"New Directions In Survey Research" Seminar Series (March-May 2010)

Session 1: Embedding Experiments in Surveys (March 1, 2010)
Alexander Kuo (Juan March Institute, Stanford University);Marieke van Londen (Radboud University Nijmegen) and Marcel Coenders (Utrecht University)

Session 2: Combining Genetic and Public Opinion Data (March 22, 2010)
Chris Dawes (University of California, San Diego) and Peter John Loewen (University of Toronto)

Session 3: Potential and Limits of Online Surveys (May 4, 2010)
Willem Saris (Universitat Pompeu Fabra), Annette Scherpenzeel (CentERdata, Tilburg University), and Melanie Revilla (Universitat Pompeu Fabra)
 

Session 1: Embedding Experiments in Surveys (March 1, 2010)
Alexander Kuo (Juan March Institute, Stanford University);Marieke van Londen (Radboud University Nijmegen) and Marcel Coenders (Utrecht University)

"Measuring Individual Identity: Experimental Evidence" (.pdf)

by Alexander Kuo (Juan March Institute, Stanford University) and Yotam Margalit (Columbia University)

What determines the identity category people feel they most belong to? What is the political significance of one's proclaimed identity? Recent research addresses this question using surveys that explicitly ask individuals about their identity. Yet little is known about the nature of the attachments conveyed in responses to identity questions. We conduct a set of studies and experiments that investigate these reported attachments. Our findings suggest that: (1) the purported identity captured in survey responses varies significantly within subjects over time; (2) changes in people's primary identity can be highly influenced by situational triggers; (3) changes in purported self-identity do not imply a corresponding change in policy preferences. Our results are drawn from three studies that vary in terms of design, country sample, and research instrument. The findings have implications for research on identity choice, as well as on the use of surveys in studying the role of identity in comparative politics.

 

"Opposition to employment policy: differences in attitudes towards ethnic and income policy among Dutch majority members" (.pdf)

by Marieke van Londen (Radboud University Nijmegen), Marcel Coenders (Utrecht University) and Peer Scheepers (Radboud University Nijmegen)

Since the mid-nineties, the Dutch government has abandoned many of its ethnic policies in favour of new 'integration' policies that focus on the socio-economic dimension of integration and target a broader category of disadvantaged people. We examine to what extent the attitudes of Dutch majority members towards employment policy vary with the policy's target-group. In a split-ballot design respondents were presented with the same policy issue, but the target-group, 'ethnic minorities' or 'people in poor neighbourhoods', varied randomly. The results indicate that opposition to employment policy is low and does not vary with the policy's target-group. Explanations for opposition to employment policy however, do vary with the policy's target-group. Perceived threat from ethnic minorities is more consequential for opposition to ethnic-targeted policy than for opposition to income-targeted policy.

 

Session 2: Combining Genetic and Public Opinion Data (March 22, 2010)
Chris Dawes (University of California, San Diego) and Peter John Loewen (University of Toronto)

The presentation will cover five things. First, we will discuss the assumptions that motivate the study of genopolitics. This includes discussion of the evolution of cooperation and the importance of prosocial orientations for modern politics. Second, we will discuss evidence from twin studies on the role of genetics in political attitudes and participation. Third, we will discuss evidence from gene association studies about political attitudes and participation. Fourth, we will discuss some new findings related directly to survey research, namely whether there is a genetic basis to survey non-response. Fifth, we will discuss future directions for genopolitics research.

"The CHRNA6 Gene, Patience, and Voter Turnout" (.pdf)

by Chris Dawes (University of California, San Diego) and Peter John Loewen (University of Toronto)

This paper explores the relationship between a gene known to regulate patience and self-reported voter turnout. Given that the costs of voting are generally required before the benefits of voting are realized, a rational calculus of voting would suggest that citizens with higher discount factors are less likely to participate. This intuition has been demonstrated empirically (see Fowler and Kam 2006). Using ADD Health, a large representative sample of American youth, we explore whether two genes known to regulate impulsivity are related to self-reported turnout in an American election. After a series of tests, we find that one single nucleotide polymorphism on one of these genes is related to turnout. We then apply further tests to check the robustness of this finding. We present this paper as an example of how genetic information on survey respondents can be used to explore gene association.

 

"The Heritability of Duty and Voter Turnout" (.pdf)

by Peter John Loewen (University of Toronto) and Chris Dawes (University of California, San Diego)

In this paper we use another technique, namely twin studies. Using a sample of approximately 300 pairs, we explore how much variance in the belief that voting is a duty is attributable to genes. We derive a heritability estimate of about one third of variance. This work demonstrates how twin studies, coupled with survey responses, can be used to test an attitude or disposition thought to be central to the decision to participate in politics.

 

"COMT, Extraversion and Partisan Attachment" (.pdf)

by Christopher T. Dawes (University of California, San Diego)

In this paper, the author shows that the personality trait of extraversion mediates the relationship between a gene (COMT) and partisan attachment. This paper leverages survey data on personality types. It also presents a replication of results out-of-sample, which is an increasingly standard practice in behaviour genetics research. In doing so, it provides a example of an empirical framework for future genopolitics research.

 

Session 3: Potential and Limits of Online Surveys (May 4, 2010)
Willem Saris (Universitat Pompeu Fabra), Annette Scherpenzeel (CentERdata, Tilburg University), and Melanie Revilla (Universitat Pompeu Fabra)

"Something has to be done to protect the public against bad web surves"

by Willem Saris (Universitat Pompeu Fabra - RECSM)

In this paper we discuss consequences for political processes of the recent paradigm change in survey research. Originally survey research was based on probability sampling, nowadays a lot of surveys are based on self selected access panels. This means that the statistical basis for generalization from the sample to the population does not exist anymore. Some alternative procedures, like weighting based on propensity scores, have been suggested to correct for possible biases. However, the empirical studies done to compare the results of access panels and probability samples are so far inconclusive. Sometimes large differences are found and it was not possible to correct for the differences. We also show that information from surveys plays a role in several different political processes such as voting, participation in elections, agenda setting, opinion forming about issues and in reaction to general political orientation of the public. Given that one can not say whether an access panel provides proper information about characteristics of the population, one runs the risk that political processes are influenced by wrong information based on access panels. This leads us to the conclusion that something has to be done to protect the public against wrong information based on bad survey research. In the Netherlands independent  agencies have been created to protect its citizens against bad quality of food and other goods, but also to check whether the media stick to the rules specified and the advertisements are in agreement with the accepted quality rules. These agencies have the right of research and to condemn companies and give them fines, if the quality is below the quality standards  Similar agencies with comparable rights are needed to check the quality of survey research in order to protect the citizens in the different countries against false information about the so called public opinion.

 

"Innovations and New Technologies in Panel Research"

by Annette Scherpenzeel (CentERdata, Tilburg University)

Survey research via the Internet has increased significantly in recent years. Several methods are used to attempt to survey the general public by Internet, including the use of volunteer and RDD samples. However, these methods may not fulfill the demands on coverage, sampling and response posed by scientific researchers. People in volunteer samples are often a rather selective part of the general population. In addition, because of vast increase of mobile-only households and declining response rates, RDD samples may not cover the population anymore. A newly established Internet panel in the Netherlands uses a different design. A true probability sample of households drawn from a population register by Statistics Netherlands is contacted with a household in-person interview, asking respondents to join the panel. The panel provides a computer and Internet connection to those households that could not otherwise participate (about 85% of households have internet access). This paper presents an overview of this new panel concept and evaluates whether the design would also be feasible in other countries having high Internet penetration rates, and could overcome the coverage problems of volunteer samples. In addition, it reports the results of some experiments carried out on the use of incentives to increase the participation in the panel and on the data quality of internet data compared to telephone and face-to-face data.

 

"A comparison of surveys using different modes of data collection: European Social Survey versus LISS panel"

by Melanie Revilla (Universitat Pompeu Fabra - RECSM)

Web surveys are becoming more and more popular in survey research, mainly because of their low costs. With the increase of the Internet coverage in most European countries, the response rates are becoming high enough to collect huge amount of data in a short period of time. However, there is a risk that changing to this new mode would lead to data incomparable with data collected in the past. Therefore it is necessary to check if data collected using Web and data collected with more traditional modes (mainly mail, telephone, face-to-face) produce similar results. This paper compares one survey completed by the Longitudinal Internet Studies for the Social sciences (LISS) panel (Web panel based on probability sample) in December 2008 with the same questions asked in the frame of the European Social Survey (face-to-face) in the Netherlands. Focusing on the quality of the questions of single items and composite scores, we find few differences between these two surveys.