Fabrication of interview data
| Published date | 03 April 2018 |
| Date | 03 April 2018 |
| Pages | 213-226 |
| DOI | https://doi.org/10.1108/QAE-06-2017-0028 |
| Author | Jörg Blasius |
| Subject Matter | Education,Curriculum, instruction & assessment,Educational evaluation/assessment |
Fabrication of interview data
Jörg Blasius
Department of Political Science and Sociology,
Rheinische Friedrich-Wilhelms-Universitat Bonn, Bonn,
Nordrhein-Westfalen, Germany
Abstract
Purpose –Evidence from past surveys suggests that some interviewees simplify their responses even in
very well-organized and highly respected surveys. This paper aims to demonstratethat some interviewers,
too, simplify their task by at least partly fabricatingtheir data, and that, in some survey research institutes,
employeessimplify their task by fabricating entire interviewsvia copy and paste.
Design/methodology/approach –Using data from the principal questionnairesin the Programme for
InternationalStudent Assessment (PISA) 2012 and the Programme forthe International Assessment of Adult
Competencies (PIAAC)data, the author applies statistical methods to search for fraudulentmethods used by
interviewersand employees at survey research organizations.
Findings –The author provides em pirical evidence for potential fraud per formed by interviewers and
employees of survey research organizations in several countries that participated in PISA 2012 and
PIAAC.
Practical implications –The proposed methods can be used as earlyas the initial phase of fieldwork to
flag potentiallyproblematic interviewer behaviorsuch as copying responses.
Originality/value –The proposed methodology may help to improve data quality in survey researchby
detectingfabricated data.
Keywords Identical response patterns, Interviewer’s fraudulent behavior,
Satisficing, Undifferentiated response patterns
Paper type Research paper
Introduction
When assessing the quality of survey data, three potential issues that can jeopardize the
validity of large-scale survey data should receive more attention because scholars believe
that:
Interviewees are at least satisfied by their own responses (Krosnick, 1991).
Survey institutes take care that their interviewers do not fabricate interviews
(American Association for Public Opinion Research, 2003).
Employees of survey research institutes do not fabricate interviews.
In the literature, thereis considerable discussion of satisficing (Krosnick, 1991;Krosnickand
Alwin, 1987) and of how to improve questionnaires (Sarisand Gallhofer, 2014), but there is
little discussion of interviewers fabricating data and almost no discussion of data
fabrication by employees of survey research organizations. Using well known, freely
available data sets, this paper focuses on the two last topics, considering interviewer
fraudulence in partly fabricated interviews, that is, some responses were filled in by the
interviewer without respondent input; completely fabricated cases can be detected via
validation callsand should be excluded before publication of the data.
From the focus of the literature, one mightassume that respondents are the only problem
in the interviewing process: it seems that the only requirements are to improve the
Interview data
213
Received30 June 2017
Revised27 October 2017
Accepted18 December 2017
QualityAssurance in Education
Vol.26 No. 2, 2018
pp. 213-226
© Emerald Publishing Limited
0968-4883
DOI 10.1108/QAE-06-2017-0028
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0968-4883.htm
To continue reading
Request your trial