Showing papers in "Public Opinion Quarterly in 2009"
TL;DR: In this paper, the same questionnaires were administered simultaneously by RDD telephone interviewing, by the In- ternet with a probability sample, and by the Internet with a nonprobability sample of people who volunteered to do surveys for money.
Abstract: In a national field experiment, the same questionnaires were administered simultaneously by RDD telephone interviewing, by the In- ternet with a probability sample, and by the Internet with a nonprobability sample of people who volunteered to do surveys for money. The probabil- ity samples were more representative of the nation than the nonprobability sample in terms of demographics and electoral participation, even after weighting. The nonprobability sample was biased toward being highly engaged in and knowledgeable about the survey's topic (politics). The telephone data manifested more random measurement error, more survey satisficing, and more social desirability response bias than did the Inter- net data, and the probability Internet sample manifested more random error and satisficing than did the volunteer Internet sample. Practice at completing surveys increased reporting accuracy among the probability Internet sample, and deciding only to do surveys on topics of personal interest enhanced reporting accuracy in the nonprobability Internet sam- ple. Thus, the nonprobability Internet method yielded the most accurate self-reports from the most biased sample, while the probability Internet sample manifested the optimal combination of sample composition ac- curacy and self-report accuracy. These results suggest that Internet data collection from a probability sample yields more accurate results than do
924 citations
TL;DR: This paper investigated how expected and actual question naire length affects cooperation rates and a variety of indicators of data quality in web surveys and found that the expected length of a web-based questionnaire is negatively related to the initial willingness to participate.
Abstract: This paper investigates how expected and actual question naire length affects cooperation rates and a variety of indicators of data quality in web surveys. We hypothesized that the expected length of a web-based questionnaire is negatively related to the initial willingness to participate. Moreover, the serial position of questions was predicted to influence four indicators of data quality. We hypothesized that ques tions asked later in a web-based questionnaire will, compared to those asked earlier, be associated with (a) shorter response times, (b) higher item-nonresponse rates, (c) shorter answers to open-ended questions, and (d) less variability to items arranged in grids. To test these assump tions, we manipulated the stated length (10, 20, and 30 minutes) and the position of questions in an online questionnaire consisting of ran domly ordered blocks of thematically related questions. As expected, the longer the stated length, the fewer respondents started and completed the questionnaire. In addition, answers to questions positioned later in the questionnaire were faster, shorter, and more uniform than answers to questions positioned near the beginning.
892 citations
TL;DR: The authors examined the accuracy of survey-based self-reports of news exposure and found that these self-reported reports can differ considerably from independent assessments, leading to severe overreporting of the news exposure.
Abstract: Many studies of media effects use self-reported news expo- sure as their key independent variable without establishing its validity. Motivated by anecdotal evidence that people's reports of their own me- dia use can differ considerably from independent assessments, this study examines systematically the accuracy of survey-based self-reports of news exposure. I compare survey estimates to Nielsen estimates, which do not rely on self-reports. Results show severe overreporting of news exposure. Survey estimates of network news exposure follow trends in Nielsen ratings relatively well, but exaggerate exposure by a factor of 3 on average and as much as eightfold for some demographics. It follows that apparent media effects may arise not because of differences in ex- posure, but because of unknown differences in the accuracy of reporting exposure.
389 citations
TL;DR: This article examined the contemporary contours of policy attitudes as they relate to ideo- logical identity and considered the implications for the way scholars conceptualize, measure, and use political ideology in empirical research.
Abstract: Given the increasingly polarized nature of American poli- tics, renewed attention has been focused on the ideological nature of the mass public. Using Bayesian Item Response Theory (IRT), we examine the contemporary contours of policy attitudes as they relate to ideo- logical identity and we consider the implications for the way scholars conceptualize, measure, and use political ideology in empirical research. Although political rhetoric today is clearly organized by a single ideolog- ical dimension, we find that the belief systems of the mass public remain multidimensional, with many in the electorate holding liberal preferences on one dimension and conservative preferences on another. These cross- pressured individuals tend to self-identify as moderate (or say "Don't Know") in response to the standard liberal-conservative scale, thereby jeopardizing the validity of this commonly used measure. Our analysis further shows that failing to account for the multidimensional nature of ideological preferences can produce inaccurate predictions about the voting behavior of the American public.
291 citations
TL;DR: This paper examined attitudinal data from the 2008 MIT Energy Survey to find that substantial majorities of Americans oppose the location of coal, natural gas, and nuclear power plants in their area, although a majority supports local siting of wind facilities.
Abstract: Increased demand for U.S. electricity generation will require the construction of hundreds of new power plants in the coming decades. We examine attitudinal data from the 2008 MIT Energy Survey to mea- sure public support for and opposition to the local siting of power plants. Substantial majorities of Americans oppose the location of coal, natural gas, and nuclear power plants in their area, although a majority supports local siting of wind facilities. We find that attitudes about plant siting depend heavily on perceptions of the environmental harm and costs of specific facilities; the effects of these attributes are similar across differ- ent types of fuel sources, suggesting that there is a common underlying structure to an individual's attitude. That is, people view all power sources in the same framework and differentiate them on perceived endowments, the most important of which is environmental harm. Expansion of electric power generation over the coming decades presents a major political challenge for the United States. The U.S. Energy Information Agency projects a 40 percent rise in electricity demand over the next 30 years. To meet rising demand with the current mix of energy sources will require ap- proximately 230 new coal-fired power plants, 160 wind farms, and 10 nuclear power facilities (U.S. EIA 2007). 1 Given the inefficiency of electricity trans- mission, most of these facilities will need to be located near population centers, raising the prospects that local opposition to such facilities may significantly delay or even prevent the development of sufficient numbers of power plants to
214 citations
TL;DR: In this article, the effects of visual and verbal answer box manipulations (i.e., altering the size of the answer box and including an explanation that answers could exceed the length of the box) and the inclusion of clarifying and motivating introductions in the question stem were examined.
Abstract: Previous research has revealed techniques to improve re- sponse quality in open-ended questions in both paper and interviewer- administered survey modes. The purpose of this paper is to test the effectiveness of similar techniques in web surveys. Using data from a se- ries of three random sample web surveys of Washington State University undergraduates, we examine the effects of visual and verbal answer-box manipulations (i.e., altering the size of the answer box and including an explanation that answers could exceed the size of the box) and the inclusion of clarifying and motivating introductions in the question stem. We gauge response quality by the amount and type of information con- tained in responses as well as response time and item nonresponse. The results indicate that increasing the size of the answer box has little effect on early responders to the survey but substantially improved response quality among late responders. Including any sort of explanation or
188 citations
TL;DR: This paper found that the individuals with whom we choose to choose to discuss politics are the same people with which we discuss other important matters in our lives, and that the network data provided by survey respondents are very similar regardless of which net-work generator procedure is used.
Abstract: Social scientists use two different methods for collecting in- formation on the people with whom individuals discuss politics. Some surveys ask respondents to provide information about the people with whom they discuss "important matters," while other studies ask for in- formation specifically on the individual's political discussants. Drawing on three of the most recently collected sources of data on this subject, we compare social network data that have been collected in these two different ways. The majority of our results show that the network data provided by survey respondents are very similar regardless of which net- work generator procedure is used. These results suggest that we do not consciously select specific individuals with whom to discuss politics. Instead, the individuals with whom we choose to discuss politics are the same people with whom we discuss other important matters in our lives. This finding has significant methodological and substantive implications for studies of social influence.
172 citations
TL;DR: This paper found that Whites remain as far apart on racial policy matters in 2008 as in 1988 and younger cohorts of Whites are no more racially liberal in 2008 than they were in 1988, and the racial divide is only partially mitigated among Obama supporters.
Abstract: A number of political commentators and social scientists have speculated about the implications of the election of Barack Obama for race relations. Some of the more optimistic have suggested that the 2008 election demonstrated that Whites' racial attitudes have undergone a fundamental transformation. In this article, I seek to determine whether the putative transformation of Whites' racial attitudes has extended to levels of support for policies designed to alleviate racial inequality, the role of racial prejudice in shaping these policy preferences and whether or not prejudice influenced the presidential vote choice in 2008. Much of the analyses in this article rely upon comparisons between the 1988 election, the last time an African American candidate achieved some success in the Democratic presidential primaries, and the 2008 election utilizing survey data from the American National Election Studies (ANES). In general, I find scant evidence of a decline in the racial divide. Blacks and Whites remain as far apart on racial policy matters in 2008 as in 1988. Second, younger cohorts of Whites are no more racially liberal in 2008 than they were in 1988. Third, the racial divide is only partially mitigated among Obama supporters. Fourth, in analysis of Whites' racial policy preferences in 2008, I find that anti-Black stereotypes and indifference to Black suffering are among the strongest correlates of these opinions. Finally, I find that these same factors also contribute substantially to opposition to Obama in the 2008 election.
149 citations
TL;DR: The authors used data from the Associated Press-Yahoo! News-Stanford University survey to explore the possibility that the election outcome might have been influenced by anti-African-American racism among voters.
Abstract: The presence of an African-American candidate on the bal- lot running for President in 2008 raises the possibility that the election outcome might have been influenced by anti-African-American racism among voters. This paper uses data from the Associated Press-Yahoo! News-Stanford University survey to explore this possibility, using mea- sures of both explicit racism (symbolic racism) and implicit racism (the Affect Misattribution Procedure). The parameters of multinomial logistic regression equations were estimated to test the hypotheses that racism might have behaved differently on election day than they would have had racism been eliminated. The findings suggest that racism's impact on the election outcome could have been substantial, by causing (1) people
131 citations
TL;DR: This paper used multilevel models to evaluate a network perspective in predicting native born whites' attitudes toward immigrants and immigration policy, finding that those who are embedded in older and tighter core networks are likely to hold anti-immigrant attitudes, controlling for individual-and group-level factors.
Abstract: Recent immigration has made the United States significantly more racially and ethnically diverse. These demographic changes prompt questions regarding intergroup conflict. With data from the 2004 General Social Survey and the 2000 Census, I add to this discussion by using multilevel models to evaluate a network perspective in predicting native born whites' attitudes toward immigrants and immigration policy. The results indicate that native-born whites who are embedded in educated core networks with nonwhite alters are likely to hold pro-immigrant atti tudes, while those who are embedded in older and tighter core networks are likely to hold anti-immigrant attitudes, controlling for individual- and group-level factors. Personal contacts play an important role in shaping native-born whites' opinions of immigrants and immigration policy, re gardless of the presence of or interaction with immigrants. At the same time, core networks also condition the effects of group threat and inter group contact on immigration attitudes, suggesting that the interpersonal environment and the broader social environment interact during opinion formation.
124 citations
TL;DR: The authors analyzes the relationship between political trust, ideology, and public support for tax cutting and finds that political trust increases support for a variety of tax cut initiatives, but only among liberals.
Abstract: This article analyzes the relationship between political trust, ideology, and public support for tax cutting. Conceptualizing tax cuts as government action under conditions of risk, I argue that political trust should increase rather than decrease public support for tax cutting, partic ularly among those for whom endorsement of tax cuts entails ideological sacrifice. The results show that political trust increases support for a va riety of tax cut initiatives, but only among liberals. The results further suggest that trust is an instrumental resource that can bolster support for a conservative as well as a liberal policy agenda.
TL;DR: This paper explored the accuracy and informational content of forecasts derived from two different types of data: polls and prediction markets, and showed that debiased prediction market-based forecasts provide more accurate probabilities of victory and more information than debiasing poll-based predictions.
Abstract: Using the 2008 elections, I explore the accuracy and informational content of forecasts derived from two different types of data: polls and prediction markets. Both types of data suffer from inherent biases, and this is the first analysis to compare the accuracy of these forecasts adjusting for these biases. Moreover, the analysis expands on previous research by evaluating state-level forecasts in Presidential and Senatorial races, rather than just the national popular vote. Utilizing several different estimation strategies, I demonstrate that early in the cycle and in not-certain races debiased prediction market-based forecasts provide more accurate probabilities of victory and more information than debiased poll-based forecasts. These results are significant because accurately documenting the underlying probabilities, at any given day before the election, is critical for enabling academics to determine the impact of shocks to the campaign, for the public to invest wisely and for practitioners to spend efficiently.
TL;DR: This paper showed that the use of multiple protocols for part of the survey effort increased the response rate, changed point estimates, and achieved lower total nonresponse error, but did not show an association between survey effort and nonresponse bias.
Abstract: Nonexperimental and experimental studies have shown a lack of association between survey effort and nonresponse bias. This does not necessarily mean, however, that additional effort could not reduce nonresponse bias. Theories on nonresponse would suggest the use of different recruiting methods for additional survey effort in order to address nonresponse bias. This study looks at changes in survey estimates as a function of making additional calls under the same protocol and additional calls under a different protocol. Respondents who were interviewed as a result of more than five call attempts were not significantly different on any of the key survey variables than those interviewed with fewer than five calls. Those interviewed under a different survey protocol, however, were different on 5 of 12 measures. Additional interviews under both the same and different protocols contributed to the reduction of total nonresponse error. In sum, the use of multiple protocols for part of the survey effort increased the response rate, changed point estimates, and achieved lower total nonresponse error. Future work is needed on optimizing survey designs that implement multiple survey protocols.
TL;DR: This paper examined error in the problem detection of both cognitive interviewers evaluating their own interviews and independent judges listening to the full set of interviews and found that participants were prone to believe that the respondent had experienced a problem when the majority of judges did not believe this to be the case (false alarms).
Abstract: Cognitive interviewing is used to identify problems in questionnaires under development by asking a small number of pretest participants to verbally report their thinking while answering the draft questions. Just as responses in production interviews include measurement error, so the detection of problems in cognitive interviews can include error. In the current study, we examine error in the problem detection of both cognitive interviewers evaluating their own interviews and independent judges listening to the full set of interviews. The cognitive interviewers were instructed to probe for additional information in one of two ways: the Conditional Probe group was instructed to probe only about what respondents had explicitly reported; the Discretionary Probe group was instructed to probe whenever they felt it appropriate. Agreement about problems was surprisingly low overall, but differed by interviewing technique. The Conditional Probe interviewers uncovered fewer potential problems but with higher inter-judge reliability than did the Discretionary Probe interviewers. These differences in reliability were related to the type of probes. When interviewers in either group probed beyond the content of respondents' verbal reports, they were prone to believe that the respondent had experienced a problem when the majority of judges did not believe this to be the case (false alarms). Despite generally poor performance at the level of individual verbal reports, judges reached relatively consistent conclusions across the interviews about which questions most needed repair. Some practical measures may improve the conclusions drawn from cognitive interviews but the quality of the findings is limited by the content of the verbal reports.
TL;DR: The authors found that episodic independent voters are more likely to vote in the midterm election than in the presidential election, suggesting that some varia- tion in midterm turnout may be a function of peripheral voters becoming engaged by ballot measures.
Abstract: Research has found that states using initiatives and refer- endums have higher turnout, particularly in midterm elections. Existing research has not examined who is mobilized to vote when issues appear on statewide ballots. Building on work by Campbell (1966. "Surge and Decline: A Study of Electoral Change." In Elections and the Political Order, eds. A. Campbell, P. E. Converse, W. E. Miller, and D. E. Stokes. New York: Wiley), we test whether ballot measures engage and mobi- lize people who do not fit the profile of regular voters. Using national opinion data from the 2004 and 2006 elections, we find that indepen- dents (relative to partisans) exhibited greater awareness of and interest in ballot measures in the midterm election. This pattern is not found in the presidential election, where peripheral voters are likely to be mobilized by the stimulus of the presidential race rather than by ballot measures. Absent salient ballot measures, some episodic independent voters may not be engaged by midterm elections. This suggests that some varia- tion in midterm turnout maybe a function of peripheral voters becoming engaged by ballot measures.
TL;DR: This article used mixed Markov latent class models and data from multi-wave national panel surveys to investigate the sta- bility of individual-level party identification in three Anglo-American democracies -the United States, Britain, and Canada.
Abstract: This paper uses mixed Markov latent class models and data from multiwave national panel surveys to investigate the sta- bility of individual-level party identification in three Anglo-American democracies - the United States, Britain, and Canada. Analyses reveal that partisan attachments exhibit substantial dynamism at the latent vari- able level in the American, British, and Canadian electorates. Large-scale partisan dynamics are not a recent development; rather, they are present in all of the national panel surveys conducted since the 1950s. In all three countries, a generalized "mover-stayer" model outperforms rival models including a partisan stability model and a "black-white" nonatti- tudes model that specifies random partisan dynamics. The superiority of generalized mover-stayer models of individual-level party identification comports well with American and British studies that document nonsta- tionary, long memory in macropartisanship. The theoretical perspective provided by party identification updating models is consistent with the mix of durable and flexible partisans found in the United States and elsewhere.
TL;DR: The authors argue that party mobilization was a critical force in boosting Black voter turnout in the 2008 presidential election, using data from the 1984 and 1996 National Black Election Studies and the 2008 American National Election Study, which contains a stratified random over-sample of blacks.
Abstract: Estimates of voter turnout indicate that African Americans cast ballots at unprecedented rates in the 2008 presidential election. Given the presence of the first Black major party presidential nominee, this should be no surprise. But were heightened interest, efficacy, and a sense of racial identity due to the candidacy of Barack Obama the main factors contributing to the surge in Black voter turnout? Using data from the 1984 and 1996 National Black Election Studies and the 2008 American National Election Study, which contains a stratified random over-sample of blacks, we argue that party mobilization was a critical force in boosting Black turnout. Attitudinal factors, in contrast, appear to have been less robust in this election than one would assume.
TL;DR: The authors presented trend data regarding public attitudes regarding attitudes toward feminists, Jews, atheists, persons with strong religious beliefs, immigrants, Muslims, militarists, and AIDS sufferers since 1990.
Abstract: Recent survey data illustrate growing tolerance of various traditionally unpopular groups and their rights to teach, speak publicly, live near us as neighbors, and have their books reside in public libraries Gay and lesbian people perhaps have enjoyed the largest shift in tolerance during the past decade and a half, though other groups have come to enjoy greater acceptance as well This article presents trend data regarding public attitudes toward feminists, Jews, atheists, persons with strong religious beliefs, immigrants, Muslims, militarists, and AIDS sufferers since 1990 Groups that have recently experienced poorer acceptance include Muslims since 9/11 and immigrants more broadly over the past decade Most other groups asked about in surveys have come to enjoy more acceptance Despite the broad patterns of growing tolerance toward people who are unlike most respondents, it remains quite possible that Americans have shifted their intolerance toward other least-liked groups
TL;DR: This paper examined the county-level presidential vote in 2008 in eleven battleground states and found that those counties in which the Obama campaign had established field offices during the general election saw a disproportionate increase in the Democratic vote share.
Abstract: Imbued with unprecedented financial resources, the Obama 2008 presidential campaign established more than 700 field offices across the country, mostly in battleground states. To what extend did this form of campaigning actually affect the presidential vote? This article examines the county-level presidential vote in 2008 in eleven battleground states. The findings show that those counties in which the Obama campaign had established field offices during the general election saw a disproportionate increase in the Democratic vote share. Furthermore, this field office- induced vote increase was large enough to flip three battleground states from Republican to Democratic.
TL;DR: The authors revisited the issue of monetary incentive effects on response rates of a mail survey sent to a random sample of adults across the United States regarding preferences for fuel ethanol and found that incentives tended to bias the sample in favor of less educated respondents, and tended to attract respondents less familiar with the subject.
Abstract: This study revisits the issue of monetary incentive effects uti- lizing data from a mail survey sent to a random sample of adults across the United States regarding preferences for fuel ethanol. The results re- ported here are consistent with those found in the literature regarding the effect of incentives on response rates: they improved them, with prepaid incentives performing relatively better. We also found that state of resi- dence was significantly correlated with choosing whether to respond to a survey. Regarding the effect of incentives on sample composition, we found that incentives tended to bias the sample in favor of less educated respondents, and tended to attract respondents less familiar with the sur- vey subject. Finally, results indicate that incentives had very little effect on item nonresponse. Instead, item nonresponse was driven by education level, gender, and familiarity with the survey subject. However, combin- ing the findings on sample composition with those of item nonresponse, it appears that the use of incentives indirectly affects item nonresponse by recruiting relatively more respondents that are less educated and/or less familiar with the survey topic, who are then less likely to respond to all questions.
TL;DR: The authors analyzed data from several national surveys using various modes of interviewing to identify the optimal design of branching questions, and found that respondents who select one of the endpoints should be asked to choose among three levels of extremity, and respondents who initially select a midpoint with a precise label should not be asked whether they lean one way or the other.
Abstract: Scholars routinely employ rating scales to measure attitudes and other bipolar constructs via questionnaires, and prior research indi- cates that this is best done using sequences of branching questions in order to maximize measurement reliability and validity. To identify the optimal design of branching questions, this study analyzed data from several national surveys using various modes of interviewing. We com- pared two branching techniques and different ways of using responses to build rating scales. Three general conclusions received empirical sup- port: (1) after an initial three-option question assessing direction (e.g., like, dislike, neither), respondents who select one of the endpoints should be asked to choose among three levels of extremity, (2) respondents who initially select a midpoint with a precise label should not be asked whether they lean one way or the other, and (3) bipolar rating scales with seven points yield measurement accuracy superior to that of three-, five-, and nine-point scales.
TL;DR: The authors explored predictors of support for gender-related affirmative action in the United States and found that gender attitudes are less important in shaping such support than racial beliefs are in shaping support for race-based affirmative action.
Abstract: Drawing on data from the General Social Survey during the period from 1996 through 2006, we explore predictors of support for gender-related affirmative action in the United States. Following the literature on race-based affirmative action, we identify three main domains of predictors, each of which also resonates well with themes within scholarship on gender: interests; gender-related attitudes; and general stratification beliefs. In multivariate analyses, at least some predictors within each domain are significant. We conclude that like support for race-based affirmative action, support for gender-based affirmative action is based on a combination of interests, gender attitudes, and general stratification beliefs, but that gender attitudes are less important in shaping such support than racial beliefs are in shaping support for race-based affirmative action. The implications of these findings for the literature on race-related affirmative action as well as for the literature on gender-related attitudes are considered.
TL;DR: For example, the authors found that 87.5 percent of those who got the "loss" framing of the request completed the second interview versus 77.9 percent of the "gain" framing.
Abstract: Every survey begins with a request to the sample members to take part. How that request is framed can have a variety of consequences, including its intended (positive) effect on the cooperation rate. Survey appeals tend to emphasize the benefits of participation, but there is reason to think that emphasizing the negative consequences of nonparticipation may sometimes be a more effective method of inducing cooperation. We carried out an experiment in which respondents in a random digit dialing (RDD) sample were asked to complete a second telephone interview. For approximately half of the respondents, we emphasized the benefits of their completing the follow-up interview; for the others, we emphasized the loss involved if they chose not to complete the follow-up. Based on Kahneman and Tversky's prospect theory, we predicted that the loss framing would be more effective than the gain framing. In line with our prediction, 87.5 percent of those who got the “loss” framing of the request completed the second interview versus 77.9 percent of those who got the “gain” framing. Multivariate models of the response rate to the second interview (conditional on completion of the first) suggest that the framing effect is fairly robust across subgroups of the sample.
TL;DR: In this article, the authors explored the use of audio-CASI versus text-CASCI in a national survey of fertility-related issues in the United States (the National Survey of Family Growth).
Abstract: Audio computer-assisted self-interviewing (audio-CASI or ACASI) has been widely adopted around the world as a method for eliciting more candid responses to sensitive questions in surveys. While few studies have explored the added advantages audio may bring over text-CASI, those that have (e.g., Tourangeau and Smith, 1996, Public Opinion Quarterly 60(2):275―304) found large effects for self administration over interviewer administration, but only modest additional gains from the audio enhancement. In this paper, we explore the use of audio-CASI versus text-CASI in a national survey of fertility-related issues in the United States (the National Survey of Family Growth). In the pretest, male and female respondents were randomly assigned to audio-CASI (n = 299) or text-CASI (n = 312). We compare the distributions of substantive responses between modes and examine a variety of paradata (e.g., keystroke files, time stamps) to examine the use of the CASI instruments. The main study, which interviewed 7,643 women and 4,928 men aged 15―44 in 2002―03, used audio-CASI only, but again we have a variety of paradata and interviewer debriefing items to examine the extent to which subjects made use of the audio enhancements to CASI. Our results indicate that most respondents make limited use of the audio features of audio-CASI and accordingly the gains produced by this technology are modest relative to text-CASI.
TL;DR: Results show that coverage patterns including shared coverage, recency, and intensity of coverage all strongly relate to false negative Medicaid reporting in the CPS, which provides important guidance for questionnaire design improvements as well as model-based adjustments to the survey data.
Abstract: The use of survey data can be significantly hampered by the occurrence of measurement error, particularly that engendered by re- spondent misreporting. Previous studies have shown that public program participation tends to be substantially misreported and underestimated. Comparison with administrative records shows such a situation with the reporting of Medicaid enrollment in the U.S. Census Bureau's Current Population Survey (CPS); for 36.2 percent of enrollees, their coverage was not reported in the survey. While false positive reporting also occurs, it is, on a percentage basis, much less frequent. To understand and correct for the false negative reporting, and to develop improvements to surveys to mitigate the occurrence of misreporting, it is useful to understand the factors that relate to it. This research sought to establish the relationship between false negative Medicaid reporting and various factors associated with the coverage (such as recency and intensity), the enrollee (such as demographics and the use of services), and differences in coverage status between the respondent and the person for whom he or she was reporting. Results show that coverage patterns including shared coverage, recency, and intensity of coverage all strongly relate to false negative Medicaid
TL;DR: The authors found that self-interest had a substantial impact on support for the Vietnam War, contrary to previous results, which suggested that there was no self-interested motivation in support of the war.
Abstract: The most striking and theoretically anomalous finding of previous research on self-interest and attitudes is the absence of a self-interest motive in support for the Vietnam War. This research note reconsiders this result using a panel survey of university students collected before and after the first Vietnam draft lottery. These data are unique because they allow the unbiased estimation of the effect of self-interest on attitudes toward the war. I find that, contrary to previous results, self-interest had a substantial impact on support for the war.
Abstract: Amid growing concern about potential noncoverage bias in random digit dial survey samples that exclude cell phones, a national face-to-face exit poll provided an opportunity to reach November 2008 Election Day voters regardless of telephone status and to evaluate how the cell-only subgroup has changed since the 2004 election. The National Election Pool's survey found a sharp increase in cell-only incidence, comparable to trends for the general public in government surveys, with cell-only status approaching the norm for voters under age 30. But vot- ers age 30 and older actually abandoned landlines at a faster rate, and the difference in presidential vote preference between the cell-only and landline-accessible voters in this age group was even greater than for younger voters. This suggests that typical poststratification weighting adjustments for age may be less likely to mitigate noncoverage bias in future landline-only RDD surveys.
TL;DR: Using a chocolate as an incentive with the first mail-out was effective in generating a significantly higher initial response than the control; however, after two follow-up mail-outs, differences were not significant.
Abstract: This study examined the effectiveness of two procedures aimed at improving response rates for mail surveys: an incentive sent with either the first mail-out or first follow-up, and the inclusion of a replacement questionnaire, with or without an incentive, with either the first or second follow-up. The survey involved a sample of 1,600 New Zealand residents aged 18 years or older, randomly selected from the 2005 Electoral Roll and randomly assigned to one of four treatment groups. The response rates after two follow-ups ranged from 62.3 to 66.5 percent. Using a chocolate as an incentive with the first mail-out was effective in generating a significantly higher initial response than the control; however, after two follow-up mail-outs, differences were not significant. Sending a replacement questionnaire plus a chocolate with the first follow-up generated a significantly higher response rate than merely sending a letter, and the effect of this procedure persisted through to the end of the survey. Overall, these results provide further compelling evidence of the importance of using follow-up mail-outs for improving mail survey response rates.
TL;DR: A likely voter model for presidential elections that efficiently balances accuracy and number of questions used is built and it is argued that the proposed model compares favorably to competing models by capturing the successful elements of those models while ignoring other elements that constrain identification.
Abstract: One often-noted difficulty in pre-election polling is the iden tification of likely voters. Our objective is to build a likely voter model for presidential elections that efficiently balances accuracy and number of questions used. We employ the Iterative Expert Data Mining tech nique and data from the American National Election Studies to identify a small number of survey questions that can be used to classify likely vot ers while maintaining or surpassing the accuracy rates of other models. Specifically, we propose two survey items that together correctly classify 78 percent of respondents as voters or nonvoters over a multielection, multidecade period. We argue that our proposed model compares favor ably to competing models by capturing the successful elements of those models while ignoring other elements that constrain identification. We end by suggesting that our model offers a new approach to identifying and evaluating likely voters that may maintain or increase accuracy without also increasing cost.