Quality & Quantity 33: 339–352, 1999.
© 1999 Kluwer Academic Publishers. Printed in the Netherlands.
Issues in the Design and Analysis of
Panel Studies: A Cursory Review
Dipartimento di Scienze Statistiche, Università di Padova, Via San Francesco 33, 35121 Padova,
Italy, phone: 049/827-4168; fax: 049/875-3930; e-mail: firstname.lastname@example.org
Abstract. This paper offers a broad review of some aspects in the design and analysis of panel
studies, chieﬂy of household panel surveys. Both the analytic beneﬁts and the potential problems
of panel surveys are brieﬂy outlined, and selected methodological and operational issues, which
crucially affect data quality are highlighted. These questions are then considered under four headings:
(i) dynamic population and its implications for initial sampling and following rules; (ii) panel length
and number of waves; (iii) tracking and tracing techniques, and other strategies for maintaining high
participation rates; (iv) questionnaire design and strategies for collecting retrospective information.
While no technical details are offered, there is some discussion of the possible drawbacks and
advantages of the different approaches described.
Key words: longitudinal analysis, panel conditioning, panel survey, questionnaire design, recall
errors, sample attrition.
Over the past two decades, one of the most important developments within ofﬁcial
statistics, and social research, has been the increased availability of panel data on
(samples of) individuals and/or households. This type of data base can be obtained
in various ways: (i) by linking administrative records (to which population census
data can be added); (ii) through a single cross-sectional survey, which collects
retrospective relevant information over a fairly long period of time; (iii) through a
panel survey, which means surveying a given sample of individuals and/or house-
holds and following them over time with a sequence of rounds of data collection
(or ‘waves’, as they are often termed); (iv) through a judicious mixture of the
Recent developments in longitudinal research clearly depend on the fact that
suitable panel data sets have become available to analysts. This is not precisely the
case for Italy, where such data sets are, or at least were, often hard to come by
(see Ghellini & Trivellato, 1996: 303–309, for a brief survey). Indeed, in order to
remedy this lack of suitable data, researchers in Italy have frequently had to exploit
longitudinally data collected by means of surveys that were not originally designed
with a proper longitudinal perspective – typically, repeated cross-sectional surveys