Random Digit Dialing¶
Build a probability-oriented telephone survey frame by generating numbers within eligible numbering blocks, screening reached numbers, and weighting the resulting sample for selection and response.
Core Idea¶
Random digit dialing is a telephone-survey sampling design that generates telephone numbers probabilistically within declared numbering strata rather than drawing only from listed subscribers. Generated numbers are dialed, screened for eligibility, and converted into respondent records under household or individual selection rules. The design’s purpose is frame coverage: unlisted but reachable numbers can have a known path into the sample. Waksberg’s classic two-stage procedure reduced the inefficiency of completely random dialing while maintaining equal unit selection probabilities under its design.
Scope of Application¶
RDD has been used in public-opinion polling, health surveillance, market research, and social surveys where telephone access offers a practical contact channel. Landline designs historically sampled households; cellular designs more often treat the reached device as linked to a person and must account for geographic mobility and safety rules. Dual-frame surveys combine landline and cellular samples while adjusting overlap among people reachable through both.
Clarity¶
Naming RDD separates frame coverage from response quality. Randomly generated numbers can reduce listed-directory bias while leaving nonresponse bias untouched. It also separates selecting a telephone number from selecting a person. The number is the contact route; the analytic unit must be specified.
This difference affects fieldwork. A sampled mobile number may reach a person while driving or outside the target geography; a landline may represent a household location more directly.
Manages Complexity¶
The numbering frame converts an incompletely listed population into a generatable contact universe. Block designs concentrate effort where working numbers are plausible, while disposition codes reduce thousands of dial attempts to analyzable outcome classes. Weights compress unequal selection, household size, dual-frame overlap, and calibration adjustments into a common estimation interface.
Abstract Reasoning¶
If every eligible number in a frame has a known nonzero generation probability, design-based inclusion reasoning becomes possible. When a household has several lines, its chance of contact increases unless multiplicity is corrected. When one line serves several eligible adults, a within-household rule is needed to avoid favoring whoever answers.
Knowledge Transfer¶
The exact method transfers across telephone surveys because number frame, generator, disposition, respondent selection, and weighting remain. Landline and mobile substrates alter role implementation but not the survey-design skeleton. The Waksberg design can transfer only where its clustering assumptions and cost structure hold.
“Generate identifiers and screen them” transfers more broadly as a sampling pattern, but random digit dialing itself remains framed by telephone numbering and contact practice. Sampling Representativeness carries the portable principle; RDD is one domain-specific realization.
Relationships to Other Abstractions¶
Current abstraction Random Digit Dialing Domain-specific
Parents (1) — more general patterns this builds on
-
Random Digit Dialing is a kind of Sampling (Representativeness) Prime
RDD strictly specializes Sampling (Representativeness) by implementing a probability-oriented contact frame.
Hierarchy paths (5) — routes to 4 parentless roots
- Random Digit Dialing → Sampling (Representativeness) → Bias
- Random Digit Dialing → Sampling (Representativeness) → Experimental Design → Comparison → Self Checking
- Random Digit Dialing → Sampling (Representativeness) → Probability → Measure → Set and Membership
- Random Digit Dialing → Sampling (Representativeness) → Probability → Measure → Aggregation → Micro Macro Linkage
- Random Digit Dialing → Sampling (Representativeness) → Experimental Design → Control Sample → Comparison → Self Checking
Neighborhood in Abstraction Space¶
Random Digit Dialing sits in a sparse region of the domain-specific corpus (89th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.
Family — Unclustered & Miscellaneous (1565 abstractions)
Nearest neighbors
- Floor Effect — 0.80
- Base Rate — 0.80
- Multilevel regression with poststratification — 0.78
- Systematic sampling — 0.78
- Transversal (Combinatorics) — 0.78
Computed from structural-signature embeddings · 2026-09-08