures
for incorporating the results of the updates on all appropriate files,
mailing lists, and other data collection control forms or listings
- Summary
of results of the frame maintenance and updates, and
- The
results of periodic coverage studies.
Guideline 3.4.3: Documentation of Data Collection Operations
The data collection operations
documentation should include:
- The
method of data collection (e.g., mail, telephone, Internet, etc.),
including methods used to track and follow up delinquent reports,
- The
data collection period, response achieved by the end of the period, and
final response achieved,
- Copies
of materials used in the data collection, including instructions given to
data providers,
- Copies
of materials used in training data collection and data provider staff,
- Schedule
of data collection operations,
- Any
response analysis or other validation surveys conducted for new data collection
efforts,
- Quantification
of response errors to the extent possible.
Related Information
Office of Management and Budget. 2002. Guidelines for Ensuring and Maximizing the
Quality, Objectivity, Utility, and Integrity of Information Disseminated by
Federal Agencies. Federal Register,
Vol. 67, No. 36, pp. 8450-8460. Washington,
DC.
February 22.
Approval Date: April 20, 2005
|
|