|
Size: 2051
Comment: Notes, needs styling
|
Size: 2348
Comment: Remove link
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 3: | Line 3: |
| The '''National Household Education Survey''' ('''NHES''') is a national survey administered by [[UnitedStates/NationalCenterForEducationStatistics|NCES]]. | The '''National Household Education Survey''' ('''NHES''') is a national survey administered by NCES. |
| Line 25: | Line 25: |
| The first screener phase collects information about the household composition. The target population is non-institutionalized, civilian members of households in the 50 states plus [[UnitedStates/WashingtonDC|DC.]] An address list is used to select the sample of about 200,000. Black and Hispanic households are oversampled. An initial invite and several reminders are sent to all selected addresses. | The first screener phase collects information about the household composition. The target population is non-institutionalized, civilian members of households in the 50 states plus [[UnitedStates/WashingtonDC|DC.]] An [[Statistics/SurveyFrame#ABS_Frames|address list]] is used to select the sample of about 200,000. Black and Hispanic households are [[Statistics/SurveySampling#Stratification|oversampled]]. An initial invite and several reminders are sent to all selected addresses. |
| Line 27: | Line 27: |
| The second phase is a topical survey, and the eligibility criteria can change year to year. Using screener responses, households with eligible children are selected for the second phase, a topical questionnaire. If there are multiple eligible children, the sampling process selects one. Note that the survey is administered to an adult on bahlf of the selected child. | The second phase is a topical survey, and the eligibility criteria can change year to year. Using screener responses, households with eligible children are selected for the second phase, a topical questionnaire. If there are multiple eligible children, the sampling process selects one. Note that the survey is administered to an adult on behalf of the selected child. |
| Line 33: | Line 33: |
| The base weights reflect inequal sampling probabilities. | The [[Statistics/SurveyWeights|base weights]] reflect unequal stage 1 sampling probabilities. |
| Line 35: | Line 35: |
| Base weights are adjusted for screener nonresponse. | Base weights are adjusted for screener [[Statistics/NonResponseBias|non-response bias]]. |
| Line 37: | Line 37: |
| A within-household sampling adjustment accounts for inequal probability of selection. | A within-household sampling adjustment accounts for unequal probability of selection. |
| Line 39: | Line 39: |
| Topical surveys then have separate weights. Principally they reflect the respective target populations. The within-household adjusted weights are further adjusted for nonresponse and coverage bias of that particular survey's sample and instrument. | Topical surveys then have separate weights. Principally they reflect the respective target populations. The within-household adjusted weights are further adjusted for non-response and [[Statistics/SurveyInference#Sampling_Error|coverage bias]] of that particular survey's sample and instrument. |
| Line 41: | Line 41: |
| Weights are raked to benchmarks from either the ACS or CPS, depending on which has a superior sample size. | Weights are raked to benchmarks from either the [[UnitedStates/CensusBureau/AmericanCommunitySurvey|ACS]] or [[UnitedStates/CensusBureau/CurrentPopulationSurvey|CPS]], depending on which has a superior sample size. |
| Line 49: | Line 49: |
| Prior to the 2012 survey's administration, the sample was selected using RDD. | Prior to the 2012 survey's administration, the sample was selected using [[Statistics/SurveyFrame#RDD_Frames|RDD]]. |
National Household Education Survey
The National Household Education Survey (NHES) is a national survey administered by NCES.
Design
Frequency
The survey is administered every 3 or 4 years.
Sampling
The survey is administered as two phases.
The first screener phase collects information about the household composition. The target population is non-institutionalized, civilian members of households in the 50 states plus DC. An address list is used to select the sample of about 200,000. Black and Hispanic households are oversampled. An initial invite and several reminders are sent to all selected addresses.
The second phase is a topical survey, and the eligibility criteria can change year to year. Using screener responses, households with eligible children are selected for the second phase, a topical questionnaire. If there are multiple eligible children, the sampling process selects one. Note that the survey is administered to an adult on behalf of the selected child.
Weighting
The base weights reflect unequal stage 1 sampling probabilities.
Base weights are adjusted for screener non-response bias.
A within-household sampling adjustment accounts for unequal probability of selection.
Topical surveys then have separate weights. Principally they reflect the respective target populations. The within-household adjusted weights are further adjusted for non-response and coverage bias of that particular survey's sample and instrument.
Weights are raked to benchmarks from either the ACS or CPS, depending on which has a superior sample size.
History
Prior to the 2012 survey's administration, the sample was selected using RDD.
Also in 2012, the within-household sampling procedure was developed to reduce survey burden. No household can be selected for more than one survey.