Sampling Methodology
Frame Precision and Structural Constraints
Address-Based Sampling (ABS) extracts records from the United States Postal Service Computerized Delivery Sequence (CDS) file, encompassing over ninety-seven percent of national households. While resolving the steep coverage deficiencies of telephone random digit dialing, ABS shifts operational pressure onto mail fulfillment outlays, push-to-web transitions, and geographic clustering variance.
Key Frame Dimensions
-
01
Primary Master Frame:
USPS Computerized Delivery Sequence File
-
02
Residential Reach:
~98% national household delivery points
-
03
Core Vulnerability:
Multi-unit drop points and rural PO Box filtering
Frame Architecture Review
Audit sampling stratification formulas, auxiliary census linkages, and nonresponse adjustment mechanisms prior to commissioning field mailings.
Consult Methodologists
01 Transitioning from Random Digit Dialing to Postal Matrices
Declining telephone survey cooperation over recent decades necessitated a fundamental rethinking of general population sampling foundations. Landline and mobile Random Digit Dialing (RDD), long the bedrock of empirical polling, encountered catastrophic response degradation alongside aggressive carrier call-blocking algorithms. In stark contrast, licensed postal delivery databases provide rigorous geographic geocoding down to census block groups, allowing researchers to link comprehensive spatial variables, community indicators, and demographic estimates directly to sampled addresses before any field contact takes place.
However, the move toward address listings reconfigures organizational budget distributions entirely. Telephone centers incur labor expenditures incrementally as interviewer contact attempts proceed. Address-based sampling requires heavy capital commitments up front, requiring large initial investments in postal prepayments, envelope printing, custom packaging, and pre-incentive inclusion before a single survey record is collected. When researchers miscalculate baseline delivery or response rates, these upfront cash outflows cannot be clawed back.
Detailed Probability Design Guidelines
Explore systematic sampling formulas, cluster strata designs, and variance estimation techniques.
Explore Guide
02 Mixed-Mode Sequencing and Within-Household Selection Friction
Contemporary address-based survey programs frequently implement a push-to-web protocol, issuing an initial mailing packet featuring a secure URL and unique household identifier. Nonresponding households receive a paper self-administered booklet only in subsequent reminder waves. This sequential mixed-mode framework reduces field expenditures relative to pure mailout designs, but it inherently introduces measurement discrepancies across collection modes. Younger, more educated respondents overwhelmingly complete web forms, whereas older, homeowning demographics disproportionately rely on mailed paper booklets, generating mode-by-demographic correlations that demand careful post-stratification.
Address registers offer unparalleled door-to-door delivery precision, yet physical delivery cannot govern who opens the envelope or completes the questionnaire inside the household.
— David Chen, Senior Survey Methodologist
Within-household selection introduces another point of potential bias. Where telephone interviewers can systematically run a structured Kish grid or random selection script with the person answering, postal mailings depend on self-directed instructions like the 'next birthday' rule. Compliance with these sorting rules is notoriously inconsistent, resulting in samples that regularly overrepresent the individual who handles the household mail unless post-stratification weights balance age and sex distributions.
03 Methodological Considerations and Implementation Rules
Essential practical nuances to address during initial sample allocation and frame cleaning.
Drop points represent locations where postal carriers deliver mail for multiple residential units into a single collective receptacle without secondary suite labels. In dense urban regions, these entries require explicit subsampling protocols or specialized listing algorithms to avoid excluding low-income and multi-family tenant populations.
Extensive methodological experiments establish that enclosing an unconditional token cash gift (typically two to five dollars) inside the initial envelope stimulates social reciprocity. This approach systematically lifts baseline participation by eight to fifteen percentage points, yielding a lower net cost per completed case than deferred conditional incentives.
Certain rural towns lack regular street delivery, routing all resident mail exclusively through post office boxes. Excluding PO Boxes to enforce strict physical street address criteria systematically removes rural residents from the frame, requiring the preservation of 'only-way-to-get-mail' PO Box classifications.
Topics:
Address-Based Sampling
Push-to-Web Protocol
Frame Undercoverage
Nonresponse Bias
