Data Archive

Sampling and Weighting Methodology for the June 2024 Texas Statewide Study

For the survey, YouGov interviewed 1,302 Texas registered voters between May 31st – June 9th, 2024 who were then matched down to a sample of 1,200 to produce the final dataset. The respondents were matched to a sampling frame on gender, age, race, and education.1 The frame was constructed by using different subsets of a politically representative "modelled frame" of US adults, based upon the American Community Survey (ACS) public use microdata file, public voter file records, the 2020 Current Population Survey (CPS) Voting and Registration supplements, the 2020 National Election Pool (NEP) exit poll, and the 2020 CES surveys, including demographics and 2020 presidential vote.

The matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. The weights were then post-stratified on 2020 Presidential vote choice, and a four-way stratification of gender, age (4-categories), race (4-categories), and education (4-categories), to produce the final weight.

The margin of error for the full sample of registered voters is +/- 2.83% (3.16% for the weighted data). 

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.5 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample. 

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the February 2024 Texas Statewide Study

For the survey, YouGov interviewed 1,313 Texas registered voters between February 2nd  – 12th, 2024 who were then matched down to a sample of 1,200 to produce the final dataset. The respondents were matched to a sampling frame on gender, age, race, and education. (A match penalty was also added to exclude respondents who saw the first version of the survey that mistakenly displayed duplicate “somewhat favorable” response options in the Q17 and Q20 batteries.) The frame was constructed by using different subsets of a politically representative "modelled frame" of US adults, based upon the American Community Survey (ACS) public use microdata file, public voter file records, the 2020 Current Population Survey (CPS) Voting and Registration supplements, the 2020 National Election Pool (NEP) exit poll, and the 2020 CES surveys, including demographics and 2020 presidential vote.

The matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. The weights were then post-stratified on 2020 Presidential vote choice, and a four-way stratification of gender, age (4-categories), race (4-categories), and education (4-categories), to produce the final weight.

The margin of error for registered voters is +/- 2.83 percentage points. The margin of error of the weighted data for registered voters is +/- 3.49 percentage points. 

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.5 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample. 

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the October 2023 Texas Statewide Study

For the survey, YouGov interviewed 1,252 Texas registered voters between October 5th – 17th, 2023 who were then matched down to a sample of 1,200 to produce the final dataset. The respondents were matched to a sampling frame on gender, age, race, and education.1 The frame was constructed by using different subsets of a politically representative "modelled frame" of US adults, based upon the American Community Survey (ACS) public use microdata file, public voter file records, the 2020 Current Population Survey (CPS) Voting and Registration supplements, the 2020 National Election Pool (NEP) exit poll, and the 2020 CES surveys, including demographics and 2020 presidential vote.

The matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. The weights were then post-stratified on 2020 Presidential vote choice, and a four-way stratification of gender, age (4-categories), race (4-categories), and education (4-categories), to produce the final weight.

The margin of error for the entire sample of registered voters is +/-2.83%, and 3.3% accounting for the weighted data. 

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.5 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample. 

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the June 2023 Texas Statewide Study

For the survey, YouGov interviewed 1,547 Texas registered voters between June 2nd – 12th, 2023 who were then matched down to a sample of 1,200 to produce the final dataset. The respondents were matched to a sampling frame on gender, age, race, and education.1 The frame was constructed by using different subsets of a politically representative "modeled frame" of US adults, based upon the American Community Survey (ACS) public use microdata file, public voter file records, the 2020 Current Population Survey (CPS) Voting and Registration supplements, the 2020 National Election Pool (NEP) exit poll, and the 2020 CES surveys, including demographics and 2020 presidential vote.

The matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. The weights were then post-stratified on 2020 Presidential vote choice, and a four-way stratification of gender, age (4-categories), race (4-categories), and education (4-categories), to produce the final weight.

The margin of error for the full sample is +/- 2.83% and is +/-3.32% for the weighted data for registered voters.  

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.5 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.
Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample.

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the October 2022 Texas Statewide Study

For the survey, YouGov interviewed 1,336 Texas registered voters between February 10 – 21, 2023 who were then matched down to a sample of 1,200 to produce the final dataset. The respondents were matched on gender, age, race, and education. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2020 Current Population Survey (CPS) Voting and Registration supplements, the 2020 National Election Pool (NEP) exit poll, and the 2020 CES surveys, including demographics and 2020 presidential vote. 

The respondents were matched to a sampling frame on gender, age, race, and education. The frame was constructed by stratified sampling from the full 2020 Current Population Survey (CPS) voter registration supplement with selection within strata by weighted sampling with replacements (using the person weights on the public use file). For the main sample, the matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity, and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. These weights were then post-stratified on baseline party identification, the 2020 and 2016 presidential vote, ideology, and a full stratification of four-category age, four-category race, gender, and four-category education. The weights were trimmed at 7 and normalized to sum to the sample size.

The margin of error of for registered voters is 2.83%, and 3.44% compensating for the relative standard deviation of the weights.

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.5 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample. 

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the October 2022 Texas Statewide Study

For the survey, YouGov interviewed 1,338 Texas registered voters between October 7 – 17, 2022 who were then matched down to a sample of 1,200 to produce the final dataset. The respondents were matched on gender, age, race, and education. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2020 Current Population survey and 2014 Pew Religious Landscape Survey.

The respondents were matched to a sampling frame on gender, age, race, and education. The frame was constructed by stratified sampling from the full 2020 Current Population Survey (CPS) voter registration supplement with selection within strata by weighted sampling with replacements (using the person weights on the public use file). For the main sample, the matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity, and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. These weights were then post-stratified on baseline party identification, the 2020 and 2016 presidential vote, ideology, and a full stratification of four-category age, four-category race, gender, and four-category education. The weights were trimmed at 7 and normalized to sum to the sample size.

The margin of error for registered voters is 2.83% for registered voters and 3.6% for the weighted data. Likely voters were defined as those respondents who indicated that they have voted in every election in the past 2-3 years in response to Q2; or those respondents who rated their likelihood to vote in the November elections on a 10-point scale as a 9 or a 10 in response to Q3B, producing an overall margin of error for the likely voter sample of 3.3%.

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.5 million U.S. residents who have agreed to participate in YouGov Web surveys. At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population. Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches. In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic. At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys. After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas. These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys. Sample matching starts with an enumeration of the target population. For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey. In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study. This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample.

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample. YouGov employs the proximity matching method to find the closest matching respondent. For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the June 2022 Texas Statewide Study

For the survey, YouGov interviewed 1,359 Texas registered voters between June 16 - 24, 2022 who were then matched down to a sample of 1,200 to produce the final dataset. The respondents were matched on gender, age, race, and education. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2020 Current Population survey and 2014 Pew Religious Landscape Survey. 

The respondents were matched to a sampling frame on gender, age, race, and education. The frame was constructed by stratified sampling from the full 2020 Current Population Survey (CPS) voter registration supplement with selection within strata by weighted sampling with replacements (using the person weights on the public use file). For the main sample, the matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity, and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. These weights were then post-stratified on baseline party identification, the 2020 and 2016 presidential vote, ideology, and a full stratification of four-category age, four-category race, gender, and four-category education. The weights were trimmed at 7 and normalized to sum to the sample size.

The margin of error of the weighted data for registered voters is 2.83% for registered voters and 3.3% for the weighted data.

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.5 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample. 

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the October 2021 Texas Statewide Study

For the survey, YouGov interviewed 1344 Texas registered voters between January 27 and February 7, 2022, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, and education. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2020 Current Population survey and 2014 Pew Religious Landscape Survey.

The respondents were matched to a sampling frame on gender, age, race, and education. The frame was constructed by stratified sampling from the full 2020 Current Population Survey (CPS) voter registration supplement with selection within strata by weighted sampling with replacements (using the person weights on the public use file). For the main sample, the matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity, and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. These weights were then post-stratified on baseline party identification, the 2020 and 2016 presidential vote, ideology, and a full stratification of four-category age, four-category race, gender, and four-category education. The weights were trimmed at 7 and normalized to sum to the sample size.

The confirmed primary voters from the main sample were then combined with the oversample and split based on their vote history into 419 Democratic and 379 Republican primary voters.

The frame for the primary voter populations was constructed by stratified sampling from the full Texas voterfile subset on active voters who had voted in at least one primary election in 2012, 2014, or 2016. For each sample of Democratic and Republican primary voters, the sampled cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, and race/ethnicity. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. These weights were then raked to match the distributions of the respective primary voter frames on gender by four category age, gender by race (4 category for Democrats, 2 category for Republicans), and four category education by race (4 category for Democrats, 2 category for Republicans). The weights were trimmed at 5 and normalized to sum to the sample size.

The margin of error of the weighted data for registered voters is 2.83% for registered voters. The margins of error of the weighted data for the full sample of Democratic and Republican primary voters are 5.98% and 5.86%, respectively.

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.5 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample. 

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the August 2021 Texas Statewide Study

For the survey, YouGov interviewed 1536 Texas registered voters between August 20 and August 30, 2021, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, and education. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2018 Current Population survey and 2014 Pew Religious Landscape Survey. 

The respondents were matched to a sampling frame on gender, age, race, and education. The frame was constructed by stratified sampling from the full 2018 Current Population Survey (CPS) voter registration supplement with selection within strata by weighted sampling with replacements (using the person weights on the public use file). For the main sample, the matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity, and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. These weights were then post-stratified on baseline party identification, the 2020 and 2016 presidential vote, ideology, and a full stratification of four-category age, four-category race, gender, and four-category education. The weights were trimmed at 7 and normalized to sum to the sample size.

The margin of error of the weighted data for registered voters is 2.8% for registered voters (if adjusted for weighting, the margin of error for registered voters is 3.4%).

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.5 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample. 

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the April 2021 Texas Statewide Study

For the survey, YouGov interviewed 1392 Texas registered voters between April 16 and April 22, 2021, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, and education. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2018 Current Population survey and 2014 Pew Religious Landscape Survey. 

The respondents were matched to a sampling frame on gender, age, race, and education. The frame was constructed by stratified sampling from the full 2018 Current Population Survey (CPS) voter registration supplement with selection within strata by weighted sampling with replacements (using the person weights on the public use file). For the main sample, the matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity, and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. These weights were then post-stratified on baseline party identification, the 2020 and 2016 presidential vote, ideology, and a full stratification of four-category age, four-category race, gender, and four-category education. The weights were trimmed at 7 and normalized to sum to the sample size.

The margin of error of the weighted data for registered voters is 2.8% for registered voters (if adjusted for weighting, the margin of error for registered voters is 3.5%).

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.5 million U.S. residents who have agreed to participate in YouGov Web surveys.  At anyt given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample. 

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the February 2021 Texas Statewide Study

For the survey, YouGov interviewed 1535 Texas registered voters between February 12 and February 18, 2021, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, and education. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2018 Current Population survey and 2014 Pew Religious Landscape Survey.

The respondents were matched to a sampling frame on gender, age, race, and education. The frame was constructed by stratified sampling from the full 2018 Current Population Survey (CPS) voter registration supplement with selection within strata by weighted sampling with replacements (using the person weights on the public use file). For the main sample, the matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity, and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. These weights were then post-stratified on baseline party identification, the 2020 and 2016 presidential vote, ideology, and a full stratification of four-category age, four-category race, gender, and four-category education. The weights were trimmed at 7 and normalized to sum to the sample size.

The margin of error of the weighted data for registered voters is 2.8% for registered voters (if adjusted for weighting, the margin of error for registered voters is 3.5%).

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.5 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample.

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the June 2020 Texas Statewide Study

For the survey, YouGov interviewed 1444 Texas registered voters between June 19 and June 29, 2020, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, and education. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2018 Current Population survey and 2014 Pew Religious Landscape Survey. 

The respondents were matched to a sampling frame on gender, age, race, and education. The frame was constructed by stratified sampling from the full 2018 Current Population Survey (CPS) voter registration supplement with selection within strata by weighted sampling with replacements (using the person weights on the public use file). For the main sample, the matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity, and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. These weights were then post-stratified on baseline party identification, the 2016 presidential vote, ideology, and a full stratification of four-category age, four-category race, gender, and four-category education. The weights were trimmed at 7 and normalized to sum to the sample size.

The margin of error of the weighted data for registered voters is 3.4%. 

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample. 

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the February 2020 Texas Statewide Study

For the survey, YouGov interviewed 1443 Texas registered voters between January 31 and February 9, 2020, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, and education. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2018 Current Population survey and 2014 Pew Religious Landscape Survey. 

The respondents were matched to a sampling frame on gender, age, race, and education. The frame was constructed by stratified sampling from the full 2016 Current Population Survey (CPS) voter registration supplement with selection within strata by weighted sampling with replacements (using the person weights on the public use file). For the main sample, the matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity, and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. These weights were then post-stratified on baseline party identification, the 2016 presidential vote, ideology, and a full stratification of four-category age, four-category race, gender, and four-category education. The weights were trimmed at 7 and normalized to sum to the sample size.

The margin of error of the weighted data for registered voters is 3.6%.

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample. 

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the Aug-Sept 2019 Texas Statewide Study

Between August 29 and September 8, 2019, YouGov interviewed 768 Texas registered voters who intend to vote in the 2020 Democratic Primary, who were then matched down to a sample of 552 to produce the final dataset. The sampling frame and targets were constructed by stratified sampling from the Texas validated registered portion of the 2018 Cooperative Congressional Election Study (CCES), which had been matched to the demographic characteristics of the 2018 Current Population Survey (CPS).

The survey was sampled such that the starts were representative of all registered Texas Democrats and Independents and matched to the sampling target. The respondents were matched on gender, age, race, education, and ideology. YouGov then weighted the matched set of survey starts to a combined sample frame from the 2018 CPS and the voter validated subset of the 2018 CCES using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity, years of education, and ideology. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. Finally, dataset was subset on the qualified completes (i.e. those who intended to vote in the Democratic Primary). The weights of the qualified completes were trimmed at 7 and normalized to sum to the sample size.

Respondents were given choice of completing survey in English and Spanish. The margin of error of the weighted data for intended Democratic Primary voters is 5.1%. 

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample.

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the February 2019 Texas Statewide Study

For the survey, YouGov interviewed 1339 Texas registered voters between February 15 and February 24, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, and education. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2016 Current Population survey and 2014 Pew Religious Landscape Survey.

The respondents were matched to a sampling frame on gender, age, race, and education. The frame was constructed by stratified sampling from the full 2016 Current Population Survey (CPS) voter registration supplement with selection within strata by weighted sampling with replacements (using the person weights on the public use file). For the main sample, the matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity, and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. These weights were then post-stratified on baseline party identification, the 2016 presidential vote, and a full stratification of four-category age, four- category race, gender, and four-category education. The weights were trimmed at 7 and normalized to sum to the sample size.

The margin of error of the weighted data for registered voters is 3.2%.
 

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who have agreed to participate in YouGov Web surveys. At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population. Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches. In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic. At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys. After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas. These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys. Sample matching starts with an enumeration of the target population. For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey. In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study. This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample.

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample. YouGov employs the proximity matching method to find the closest matching respondent. For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the June 2018 Texas Statewide Study

For the survey, YouGov interviewed 1342 Texas registered voters between June 8 and June 15, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, and education. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2016 Current Population survey and 2014 Pew Religious Landscape Survey.
The respondents were matched to a sampling frame on gender, age, race, and education. The frame was constructed by stratified sampling from the full 2016 Current Population Survey (CPS) voter registration supplement with selection within strata by weighted sampling with replacements (using the person weights on the public use file). For the main sample, the matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, gender, race/ethnicity, and years of education. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. These weights were then post-stratified on baseline party identification, the 2016 presidential vote, and a full stratification of four-category age, four- category race, gender, and four-category education. The weights were trimmed at 7 and normalized to sum to the sample size.

The margin of error of the weighted data for registered voters is 3.4%.

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who have agreed to participate in YouGov Web surveys. At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population. Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches. In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic. At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys. After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

Additionally, YouGov augments their panel with difficult to recruit respondents by soliciting panelists in telephone and mail surveys. For example, in 2006 and 2010, YouGov completed telephone interviews using RDD sampling and invited respondents to join the online panel. Respondents provided a working email where they could confirm their consent and request to receive online survey invitations. YouGov also employed registration based sampling, inviting respondents to complete a pre-election survey online. At the conclusion of that survey, respondents were invited to become YouGov members and receive additional survey invitations at their email address.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas. These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys. Sample matching starts with an enumeration of the target population. For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high-quality survey, such as the American Community Survey. In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study. This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample.

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample. YouGov employs the proximity matching method to find the closest matching respondent. For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the October 2017 Texas Statewide Study

For the survey, YouGov interviewed 1311 respondents between October 5 and October 17, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, and education. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2016 Current Population survey and 2014 Pew Religious Landscape Survey.

Sampling Frame and Target Sample

YouGov constructed a national sampling frame from the 2016 Current Population Survey (CPS), including data on age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state, voter registration status, turnout, and metropolitan area. The frame was constructed by stratified sampling from the full 2016 CPS sample with selection within strata by weighted sampling with replacements (using the person weights on the public use file). Data on interest in politics, ideology, and party identification were then matched to this frame from the 2014 Pew Religious Life survey, using the following variables for the match: age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state. The target sample of 1200 Texas registered voters was selected with stratification by age, race, gender, education, and with simple random sampling within strata.

Weighting

Because matching is approximate, rather than exact, and response rates vary by group, the sample of completed interviews normally shows small amounts of imbalance that can be corrected by post-stratification weighting.
The matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, years of education, gender, race/ethnicity, ideology, interest in politics. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. The propensity scores were then post-stratified to the estimated distribution of party identifiers, and on a full stratification of four category race, four category age, gender, and four category education. Weights larger than 7 were trimmed and the final weights normalized to equal sample size. The margin of error of the weighted data is 3.59%.

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who have agreed to participate in YouGov Web surveys. At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population. Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches. In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic. At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys. After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

Additionally, YouGov augments their panel with difficult to recruit respondents by soliciting panelists in telephone and mail surveys. For example, in 2006 and 2010, YouGov completed telephone interviews using RDD sampling and invited respondents to join the online panel. Respondents provided a working email where they could confirm their consent and request to receive online survey invitations. YouGov also employed registration based sampling, inviting respondents to complete a pre-election survey online. At the conclusion of that survey, respondents were invited to become YouGov members and receive additional survey invitations at their email address.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas. These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys. Sample matching starts with an enumeration of the target population. For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high quality survey, such as the American Community Survey. In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study. This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample.

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample. YouGov employs the proximity matching method to find the closest matching respondent. For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the February 2017 Texas Statewide Study

For the survey, YouGov interviewed 1280 respondents between February 3 and February 10, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, education, party identification, ideology and political interest. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2012 Current Population survey and the 2007 Pew Religious Landscape Survey.

Sampling Frame and Target Sample

YouGov constructed a national sampling frame from the 2012 American Community Survey, including data on age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state, and metropolitan area. The frame was constructed by stratified sampling from the full 2012 ACS sample with selection within strata by weighted sampling with replacements (using the person weights on the public use file). Data on voter registration status and turnout were matched to this frame using the November 2012 Current Population Survey. Data on interest in politics and party identification were then matched to this frame from the 2007 Pew Religious Life survey, using the following variables for the match: age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state. The target sample of 1200 Texas registered voters was selected with stratification by age, race, gender, education, and with simple random sampling within strata.

Weighting

Because matching is approximate, rather than exact, and response rates vary by group, the sample of completed interviews normally shows small amounts of imbalance that can be corrected by post-stratification weighting.
The matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, years of education, gender, race/ethnicity, ideology, interest in politics. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. The propensity scores were then post-stratified to the estimated distribution of party identifiers, and on a full stratification of four category race, four category age, gender, and four category education. Weights larger than 7 were trimmed and the final weights normalized to equal sample size.

The margin of error of the weighted data is 3.73%.

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who have agreed to participate in YouGov Web surveys. At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population. Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches. In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic. At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys. After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

Additionally, YouGov augments their panel with difficult to recruit respondents by soliciting panelists in telephone and mail surveys. For example, in 2006 and 2010, YouGov completed telephone interviews using RDD sampling and invited respondents to join the online panel. Respondents provided a working email where they could confirm their consent and request to receive online survey invitations. YouGov also employed registration based sampling, inviting respondents to complete a pre-election survey online. At the conclusion of that survey, respondents were invited to become YouGov members and receive additional survey invitations at their email address.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas. These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys. Sample matching starts with an enumeration of the target population. For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high quality survey, such as the American Community Survey. In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study. This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample.

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample. YouGov employs the proximity matching method to find the closest matching respondent. For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the June 2016 Texas Statewide Study

For the survey, YouGov interviewed 1310 respondents between June 10 and June 20, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, education, party identification, ideology and political interest. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2012 Current Population survey and the 2007 Pew Religious Landscape Survey.

Sampling Frame and Target Sample

YouGov constructed a national sampling frame from the 2012 American Community Survey, including data on age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state, and metropolitan area. The frame was constructed by stratified sampling from the full 2012 ACS sample with selection within strata by weighted sampling with replacements (using the person weights on the public use file). Data on voter registration status and turnout were matched to this frame using the November 2012 Current Population Survey. Data on interest in politics and party identification were then matched to this frame from the 2007 Pew Religious Life survey, using the following variables for the match: age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state. The target sample of 1200 Texas registered voters was selected with stratification by age, race, gender, education, and with simple random sampling within strata.

Weighting

Because matching is approximate, rather than exact, and response rates vary by group, the sample of completed interviews normally shows small amounts of imbalance that can be corrected by post-stratification weighting.

The matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, years of education, gender, race/ethnicity, ideology, and interest in politics. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. Weights larger than 7 were trimmed and the final weights normalized to equal sample size.

The margin of error of the weighted data is 3.52%.

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who ha ve agreed to participate in Y ouGov W eb surveys. At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population. Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches. In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the searchtopic. At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys. After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.
Additionally, YouGov augments their panel with difficult to recruit respondents by soliciting panelists in telephone and mail surveys. For example, in 2006 and 2010, YouGov completed telephone interviews using RDD sampling and invited respondents to join the online panel. Respondents provided a working email where they could confirm their consent and request to receive online survey invitations. YouGov also employed registration based sampling, inviting respondents to complete a pre-election survey online. At the conclusion of that survey, respondents were invited to become YouGov members and receive additional survey invitations at their email address.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas. These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys. Sample matching starts with an enumeration of the target population. For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high quality survey, such as the American Community Survey. In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study. This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the targetsample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample.

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample. YouGov employs the proximity matching method to find the closest matching respondent. For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the October 2015 Texas Statewide Study

For the survey, YouGov interviewed 1484 respondents between October 30 and November 8, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, education, party identification, ideology and political interest. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2012 Current Population survey and the 2007 Pew Religious Landscape Survey. 

Sampling Frame and Target Sample

YouGov constructed a national sampling frame from the 2012 American Community Survey, including data on age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state, and metropolitan area. The frame was constructed by stratified sampling from the full 2012 ACS sample with selection within strata by weighted sampling with replacements (using the person weights on the public use file). Data on voter registration status and turnout were matched to this frame using the November 2012 Current Population Survey. Data on interest in politics and party identification were then matched to this frame from the 2007 Pew Religious Life survey, using the following variables for the match: age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state. The target sample of 1200 Texas registered voters was selected with stratification by age, race, gender, education, and with simple random sampling within strata.

Weighting

Because matching is approximate, rather than exact, and response rates vary by group, the sample of completed interviews normally shows small amounts of imbalance that can be corrected by post-stratification weighting.

The matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, years of education, gender, race/ethnicity, and interest in politics. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. Weights larger than 7 were trimmed and the final weights normalized to equal sample size.

The margin of error of the weighted data is 3.24%.  

 
Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

Additionally, YouGov augments their panel with difficult to recruit respondents by soliciting panelists in telephone and mail surveys.  For example, in 2006 and 2010, YouGov completed telephone interviews using RDD sampling and invited respondents to join the online panel.  Respondents provided a working email where they could confirm their consent and request to receive online survey invitations.  YouGov also employed registration based sampling, inviting respondents to complete a pre-election survey online.  At the conclusion of that survey, respondents were invited to become YouGov members and receive additional survey invitations at their email address.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample. 

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the February 2015 Texas Statewide Study

For the survey, YouGov interviewed 1387 respondents between February 6-16, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, education, party identification, ideology and political interest. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2012 Current Population survey and the 2007 Pew Religious Landscape Survey. 

Sampling Frame and Target Sample

YouGov constructed a national sampling frame from the 2012 American Community Survey, including data on age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state, and metropolitan area. The frame was constructed by stratified sampling from the full 2012 ACS sample with selection within strata by weighted sampling with replacements (using the person weights on the public use file). Data on voter registration status and turnout were matched to this frame using the November 2012 Current Population Survey. Data on interest in politics and party identification were then matched to this frame from the 2007 Pew Religious Life survey, using the following variables for the match: age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state. The target sample of 1200 Texas registered voters was selected with stratification by age, race, gender, education, and with simple random sampling within strata.

Weighting

Because matching is approximate, rather than exact, and response rates vary by group, the sample of completed interviews normally shows small amounts of imbalance that can be corrected by post-stratification weighting.

The matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, years of education, gender, race/ethnicity, and interest in politics. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. Weights larger than 7 were trimmed and the final weights normalized to equal sample size.

The margin of error of the weighted data is 3.36%.   

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

Additionally, YouGov augments their panel with difficult to recruit respondents by soliciting panelists in telephone and mail surveys.  For example, in 2006 and 2010, YouGov completed telephone interviews using RDD sampling and invited respondents to join the online panel.  Respondents provided a working email where they could confirm their consent and request to receive online survey invitations.  YouGov also employed registration based sampling, inviting respondents to complete a pre-election survey online.  At the conclusion of that survey, respondents were invited to become YouGov members and receive additional survey invitations at their email address.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample. 

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the May-June 2014 Texas Statewide Study

For the survey, YouGov interviewed 1350 respondents between May 30 and June 9, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, education, party identification, ideology and political interest. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2010 Current Population survey and the 2007 Pew Religious Landscape Survey. 

Sampling Frame and Target Sample

YouGov constructed a national sampling frame from the 2010 American Community Survey, including data on age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state, and metropolitan area. The frame was constructed by stratified sampling from the full 2010 ACS sample with selection within strata by weighted sampling with replacements (using the person weights on the public use file). Data on voter registration status and turnout were matched to this frame using the November 2010 Current Population Survey. Data on interest in politics and party identification were then matched to this frame from the 2007 Pew Religious Life survey, using the following variables for the match: age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state. The target sample of 1200 Texas registered voters was selected with stratification by age, race, gender, education, and with simple random sampling within strata.

Weighting

Because matching is approximate, rather than exact, and response rates vary by group, the sample of completed interviews normally shows small amounts of imbalance that can be corrected by post-stratification weighting.

The matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, years of education, gender, race/ethnicity, and interest in politics. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. Weights larger than 7 were trimmed and the final weights normalized to equal sample size.

The margin of error of the weighted data is 3.58%.  The margin of error for Republican primary voters is 5.64%.  The margin of error for Democratic primary voters is 6.04%.

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who have agreed to participate in YouGov Web surveys.  At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population.  Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches.  In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic.  At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys.  After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

Additionally, YouGov augments their panel with difficult to recruit respondents by soliciting panelists in telephone and mail surveys.  For example, in 2006 and 2010, YouGov completed telephone interviews using RDD sampling and invited respondents to join the online panel.  Respondents provided a working email where they could confirm their consent and request to receive online survey invitations.  YouGov also employed registration based sampling, inviting respondents to complete a pre-election survey online.  At the conclusion of that survey, respondents were invited to become YouGov members and receive additional survey invitations at their email address.

The YouGov panel currently has over 20,000 active panelists who are residents of Texas.  These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of “representative” samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys.  Sample matching starts with an enumeration of the target population.  For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high quality survey, such as the American Community Survey.  In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study.  This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample. 

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample.  YouGov employs the proximity matching method to find the closest matching respondent.  For each variable used for matching, we define a distance function, d(x,y), which describes how “close” the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the October 2013 Texas Statewide Study

For the survey, YouGov interviewed 1618 respondents between October 18-29, 2013, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, education, party identification, ideology and political interest. YouGov then weighted the matched set of survey respondents to known characteristics of registered voters of Texas from the 2010 Current Population survey and the 2007 Pew Religious Landscape Survey.

Sampling Frame and Target Sample

YouGov constructed a national sampling frame from the 2010 American Community Survey, including data on age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state, and metropolitan area. The frame was constructed by stratified sampling from the full 2010 ACS sample with selection within strata by weighted sampling with replacements (using the person weights on the public use file). Data on voter registration status and turnout were matched to this frame using the November 2010 Current Population Survey. Data on interest in politics and party identification were then matched to this frame from the 2007 Pew Religious Life survey, using the following variables for the match: age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state. The target sample of 1200 Texas registered voters was selected with stratification by age, race, gender, education, and with simple random sampling within strata.

Weighting

Because matching is approximate, rather than exact, and response rates vary by group, the sample of completed interviews normally shows small amounts of imbalance that can be corrected by post-stratification weighting. The matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, years of education, gender, race/ethnicity, and interest in politics. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. Weights larger than 7 were trimmed and the final weights normalized to equal sample size.

The margin of error of the weighted data is 3.1%.

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who have agreed to participate in YouGov Web surveys. At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population. Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling). The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches. In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic. At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys. After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

Additionally, YouGov augments their panel with difficult to recruit respondents by soliciting panelists in telephone and mail surveys. For example, in 2006 and 2010, YouGov completed telephone interviews using RDD sampling and invited respondents to join the online panel. Respondents provided a working email where they could confirm their consent and request to receive online survey invitations. YouGov also employed registration based sampling, inviting respondents to complete a pre-election survey online. At the conclusion of that survey, respondents were invited to become YouGov members and receive additional survey invitations at their email address.

The YouGov panel currently has nearly 20,000 active panelists who are residents of Texas. These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of "representative" samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys. Sample matching starts with an enumeration of the target population. For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high quality survey, such as the American Community Survey. In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study. This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel. The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample.

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample. YouGov employs the proximity matching method to find the closest matching respondent. For each variable used for matching, we define a distance function, d(x,y), which describes how "close" the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the February 2013 Texas Statewide Study

For the survey, YouGov interviewed 1420 respondents between February 15-25, 2013, who were then matched down to a sample of 1200 to produce the final dataset. The respondents were matched on gender, age, race, education, party identification, ideology and political interest. YouGov then weighted the matched set of survey respondents to known marginals for the registered voters of Texas from the 2008 Current Population survey and the 2007 Pew Religious Landscape Survey.

Sampling Frame and Target Sample

YouGov constructed a national sampling frame from the 2007 American Community Survey, including data on age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state, and metropolitan area. The frame was constructed by stratified sampling from the full 2007 ACS sample with selection within strata by weighted sampling with replacements (using the person weights on the public use file). Data on voter registration status and turnout were matched to this frame using the November 2008 Current Population Survey. Data on interest in politics and party identification were then matched to this frame from the 2007 Pew Religious Life survey, using the following variables for the match: age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state. The target sample of 1200 Texas registered voters was selected with stratification by age, race, gender, education, and with simple random sampling within strata.

Weighting

Because matching is approximate, rather than exact, and response rates vary by group, the sample of completed interviews normally shows small amounts of imbalance that can be corrected by post-stratification weighting.

The matched cases were weighted to the sampling frame using propensity scores. The matched cases and the frame were combined and a logistic regression was estimated for inclusion in the frame. The propensity score function included age, years of education, gender, race/ethnicity, and interest in politics. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles. Weights larger than 7 were trimmed and the final weights normalized to equal sample size.

The margin of error of the weighted data is 3.3%.

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who have agreed to participate in YouGov Web surveys. At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population. Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches. In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic. At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys. After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

Additionally, YouGov augments their panel with difficult to recruit respondents by soliciting panelists in telephone and mail surveys. For example, in 2006 and 2010, YouGov completed telephone interviews using RDD sampling and invited respondents to join the online panel. Respondents provided a working email where they could confirm their consent and request to receive online survey invitations. YouGov also employed registration based sampling, inviting respondents to complete a pre-election survey online. At the conclusion of that survey, respondents were invited to become YouGov members and receive additional survey invitations at their email address.

The YouGov panel currently has nearly 20,000 panelists who are residents of Texas. These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of "representative" samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys. Sample matching starts with an enumeration of the target population. For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high quality survey, such as the American Community Survey. In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study. This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample.

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample. YouGov employs the proximity matching method to find the closest matching respondent. For each variable used for matching, we define a distance function, d(x,y), which describes how "close" the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

For the survey May 2012 , YouGov interviewed 909 respondents between May 7-13 2012, who were then matched down to a sample of 800 to produce the final dataset. The respondents were matched on gender, age, race, education, party identification, ideology and political interest. YouGov then weighted the matched set of survey respondents to known marginals for the registered voters of Texas from the 2008 Current Population survey and the 2007 Pew Religious Landscape Survey.

Sampling Frame and Target Sample

YouGov constructed a national sampling frame from the 2007 American Community Survey, including data on age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state, and metropolitan area. The frame was constructed by stratified sampling from the full 2007 ACS sample with selection within strata by weighted sampling with replacements (using the person weights on the public use file). Data on voter registration status and turnout were matched to this frame using the November 2008 Current Population Survey. Data on interest in politics and party identification were then matched to this frame from the 2007 Pew Religious Life survey, using the following variables for the match: age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state. The target sample of 800 Texas registered voters was selected with stratification by age, race, gender, education, and with simple random sampling within strata.

Weighting Because matching is approximate, rather than exact, and response rates vary by group, the sample of completed interviews normally shows small amounts of imbalance that can be corrected by post-stratification weighting.

Raking, first proposed by Deming and Stephan (1940), adjusts an initial set of weights to match a known set of population marginals, using a method of iterative proportional fitting (see Bishop, Fienberg and Holland, 1975 for details). In this procedure, the weights are adjusted sequentially to match the marginal distribution of each weight variable. The process proceeds until all marginals are matched. It does not require any information about the joint distribution of the variables (though, if these data are available and believed to be important, they can be employed by defining a marginal distribution involving a cross-classification of two variables).

Post-stratification weights are calculated by raking the completed interviews to known marginals for Texas registered voters from the November 2008 Current Population Survey for the following variables: age, race, gender, and education.

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who have agreed to participate in YouGov Web surveys. At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population. Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches. In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic. At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys. After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

Additionally, YouGov augments their panel with difficult to recruit respondents by soliciting panelists in telephone and mail surveys. For example, in 2006 and 2010, YouGov completed telephone interviews using RDD sampling and invited respondents to join the online panel. Respondents provided a working email where they could confirm their consent and request to receive online survey invitations. YouGov also employed registration based sampling, inviting respondents to complete a pre-election survey online. At the conclusion of that survey, respondents were invited to become YouGov members and receive additional survey invitations at their email address.

The YouGov panel currently has nearly 20,000 panelists who are residents of Texas. These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of "representative" samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys. Sample matching starts with an enumeration of the target population. For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high quality survey, such as the American Community Survey. In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study. This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample.

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample. YouGov employs the proximity matching method to find the closest matching respondent. For each variable used for matching, we define a distance function, d(x,y), which describes how "close" the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the October 2011 Texas Statewide Study

For the survey, YouGovPolimetrix interviewed 889 respondents between October 19-26 2011, who were then matched down to a sample of 800 to produce the final dataset. The respondents were matched on gender, age, race, education, party identification, ideology and political interest. YouGovPolimetrix then weighted the matched set of survey respondents to known marginals for the registered voters of Texas from the 2008 Current Population survey and the 2007 Pew Religious Life Survey.

Sampling Frame and Target Sample

YouGovPolimetrix constructed a national sampling frame from the 2007 American Community Survey, including data on age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state, and metropolitan area. The frame was constructed by stratified sampling from the full 2007 ACS sample with selection within strata by weighted sampling with replacements (using the person weights on the public use file). Data on voter registration status and turnout were matched to this frame using the November 2008 Current Population Survey. Data on interest in politics and party identification were then matched to this frame from the 2007 Pew Religious Life survey, using the following variables for the match: age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state. The target sample of 800 Texas registered voters was selected with stratification by age, race, gender, education, and with simple random sampling within strata.

Weighting

Because matching is approximate, rather than exact, and response rates vary by group, the sample of completed interviews normally shows small amounts of imbalance that can be corrected by post-stratification weighting.

Raking, first proposed by Deming and Stephan (1940), adjusts an initial set of weights to match a known set of population marginals, using a method of iterative proportional fitting (see Bishop, Fienberg and Holland, 1975 for details). In this procedure, the weights are adjusted sequentially to match the marginal distribution of each weight variable. The process proceeds until all marginals are matched. It does not require any information about the joint distribution of the variables (though, if these data are available and believed to be important, they can be employed by defining a marginal distribution involving a cross-classification of two variables).

Post-stratification weights are calculated by raking the completed interviews to known marginals for Texas registered voters from the November 2008 Current Population Survey for the following variables: age, race, gender, and education.

Survey Panel Data

The YouGov panel, a proprietary opt-in survey panel, is comprised of 1.2 million U.S. residents who have agreed to participate in YouGov Web surveys. At any given time, YouGov maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population. Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the YouGov Panel is Web advertising campaigns that appear based on keyword searches. In practice, a search in Google may prompt an active YouGov advertisement soliciting opinion on the search topic. At the conclusion of the short survey respondents are invited to join the YouGov panel in order to receive and participate in additional surveys. After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

Additionally, YouGov augments their panel with difficult to recruit respondents by soliciting panelists in telephone and mail surveys. For example, in 2006 and 2010, YouGov completed telephone interviews using RDD sampling and invited respondents to join the online panel. Respondents provided a working email where they could confirm their consent and request to receive online survey invitations. YouGov also employed registration based sampling, inviting respondents to complete a pre-election survey online. At the conclusion of that survey, respondents were invited to become YouGov members and receive additional survey invitations at their email address.

The YouGov panel currently has nearly 20,000 panelists who are residents of Texas. These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of "representative" samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys. Sample matching starts with an enumeration of the target population. For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high quality survey, such as the American Community Survey. In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study. This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process. First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn.

Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample.

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample. YouGov employs the proximity matching method to find the closest matching respondent. For each variable used for matching, we define a distance function, d(x,y), which describes how "close" the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling and Weighting Methodology for the February 11 Texas Statewide Study

For the February 2011 survey, YouGovPolimetrix interviewed 963 respondents between February 9-18 2011, who were then matched down to a sample of 800 to produce the final dataset. The respondents were matched on gender, age, race, education, party identification, ideology and political interest. YouGovPolimetrix then weighted the matched set of survey respondents to known marginals for the registered voters of Texas from the 2008 Current Population survey and the 2007 Pew Religious Life Survey.

Sampling Frame and Target Sample

YouGovPolimetrix constructed a national sampling frame from the 2007 American Community Survey, including data on age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state, and metropolitan area. The frame was constructed by stratified sampling from the full 2007 ACS sample with selection within strata by weighted sampling with replacements (using the person weights on the public use file). Data on voter registration status and turnout were matched to this frame using the November 2008 Current Population Survey. Data on interest in politics and party identification were then matched to this frame from the 2007 Pew Religious Life survey, using the following variables for the match: age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state. The target sample of 800 Texas registered voters was selected with stratification by age, race, gender, education, and with simple random sampling within strata.

Weighting

Because matching is approximate, rather than exact, and response rates vary by group, the sample of completed interviews normally shows small amounts of imbalance that can be corrected by post-stratification weighting. Raking, first proposed by Deming and Stephan (1940), adjusts an initial set of weights to match a known set of population marginals, using a method of iterative proportional fitting (see Bishop, Fienberg and Holland, 1975 for details). In this procedure, the weights are adjusted sequentially to match the marginal distribution of each weight variable. The process proceeds until all marginals are matched. It does not require any information about the joint distribution of the variables (though, if these data are available and believed to be important, they can be employed by defining a marginal distribution involving a cross-classification of two variables). Post-stratification weights are calculated by raking the completed interviews to known marginals for Texas registered voters from the November 2008 Current Population Survey for the following variables: age, race, gender, and education.

For the University of Texas / Texas Tribune survey, YouGovPolimetrix interviewed 906 respondents between September 3 and 8, 2010, who were then matched down to a sample of 800 to produce the final dataset. The respondents were matched on gender, age, race, education, party identification, ideology and political interest. YouGovPolimetrix then weighted the matched set of survey respondents to known marginals for the registered voters of Texas from the 2008 Current Population survey and the 2007 Pew Religious Life Survey.

Sampling Frame and Target Sample

YouGovPolimetrix constructed a national sampling frame from the 2007 American Community Survey, including data on age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state, and metropolitan area. The frame was constructed by stratified sampling from the full 2007 ACS sample with selection within strata by weighted sampling with replacements (using the person weights on the public use file). Data on voter registration status and turnout were matched to this frame using the November 2008 Current Population Survey. Data on interest in politics and party identification were then matched to this frame from the 2007 Pew Religious Life survey, using the following variables for the match: age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state. The target sample of 800 Texas registered voters was selected with stratification by age, race, gender, education, and with simple random sampling within strata.

Weighting

Because matching is approximate, rather than exact, and response rates vary by group, the sample of completed interviews normally shows small amounts of imbalance that can be corrected by post-stratification weighting. Raking, first proposed by Deming and Stephan (1940), adjusts an initial set of weights to match a known set of population marginals, using a method of iterative proportional fitting (see Bishop, Fienberg and Holland, 1975 for details). In this procedure, the weights are adjusted sequentially to match the marginal distribution of each weight variable. The process proceeds until all marginals are matched. It does not require any information about the joint distribution of the variables (though, if these data are available and believed to be important, they can be employed by defining a marginal distribution involving a cross-classification of two variables). Post-stratification weights are calculated by raking the completed interviews to known marginals for the general population of Texas from the November 2008 Current Population Survey for the following variables: age, race, gender, and education.

Survey Panel Data

The PollingPoint panel, a proprietary opt-in survey panel, is comprised of 1.6 million U.S. residents who have agreed to participate in YouGov Polimetrix's Web surveys. At any given time, YouGov Polimetrix maintains a minimum of five recruitment campaigns based on salient current events.

Panel members are recruited by a number of methods and on a variety of topics to help ensure diversity in the panel population. Recruiting methods include Web advertising campaigns (public surveys), permission-based email campaigns, partner sponsored solicitations, telephone-to-Web recruitment (RDD based sampling), and mail-to-Web recruitment (Voter Registration Based Sampling).

The primary method of recruitment for the PollingPoint Panel is Web advertising campaigns that appear based on keyword searches. In practice, a search in Google may prompt an active PollingPoint advertisement soliciting opinion on the search topic. At the conclusion of the short survey respondents are invited to join the PollingPoint panel in order to receive and participate in additional surveys. After a double opt-in procedure, where respondents must confirm their consent by responding to an email, the database checks to ensure the newly recruited panelist is in fact new and that the address information provided is valid.

Additionally, YouGov Polimetrix augments their panel with difficult to recruit respondents by soliciting panelists in telephone and mail surveys. For example, in the fall and winter of 2006, YouGov Polimetrix completed telephone interviews using RDD sampling and invited respondents to join the online panel. Respondents provided a working email where they could confirm their consent and request to receive online survey invitations. YouGov Polimetrix also employed registration based sampling, inviting respondents to complete a pre-election survey online. At the conclusion of that survey, respondents were invited to become PollingPoint members and receive additional survey invitations at their email address.

The PollingPoint panel currently has over 55,000 active panelists who are registered voters in Texas. These panelists cover a wide range of demographic characteristics.

Sampling and Sample Matching

Sample matching is a methodology for selection of "representative" samples from non-randomly selected pools of respondents. It is ideally suited for Web access panels, but could also be used for other types of surveys, such as phone surveys. Sample matching starts with an enumeration of the target population. For general population studies, the target population is all adults, and can be enumerated through the use of the decennial Census or a high quality survey, such as the American Community Survey. In other contexts, this is known as the sampling frame, though, unlike conventional sampling, the sample is not drawn from the frame. Traditional sampling, then, selects individuals from the sampling frame at random for participation in the study. This may not be feasible or economical as the contact information, especially email addresses, is not available for all individuals in the frame and refusals to participate increase the costs of sampling in this way.

Sample selection using the matching methodology is a two-stage process.

First, a random sample is drawn from the target population. We call this sample the target sample. Details on how the target sample is drawn are provided below, but the essential idea is that this sample is a true probability sample and thus representative of the frame from which it was drawn. Second, for each member of the target sample, we select one or more matching members from our pool of opt-in respondents. This is called the matched sample. Matching is accomplished using a large set of variables that are available in consumer and voter databases for both the target population and the opt-in panel.

The purpose of matching is to find an available respondent who is as similar as possible to the selected member of the target sample. The result is a sample of respondents who have the same measured characteristics as the target sample. Under certain conditions, described below, the matched sample will have similar properties to a true random sample. That is, the matched sample mimics the characteristics of the target sample. It is, as far as we can tell, "representative" of the target population (because it is similar to the target sample).

When choosing the matched sample, it is necessary to find the closest matching respondent in the panel of opt-ins to each member of the target sample. Polimetrix employs the proximity matching method to find the closest matching respondent. For each variable used for matching, we define a distance function, d(x,y), which describes how "close" the values x and y are on a particular attribute. The overall distance between a member of the target sample and a member of the panel is a weighted sum of the individual distance functions on each attribute. The weights can be adjusted for each study based upon which variables are thought to be important for that study, though, for the most part, we have not found the matching procedure to be sensitive to small adjustments of the weights. A large weight, on the other hand, forces the algorithm toward an exact match on that dimension.

Sampling Frame and Target Sample

YouGov/Polimetrix constructed a national sampling frame from the 2007 American Community Survey, including data on age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state, and metropolitan area. The frame was constructed by stratified sampling from the full 2007 ACS sample with selection within strata by weighted sampling with replacements (using the person weights on the public use file). Data on voter registration status and turnout were matched to this frame using the November 2006 Current Population Survey. Data on interest in politics and party identification were then matched to this frame from the 2007 Pew Religious Life survey, using the following variables for the match: age, race, gender, education, marital status, number of children under 18, family income, employment status, citizenship, state. The target sample of 800 Texas registered voters was selected with stratification by age, race, gender, education, and with simple random sampling within strata.

Weighting

Because matching is approximate, rather than exact, and response rates vary by group, the sample of completed interviews normally shows small amounts of imbalance that can be corrected by post-stratification weighting.

Raking, first proposed by Deming and Stephan (1940), adjusts an initial set of weights to match a known set of population marginals, using a method of iterative proportional fitting (see Bishop, Fienberg and Holland, 1975 for details). In this procedure, the weights are adjusted sequentially to match the marginal distribution of each weight variable. The process proceeds until all marginals are matched. It does not require any information about the joint distribution of the variables (though, if these data are available and believed to be important, they can be employed by defining a marginal distribution involving a cross-classification of two variables).

You Gov Politmetrix calculated post-stratification weights by raking the completed interviews to known marginals for the general population of Texas from the November 2006 Current Population Survey and Pew Religious Life survey for the following variables: age, race, gender, education, and ideology.

The June 2009 UT-Austin Texas Politics Poll was designed by researchers in the UT-Austin Department of Government and conducted by YouGov/Polimetrix, a firm with demonstrated success in internet polling. YouGov/Polimetrix accomplishes internet polling through a unique sampling procedure known as "matched random sampling." The firm begins with two lists: (1) a list of all adult "consumers" in Texas (covering approximately 95 percent of the adult population), and (2) a list of people who have agreed to take YouGov/Polimetrix's surveys. For each list, Polimetrix has an extensive set of demographics.

The sampling procedure then progresses in two stages. First, a random sample of consumers is drawn. For each person drawn from this sample a list of key demographics is recorded. In essence, each individual drawn is represented as a cluster of demographic characteristics, including age, income, education, race, gender, longitude and latitude, etc. Second, YouGov/Polimetrix uses a matching algorithm to find the PollingPoint panelist who is the closest match to the person drawn off the consumer file. In this way an entire "matched" random sample is constructed for all people in the "drawn" sample.

The June 2009 poll consists of 924 adult Texans, and has a margin of error of +/- 3.22 percentage points at the 95% confidence level. The poll includes interviews with 791 registered voters, with an attendant margin of error of +/- 3.66 percentage points. Response rates are almost 100% given the matching methodology. The YouGov/Polimetrix pool includes people who are much less likely to have access to the Internet or a personal computer. YouGov/Polimetrix has been especially assiduous at enlisting ethnic and racial minorities, as well as people who are less affluent, as part of their attempt to ensure the representativeness of their samples. Surveys were completed between June 11 and June 22, 2009.

Polimetrix interviewed 924 respondents who were then matched down to a sample of 800 to produce the final data set. The respondents were matched on gender, age, race, education, party identification and political interest. YouGov/Polimetrix then weighted the matched set of survey respondents to known marginals for the general population of Texas from the 2006 American Community Survey. 

The October 2008 UT-Austin Texas Politics Poll was designed by researchers at UT-Austin and conducted by YouGov/Polimetrix, a firm with demonstrated success in internet polling. YouGov/Polimetrix accomplishes internet polling through a unique sampling procedure known as "matched random sampling." The firm begins with two lists: (1) a list of all adult "consumers" in Texas (covering approximately 95 percent of the adult population), and (2) a list of people who have agreed to take YouGov/Polimetrix's surveys. For each list, Polimetrix has an extensive set of demographics.

The sampling procedure then progresses in two stages. First, a random sample of consumers is drawn. For each person drawn from this sample a list of key demographics is recorded. In essence, each individual drawn is represented as a cluster of demographic characteristics, including age, income, education, race, gender, longitude and latitude, etc. Second, YouGov/Polimetrix uses a matching algorithm to find the PollingPoint panelist who is the closest match to the person drawn off the consumer file. In this way an entire "matched" random sample is constructed for all people in the "drawn" sample.

The October 2008 poll consists of 613 adult Texans, and has a margin of error of +/- 3.98 percentage points at the 95% confidence level. The poll includes interviews with 550 registered voters, with an attendant margin of error of +/- 4.20 percentage points. Response rates are almost 100% given the matching methodology. The YouGov/Polimetrix pool includes people who are much less likely to have access to the Internet or a personal computer. YouGov/Polimetrix has been especially assiduous at enlisting ethnic and racial minorities, as well as people who are less affluent, as part of their attempt to ensure the representativeness of their samples.The poll was administered by YouGov/Polimetrix. Polimetrix interviewed 899 respondents who were then matched down to a sample of 613 to produce the final data set. The respondents were matched on (among other items) gender, age, race, education, party identification and political interest. YouGov/Polimetrix then weighted the matched set of survey respondents to known marginals for the general population of Texas from the 2006 American Community Survey.