New data strategies: nonprobability sampling, mobile, big data

Publication Date03 April 2018
Pages303-314
DOIhttps://doi.org/10.1108/QAE-06-2017-0029
AuthorMichael Link
SubjectEducation,Curriculum, instruction & assessment,Educational evaluation/assessment
New data strategies:
nonprobability sampling, mobile,
big data
Michael Link
Data Science, Surveys and Enabling Technologies (DSET), Abt Associates,
Rockville, MD, USA
Abstract
Purpose Researchers now have more ways than ever before to capture information about groups of
interest. In many areas, these are augmenting traditional survey approaches in others, new methods are
potential replacements. Thispaper aims to explore three key trends: use of nonprobability samples,mobile
data collectionand administrative and big data.
Design/methodology/approach Insights and lessonslearned about these emerging trends are drawn
from recent publishedarticles and relevant scientic conference papers.
Findings Each new trend has its own timeline in terms of methodological maturity. While mobile
technologies for data capture are being rapidly adopted, particularly the use of internet-based surveys
conducted on mobile devices, nonprobability sampling methods remainrare in most government research.
Resource and quality pressures combined with the intensiveresearch focus on new sampling methods, are,
however, making nonprobability sampling a more attractive option. Finally, exploration of big datais
becoming more common,although there are still many challenges to overcome methodological,quality and
access beforesuch data are used routinely.
Originality/value This paper provides a timely review of recent developments in the eld of data
collectionstrategies, drawing on numerous currentstudies and practical applications in the eld.
Keywords Big data, Survey, Mobile, Sampling, Mobile app, Nonprobability
Paper type Viewpoint
For more than a decade, the eld of survey research has faced a series of mounting
challenges as many of the suppositions underlying the theory of how high-quality surveys
should function rarely hold in practice if, in fact, they ever did. Even the most well-
designed and well-funded surveys these days face a litany of practical issues from
incomplete sampling frames, rising nonresponse, recognition that factors such as language,
culture and cognitive understanding can lead to a variety of interpretations of even well-
written questions, variations produced by modes of survey administration, and a series of
other problems. All of this has led to a corresponding rise in costs for conducting well-
designed surveys. Debates havealso been playing out within the survey industry itself. We
have witnessed a shift in discussions over what constitutes a goodresponse rate to
debates about whether response rates are even important at all (Groves, 2006). We have
gone from (fairly) strict ideas of what constitutes a proper probability sampleto serious
questions about whetherprobability samplingis an essential component of social research
(Baker et al., 2013). More recently,there is debate as to whether surveys themselves are even
necessary or if questions can be answered using the volumes of data generated across the
digital world –“big data(Japec et al., 2015). Today more than ever, survey practice rarely
reects survey theory.
New data
strategies
303
Received30 June 2017
Revised23 October 2017
Accepted13 December 2017
QualityAssurance in Education
Vol.26 No. 2, 2018
pp. 303-314
© Emerald Publishing Limited
0968-4883
DOI 10.1108/QAE-06-2017-0029
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

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT