ESRA 2013 Sessions

Effect of nonresponse on results of statistical models 1Professor Christof Wolf
The research on nonresponse has made very important progress these last years by looking to many different aspect of the survey process linked to the nonresponse question and on the way to estimate a potential bias on some indicators. However, research on the effects of nonresponse has mostly focused on possible bias of point estimates and how to correct for this bias. Very little research has been done on the consequences nonresponse can have on estimates of covariance structures and multivariate models; data structures that are much more common in substantive social science research. The aim of this session is to explore how nonresponse can effect estimates of covariance and effect sizes and ways to counteract these effects. All papers -- theoretical contributions, empirical analysis, results from simulations or experimental studies -- are welcome.

Effect of nonresponse on results of statistical models 2Professor Christof Wolf
The research on nonresponse has made very important progress these last years by looking to many different aspect of the survey process linked to the nonresponse question and on the way to estimate a potential bias on some indicators. However, research on the effects of nonresponse has mostly focused on possible bias of point estimates and how to correct for this bias. Very little research has been done on the consequences nonresponse can have on estimates of covariance structures and multivariate models; data structures that are much more common in substantive social science research. The aim of this session is to explore how nonresponse can effect estimates of covariance and effect sizes and ways to counteract these effects. All papers -- theoretical contributions, empirical analysis, results from simulations or experimental studies -- are welcome.

Estimation and Imputation Under Informative Sampling and Nonresponse 1Professor Danny Pfeffermann
Survey data are frequently used for analytic inference on statistical models, which are assumed to hold for the population from which the sample is taken. Familiar examples include the analysis of labour market dynamics from labour force surveys, comparisons of pupils' achievements from educational surveys and the search for causal relationships between risk factors and disease prevalence from health surveys.The sample selection probabilities in at least some stages of the sample selection are often unequal; when these probabilities are related to the model outcome variable, the sampling process becomes informative and the model holding for the sample is then different from the target population model. Another related problem is unit noresponse which again may distort the population model if the response propensity is associated with the outcome of interest, known as not missing at random (NMAR) nonresponse.

Accounting for informative sampling is relatively simple, because the sample selection probabilities are usually known, and several approaches have been proposed in the literature to deal with this problem. On the other hand, accounting for NMAR nonresponse is much harder because the response probabilities are unknown, requiring to assume some structure on the response mechanism, or in the case of longitudinal surveys with attrition, to model the response probabilities as functions of previously measured values.

The aim of this session is to present different practical scenarios giving rise to informative sampling and/or NMAR nonresponse and to discuss alternative approaches to deal with these problems. The focus of the different presentations will be on imputations of missing sample data under NMAR nonresponse and estimation of model parameters and/or finite population means under both informative sampling and NMAR nonresponse.

Estimation and Imputation Under Informative Sampling and Nonresponse 2Professor Danny Pfeffermann
Survey data are frequently used for analytic inference on statistical models, which are assumed to hold for the population from which the sample is taken. Familiar examples include the analysis of labour market dynamics from labour force surveys, comparisons of pupils' achievements from educational surveys and the search for causal relationships between risk factors and disease prevalence from health surveys.The sample selection probabilities in at least some stages of the sample selection are often unequal; when these probabilities are related to the model outcome variable, the sampling process becomes informative and the model holding for the sample is then different from the target population model. Another related problem is unit noresponse which again may distort the population model if the response propensity is associated with the outcome of interest, known as not missing at random (NMAR) nonresponse.

Accounting for informative sampling is relatively simple, because the sample selection probabilities are usually known, and several approaches have been proposed in the literature to deal with this problem. On the other hand, accounting for NMAR nonresponse is much harder because the response probabilities are unknown, requiring to assume some structure on the response mechanism, or in the case of longitudinal surveys with attrition, to model the response probabilities as functions of previously measured values.

The aim of this session is to present different practical scenarios giving rise to informative sampling and/or NMAR nonresponse and to discuss alternative approaches to deal with these problems. The focus of the different presentations will be on imputations of missing sample data under NMAR nonresponse and estimation of model parameters and/or finite population means under both informative sampling and NMAR nonresponse.

Investigating Non Respondents: How to Get Reliable Data and How to Use ThemDr Michele Ernst Staehli


Not able to participate: a neglected cause of nonresponse 1Dr Ineke Stoop
There are two main causes of survey nonresponse: refusal and noncontact. It can also happen that a person is not able to participate because of language problems, or because someone is mentally or physically not able to participate. Sometimes "being unable" is the consequence of design aspects (someone who cannot hear cannot participate in a telephone survey, and someone who is blind cannot complete a paper questionnaire). In these cases practical solutions can be found. In other cases someone may not be able to answer survey questions for more general reasons: persons with severe learning disabilities may not be able to participate in a general survey whatever the mode. They may however be able to participate in a survey that has especially been designed to accommodate these groups.

Not being able to participate is usually assumed to be a minor factor in survey nonresponse and the bias caused by this is usually assumed to be small. In the European Social Survey, however, the share of persons not being able to participate cannot be ignored in a number of countries.

We invite submissions that address : a) the measurement of reasons for nonparticipation, b) the relationship with design issues (survey mode, sampling frame) and reasons for nonparticipation, c) ways to minimise nonparticipation due to not being able (and potential effects on measurement error and comparability), and d) effects of "not able" on nonresponse bias.

Not able to participate: a neglected cause of nonresponse 2Dr Ineke Stoop
There are two main causes of survey nonresponse: refusal and noncontact. It can also happen that a person is not able to participate because of language problems, or because someone is mentally or physically not able to participate. Sometimes "being unable" is the consequence of design aspects (someone who cannot hear cannot participate in a telephone survey, and someone who is blind cannot complete a paper questionnaire). In these cases practical solutions can be found. In other cases someone may not be able to answer survey questions for more general reasons: persons with severe learning disabilities may not be able to participate in a general survey whatever the mode. They may however be able to participate in a survey that has especially been designed to accommodate these groups.

Not being able to participate is usually assumed to be a minor factor in survey nonresponse and the bias caused by this is usually assumed to be small. In the European Social Survey, however, the share of persons not being able to participate cannot be ignored in a number of countries.

We invite submissions that address : a) the measurement of reasons for nonparticipation, b) the relationship with design issues (survey mode, sampling frame) and reasons for nonparticipation, c) ways to minimise nonparticipation due to not being able (and potential effects on measurement error and comparability), and d) effects of "not able" on nonresponse bias.

Participation rates and recruitment methods for health examination surveysDr Hanna Tolonen
Health examination surveys, which include both questionnaires and physical measurements, and often also collection of biological samples like blood and urine, are challenging for the survey participants. These surveys require more time than questionnaire based surveys and they can cause some uncomfort for survey participant through blood sample collection etc. Examinations are also in many surveys conducted in health clinics, assuming that survey participants are physically fit to travel there and have sufficient time and form of transportation to do that.

The participation rates of health examination surveys have declined in past decades in similar pattern than for other surveys. What is known about non-participants to the health examination surveys and their effect on survey representativeness of the results is still limited. Some studies have shown that non-participants have at least twice as high mortality risk than participants, implying that non-participants are less healthy and have worse lifestyle than survey participants. More studies on characteristics of health examination survey non-participants are needed to better understand which population groups are not participating. This information could also help to target recruitment methods and to develop new ways to recruit survey participants.


The Contribution of Paradata in Analysing Unit Nonresponse Processes and Nonresponse BiasMs Verena Halbherr
Researchers are invited to submit presentation proposals at the session "The contribution of Paradata in Analysing Nonresponse Processes and Nonresponse Bias" at the European Survey Research Association conference, July, 15-19, 2013 in Ljubljana, Slovenia.

Unit Nonresponse is one of the major issues affecting data quality in surveys. Nonresponse occurs in every survey and may cause biases in estimates. Over the last decades an increase of unit nonresponse has been observed generating a growing interest in understanding nonresponse processes.

One promising way of analysing nonresponse is by the use of auxiliary variables, especially paradata. These data might include information on contact strategies, interviewer observations, interview duration and time stamps.

This session focuses on two aspects of survey nonresponse: Analysing the processes leading to nonresponse during fieldwork and the induced nonresponse bias. This might include presentations on fieldwork monitoring and responsive designs, as well as the use of response propensities and R-indicators. We specifically encourage submissions using paradata in the nonresponse analyses of face-to-face surveys based on strict probability samples.


The use of respondent incentives in face-to-face surveys: Effects on response rates, survey error and survey costsMr Klaus Pforr
Decreasing response rates have become a major concern for face-to-surveys in modern societies in the last decades. To counter this downward trend, one possible and often used measure is respondent incentives. Incentives are used in a large variety of forms, modes and value, and survey modes.

While there is evidence for a positive effect of incentives on response rates, there is still debate on the effects on survey error. Respondent incentives may increase sample selectivity by attracting a specific subset of respondents over proportionally to the survey. Incentives may also systematically change answers by survey respondents, thus producing measurement error, by changing survey participants' perception of the study or their motivation for participation.
As incentives increase direct survey costs, research on the overall cost effectiveness of incentives is needed. Incentives may ease contact processes or increase data quality, thus save the survey enterprise fieldwork or data editing costs.

Contributions sought for this session will address one or more of following research questions:
* Effects of respondent incentives on survey outcome: How are contact, cooperation, response rates influenced by incentives?
* Effects of respondent incentives on sample composition and nonresponse bias. Do incentives differentially affect the response propensity of various subgroups of the population?
* Effects of respondent incentives on measurement error. Do respondent incentives change the response behavior during the interview?
* Are respondent incentives cost effective? Can savings in terms of fieldwork effort outweigh the direct expenses for incentives?
We want to focus the session on contributions from large-scale face-to-face surveys. We prefer results from experimental studies; however, all studies addressing the research questions are welcome.