Missing Nonvoters and Misweighted Samples

Mellon, Jonathan; Prosser, Christopher

doi:10.1093/poq/nfx015

Cited by 42 publications

(11 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…However the question remains whether, when controlling for demographic variables, there remain unobserved, nonignorable, differences between social media users and nonusers. If there are non-demographic differences in the data, adjusting it with weights to appear demographically representative could lead to large errors (Mellon and Prosser, 2017) and would require more sophisticated adjustment methods (e.g. Wang et al, 2014).…”

mentioning

confidence: 99%

Twitter and Facebook are not representative of the general population: Political attitudes and demographics of British social media users

Mellon¹,

Prosser

2017

Research & Politics

Self Cite

313

187

View full text Add to dashboard Cite

A growing social science literature has used Twitter and Facebook to study political and social phenomena including for election forecasting and tracking political conversations. This research note uses a nationally representative probability sample of the British population to examine how Twitter and Facebook users differ from the general population in terms of demographics, political attitudes and political behaviour. We find that Twitter and Facebook users differ substantially from the general population on many politically relevant dimensions including vote choice, turnout, age, gender, and education. On average social media users are younger and better educated than non-users, and they are more liberal and pay more attention to politics. Despite paying more attention to politics, social media users are less likely to vote than non-users, but they are more likely to support the left leaning Labour Party when they do vote. However, we show that these apparent differences mostly arise due to the demographic composition of social media users. After controlling for age, gender, and education, no statistically significant differences arise between social media users and non-users on political attention, values or political behaviour.

show abstract

mentioning

confidence: 99%

Twitter and Facebook are not representative of the general population: Political attitudes and demographics of British social media users

Mellon¹,

Prosser

2017

Research & Politics

Self Cite

313

187

View full text Add to dashboard Cite

show abstract

“…The idea that mass polarization might be partially an artifact of differential nonresponse seems plausible at face value. It is well established that surveys over-represent politically interested respondents (Groves, Presser and Dipko, 2004;Keeter et al, 2006;Tourangeau, Groves and Redline, 2010;Mellon and Prosser, 2017), and other research has shown that politically interested respondents are more polarized (Abramowitz and Saunders, 2008). If lower response rates have increased nonresponse bias, i it seems plausible that lower response rates might result in more polarized survey responses.…”

Section: Nonresponse Bias and Polarizationmentioning

confidence: 99%

Correlation with Time Explains the Relationship between Survey Nonresponse and Mass Polarization

Mellon

Prosser

2021

The Journal of Politics

Self Cite

View full text Add to dashboard Cite

While elite polarization in US politics has been well established, there has been substantial debate about whether US mass opinion has polarized in recent decades. Cavari and Freedman (2018) suggest the apparent polarization of the US public may in part be an artifact of declining survey response rates. Cavari and Freedman argue that declining response rates may have led to only the most politicized-and therefore polarized-respondents being willing to give their opinions. We argue that, while the mechanism posited is plausible, other theoretical models of nonresponse predict no change in nonresponse bias or even declining nonresponse bias, depending on the cause of increased nonresponse. We show that the results of Cavari and Freedman's analysis are spuriously driven by each variable's correlation with time. After accounting for the time correlation there is no evidence that nonresponse bias has inflated our estimates of polarization.

show abstract

“…Las personas interesadas en la política son más proclives que las no interesadas a aceptar el requerimiento de contestar sobre ese tópico (Groves et al, 2000(Groves et al, , 2004. Una muestra compuesta en forma desproporcionada por individuos políticamente implicados puede producir grandes errores en la estimación de los porcentajes de voto (Mellon y Prosser, 2017). A raíz de su proceso de autoselección de los encuestados, los paneles online no probabilísticos están aún más expuestos a sufrir esta distorsión.…”

Section: La Transformación De Los Estudios Por Encuestaunclassified

La Falla de las Encuestas en las Elecciones Argentinas de 2019. Un Análisis en Perspectiva Comparada Internacional (The Failure of the Polls in the 2019 Argentine Elections. An Analysis in International Comparative Perspective)

Jorge¹,

Miró²

2021

Preprint

View full text Add to dashboard Cite

(English): The performance of polls in the 2019 Argentine elections reached its lowest point since democratic restoration in 1983. Their errors were much greater than those observed in recent years in mature democracies, where there is talk about a crisis of this type of surveys. In the primaries, Argentine pollsters widely underestimated the advantage of the Everyone's Front opposition alliance over the governing Together for Change. But after adjusting their methods, in the general election they overestimated that advantage by similar margins. We analyze both failures in an international comparative perspective. Vote intention data for president was used to calculate error indicators and compare them with current international averages and those of past decades, and with errors in several electoral cycles in the US and the UK. We examine ongoing changes in the survey industry, falling response rates, and causes of error identified by new academic studies and work commissioned by professional associations in the field of public opinion research. This literature shows that pre-election polls are no more inaccurate today than in the past and underlines the importance of "non-sampling" errors, particularly those produced by non-response bias, when the probability of participating in a poll is lower for certain groups, such as people with little interest in politics or low social or institutional trust. These types of errors and their causes are now better understood and procedures can be designed to improve the accuracy of estimations. We also discuss the impact of inaccurate polls on political strategies and the media narrative of elections, as well as the difficulty for pollsters, journalists and political actors to acknowledge and communicate the levels of uncertainty associated with electoral predictions. Our analysis highlights basic limitations of the polling industry and public opinion research in Argentina, such as the lack of a professional association and the incipient development of electoral behavior studies. Pollsters were surprised when voters punished the government because of the economic crisis, even though it was the most likely scenario according to a not very large but significant set of empirical studies on economic voting in Argentina and Latin America. The substantial errors in 2019 have implications for the reliability of issue polling, which may show a distorted picture of opinions in Argentine society.

show abstract

Missing Nonvoters and Misweighted Samples

Cited by 42 publications

References 41 publications

Twitter and Facebook are not representative of the general population: Political attitudes and demographics of British social media users

Twitter and Facebook are not representative of the general population: Political attitudes and demographics of British social media users

Correlation with Time Explains the Relationship between Survey Nonresponse and Mass Polarization

La Falla de las Encuestas en las Elecciones Argentinas de 2019. Un Análisis en Perspectiva Comparada Internacional (The Failure of the Polls in the 2019 Argentine Elections. An Analysis in International Comparative Perspective)

Contact Info

Product

Resources

About