Ask: Research and Methods. Volume 20, Issue 1 (2011)
Permanent URI for this collection
A Good Mix? Mixed Mode Data Collection and Cross-National Surveys
Martin, Piotr pp. 5-26
Article description |
Article Full Text PDF
Do More Contact Attempts Reduce Non-Response Bias in Representative Face-to-Face Interviews? Findings from a PAPI Survey with a Low Response Rate
Jabkowski, Piotr pp. 27-58
Article description |
Article Full Text PDF
Developing Attitudinal Indicators of Societal Progress
Harrison, Eric; Jowell, Roger; Sibley, Elissa pp. 59-80
Article description |
Article Full Text PDF
Missing by Design: Planned Missing-Data Designs in Social Science
Pokropek, Artur pp. 81-105
Article description |
Article Full Text PDF
How to Prepare an Advance Letter? The ESS Experience in Poland
Sztabinski, Pawel B. pp. 107-148
Article description |
Article Full Text PDF
Browse
Recent Submissions
Item How to Prepare an Advance Letter? The ESS Experience in Poland(IFiS Publishers, 2011) Sztabiński, Paweł B.When we send advance letters to the sampled persons in a survey, we expect that the recipients will read them and, based on the information provided, will decide on their participation in the survey. Therefore, the letter is expected to play an important role paving the way for the interaction with the interviewer. Findings from in-depth interviews with hard refusers from ESS 3 (2006) and ESS 4 (2008) presented in this paper indicate that such individuals are generally not interested in the mailing received and can hardly remember anything from it. The paper also shows how, following the findings of this research, the advance letters in subsequent ESS rounds in Poland were modified in order to generate recipients’ interest and drive participation.Item Missing by Design: Planned Missing-Data Designs in Social Science(IFiS Publishers, 2011) Pokropek, ArturThis article presents research designs that employ modern statistical tools to optimize costs and precision of research along with some additional methodological advantages. In planned missing-data designs some parts of information about respondent are purposely not collected. This gives flexibility and opportunity to explore a broad range of solutions with considerably lower cost. Modern statistical tools for coping with missing-data, namely multiple imputation (MI) and maximum likelihood estimation with missing data (ML) are presented. Several missing-data designs are introduced and assessed by Monte Carlo simulation studies. Designs particularly useful in surveys, longitudinal analysis and measurement applications are showed and tested in terms of statistical power and bias reduction. Article shows advantages, opportunities and problems connected with missing-data designs and their application in social science researches.Item Developing Attitudinal Indicators of Societal Progress(IFiS Publishers, 2011) Harrison, Eric; Jowell, Roger; Sibley, ElissaThis paper reports progress on a project to develop a set of ‘attitudinal indicators of societal progress’, as part of the wider methodological research agenda associated with the European Social Survey. It recognises the recent contribution of ‘happiness economics’ in moving the debate about the progress of societies ‘beyond GDP’, but seeks to move the agenda on in two ways. Firstly it focuses more on cognitive evaluations of society’s functioning than measures of affect like happiness. Secondly it is less concerned with the psychological wellbeing of individuals and more with the social wellbeing of aggregates, whether these are neighbourhoods, regions or nations. While the study of data relating to overall measures of individuals’ life satisfaction has a long history, the recent Stiglitz-Sen-Fitoussi report called for recognition of the multi-dimensional nature of wellbeing. In examining the feasibility of such an indicator set, we were constrained by two considerations: a) it must comprise items that represented a comprehensive set of domains, in order to provide a full picture of the experiences that constitute societal progress; and b) it must initially be drawn from existing cross-national surveys. Unfortunately it is currently the case that while comparative data exist for most of the key domains, these are often scattered across many different general social surveys, meaning that they can only be analysed at the aggregate level. This makes it difficult to implement another of the Stiglitz recommendations, namely that surveys should ‘assess the links between various quality-of-life domains for each person, and this information should be used when designing policies in various fields’. With this in mind, we present a prototype set of indicators that could be used to measure citizens’ perceptions of the quality of their society. The eventual aim is to field all the items together in one survey, in order to examine their inter-correlations, as well as their relationships with measures of individual life satisfaction and with socio-demographic characteristics.Item Do More Contact Attempts Reduce Non-Response Bias in Representative Face-to-Face Interviews? Findings from a PAPI Survey with a Low Response Rate(IFiS Publishers, 2011) Jabkowski, PiotrThe aim of this article is to examine whether providing pollsters with multiple opportunities to carry out interviews with a sample of randomly selected respondents reduces non-response bias. First, I present the procedures for assessing non-response bias in sociological surveys, opting for a method based on post-stratification weights. Next, using a unique dataset, I address four basic questions related to the research problem: (1) Do more contact attempts reduce the dissimilarity between the sample and population distributions? (2) Do more contact attempts reduce the need for post-stratification weighting? (3) Which categories of respondents are underrepresented and/or overrepresented in the sample in the early and late contact attempts? (4) Do more contact attempts reduce non-response bias? While more contact attempts increased the response rate and closed the gap between sample and population, they do not always reduce the level of non-response bias. Indeed, for some types of variables, the more contact attempts, the larger the non-response bias.Item A Good Mix? Mixed Mode Data Collection and Cross-National Surveys(IFiS Publishers, 2011) Martin, PiotrCan cross-national surveys benefit from mixed mode data collection? This article provides a classification of the different ways in which modes of data collection may be mixed within a cross-national survey, and investigates the methodological consequences of such designs. Mixed mode designs have the potential to lower survey costs relative to single-mode face-to-face surveys, while maintaining higher response rates than cheaper modes alone could. Yet since responses to survey questions are not always independent of the survey mode, mixed mode designs endanger cross-national measurement equivalence (as well as, in the case of time series surveys, diachronic equivalence), so that cross-national comparisons (and analyses of change over time) lose internal validity. These problems can be mitigated by careful questionnaire and survey design, but won’t be entirely overcome in many cases. The use of mixed mode designs in cross-national surveys therefore needs to be accompanied by methodological research to establish the likely consequences for measurement.