|
Size: 1334
Comment: Killing SurveyStatistics page
|
← Revision 10 as of 2025-11-03 01:17:04 ⇥
Size: 2718
Comment: RDD notes
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 1: | Line 1: |
| ## page was renamed from SurveyStatistics/Frames = Survey Frames = |
= Survey Frame = |
| Line 4: | Line 3: |
| In survey methodology, a '''frame''' is the set of records from which a [[SurveyStatistics/Samples|sample]] is drawn. | A '''survey frame''' is a list used for [[Statistics/SurveySampling|survey sampling]]. |
| Line 12: | Line 11: |
| == Definition == | == Description == |
| Line 18: | Line 17: |
| 3. Random-digit dial telephone frames | 3. Random-digit dial (RDD) telephone frames |
| Line 23: | Line 22: |
| ---- | The details of a frame establish the '''universe''' for [[Statistics/SurveyInference|survey inference]]. |
| Line 27: | Line 26: |
| == List Frames == | === List Frames === |
| Line 29: | Line 28: |
| A '''list frame''' is simply a list of specifically-named individuals. | A list of specifically-named individuals. |
| Line 31: | Line 30: |
| Generally, if a list frame is available, it is assumed to be the population. Frame over- and under-coverage assumptions should be examined. ---- |
List frames are commonly taken as a complete population. Assumptions about over-coverage and under-coverage should be examined. |
| Line 37: | Line 34: |
| == ABS Frames == | === ABS Frames === |
| Line 39: | Line 36: |
| '''Address-based samples''' ('''ABS'''), or address-based frames, are lists of specific addresses without the specific individuals who are believed to live there. ---- |
A list of specific addresses; the specific residents are not necessarily known, nor whether they are eligible. |
| Line 45: | Line 40: |
| == RDD Frames == | === RDD Frames === |
| Line 47: | Line 42: |
| '''Random-digit dial''' ('''RDD''') frames are randomly-selected telephone numbers. This is inherently a [[SurveyStatistics/Samples|sample]]. | A procedure for randomly generating telephone numbers. |
| Line 49: | Line 44: |
| '''Dual-frame RDD''' ('''DFRDD''') refers to drawing two random samples for landline teleophone numbers and cellphone numbers, respectively. | '''Dual-frame RDD''' ('''DFRDD''') refers to drawing two random samples for landline telephone numbers and cellphone numbers. |
| Line 51: | Line 46: |
| ---- | First, all possible '''hundred-digit blocks''' form the frame. These are (in the [[UnitedStates|U.S.]]) the full set of 8-digit prefixes for which there are 100 10-digit phone numbers. The best practice, when it is feasible, is to remove the hundred-digit blocks with 0 or 1 residential numbers (according to either published phone books or collected data). A [[Statistics/SimpleRandomSample|SRS]] with replacement is then drawn from this frame. This forms the first important probability of selection. Stratified RDD sampling relies on the segmentation of the hundred-digit block frame. Generally, this is done according to geographic bounds (state, county, or ZIP code) for which 3-digit area codes can be roughly matched. Given this information and additional statistics, e.g. which ZIP codes have higher concentrations of black residents, it can be possible to segment the frame by demographic propensities. For each selection (because a hundred-digit block can be selected multiple times), a random 10-digit phone number from the hundred-digit block is called. In some cases, 10-digit phone numbers are tried sequentially until a response is collected. This reveals one strength of the hundred-digit block--a call center can be directed to contact some number of households with a prefix, and the interviewers are free to replace the specific households as needed. From a household, there are often multiple possible respondents. A screening protocol is used to decide on the best respondent. This forms the second important probability of selection. |
| Line 55: | Line 56: |
| == Probability-based Panels == | === Probability-based Panels === |
Survey Frame
A survey frame is a list used for survey sampling.
Description
AAPOR recognizes four types of frames:
- List frames
- Address-based frames
- Random-digit dial (RDD) telephone frames
- Probability-based panel frames
This does purposefully exclude the statistics of non-probability web panels.
The details of a frame establish the universe for survey inference.
List Frames
A list of specifically-named individuals.
List frames are commonly taken as a complete population. Assumptions about over-coverage and under-coverage should be examined.
ABS Frames
A list of specific addresses; the specific residents are not necessarily known, nor whether they are eligible.
RDD Frames
A procedure for randomly generating telephone numbers.
Dual-frame RDD (DFRDD) refers to drawing two random samples for landline telephone numbers and cellphone numbers.
First, all possible hundred-digit blocks form the frame. These are (in the U.S.) the full set of 8-digit prefixes for which there are 100 10-digit phone numbers. The best practice, when it is feasible, is to remove the hundred-digit blocks with 0 or 1 residential numbers (according to either published phone books or collected data). A SRS with replacement is then drawn from this frame. This forms the first important probability of selection.
Stratified RDD sampling relies on the segmentation of the hundred-digit block frame. Generally, this is done according to geographic bounds (state, county, or ZIP code) for which 3-digit area codes can be roughly matched. Given this information and additional statistics, e.g. which ZIP codes have higher concentrations of black residents, it can be possible to segment the frame by demographic propensities.
For each selection (because a hundred-digit block can be selected multiple times), a random 10-digit phone number from the hundred-digit block is called. In some cases, 10-digit phone numbers are tried sequentially until a response is collected. This reveals one strength of the hundred-digit block--a call center can be directed to contact some number of households with a prefix, and the interviewers are free to replace the specific households as needed.
From a household, there are often multiple possible respondents. A screening protocol is used to decide on the best respondent. This forms the second important probability of selection.