Assessing the Quality of Survey Data

Jörg Blasius is a professor of Sociology at the Institute for Political Science and Sociology, University of Bonn, Germany. His research interests are mainly focused on explorative data analysis, especially correspondence analysis and related methods, data collection methods, sociology of lifestyles, and urban sociology. He is a coeditor of the SAGE Series “Survey Research Methods in the Social Sciences.” Together with Michael Greenacre (University Pompeu Fabra, Barcelona), he founded Carme (Correspondence Analysis and Related Methods Network, see www.carme-n.org). From 1998 to 2016, he belonged to the Board of RC33 (Research Committee on Logic and Methodology in Sociology) of the ISA (International Sociological Association), from 2006 to 2010 he served as president. Website: https://www.politik-soziologie.uni-bonn.de/de/personal/Prof.-Dr.-Joerg-Blasius

This is an innovative book. Blasius and Thiessen show how careful data-analysis can uncover defects in survey data, without having recourse to meta-data or other extra information. This is good news for researchers who work with existing data sets and wish to assess their qualityJoop HoxProfessor of social science methodology, Utrecht University ′This illuminating and innovative book on the quality of survey data focuses on screening procedures that should be conducted prior to assessing substantive relations. Numerous tables and visual displays based on appropriate analyses show convincing examples of procedural deficiencies, data duplicates, and faked interview, and unusual patterns in well known international datasets. Blasius and Thiessen discuss openly what is mostly covered in survey reporting, and they show ways to improve the reliability of conclusions based on survey data. A must for survey practitioners and usersJaak BillietEmeritus professor in social methodology, University of Leuven ′I hope that many social science researchers will read Jörg Blasius and Victor Thiessen′s book and realize the importance of the lesson it provides, which is that many of our data sets contain serious errors which may lead to wrong conclusions if not wrong theories. Applying the suggested tests can prevent this.Willem SarisDirector of RECSM, Universitat Pompeu Fabra, Barcelona This is an important, groundbreaking analysis of survey data quality that serves as an indispensable vade mecum for anyone engaged in gathering, analysing or utilising survey data. Clearly quality is not always what is appears to be... Assessing the Quality of Survey Data... addresses this issue in far more detail than any other work known to this reviewer. Maryam Nazari and G. E. GormanOnline Information Review Assessing the Quality of Survey Data has been written for all researchers who want to be good data detectives... This book has some considerable strengths. It really got me to reflect on the assumptions I had made about data sets I had handled. It highlighted how relatively easy it is to spot data issues early on in analysis. It also made me think of response style challenges to be more than a survey design/sampling issues and to include much more data screening. Although I won’t be setting up my own detective agency, I have certainly picked up some sleuthing tips.Catherine OwensSR: Reviews

• Conceptualizing Data Quality: Respondent Attributes, Study Architecture, and Institutional PracticesConceptualizing Response QualityStudy ArchitectureInstitutional Quality Control PracticesData Screening MethodologyChapter OutlineEmpirical Findings on Quality and Comparability of Survey Data Response QualityApproaches to Detecting Systematic Response ErrorsQuestionnaire ArchitectureCognitive Maps in Cross-Cultural PerspectiveConclusionStatistical Techniques for Data ScreeningPrincipal Component Analysis Categorical Principal Component Analysis Multiple Correspondence Analysis ConclusionInstitutional Quality Control PracticesDetecting Procedural DeficienciesData DuplicationDetecting Faked and Partly Faked InterviewsData Entry Errors ConclusionSubstantive or Methodology-Induced Factors? A Comparison of PCA, CatPCA and MCA SolutionsDescriptive Analysis of Personal Feelings DomainRotation and Structure of DataConclusionItem Difficulty and Response QualityDescriptive Analysis of Political Efficacy DomainDetecting Patterns with Subset Multiple Correspondence AnalysisModerator EffectsConclusion Questionnaire Architecture Fatigue EffectQuestion Order EffectsMeasuring Data Quality: The Dirty Data IndexConclusionCognitive Competencies and Response QualityData and MeasuresResponse Quality, Task Simplification and Complexity of Cognitive MapsConclusionConclusion