Quality & Quantity 33: 215–218, 1999.
© 1999 Kluwer Academic Publishers. Printed in the Netherlands.
Longitudinal Research: A Bridge Between Quantitative and
Qualitative Social Research?
In the social sciences, longitudinal analysis is, simultaneously, a necessity, a luxury
and a riddle.
It is a necessity because one assumes that the actor’s experience, including
its length and the speciﬁc historical period in which it took place, will have had
a precise inﬂuence on the actor’s behaviour and sociologists, unlike economists,
think that one ought to take this factor into account. As will be seen in this issue
of Quality and Quantity, different methods of longitudinal analysis are brought
into play. Here ‘longitudinal’ is increasingly being taken to mean not so much the
classic, retrospective reconstruction of life histories but rather a whole series of
methodologies and analytical techniques for collecting data – from panels (see the
contribution by Trivellato) to archives on users – which track the same subjects
over time. Indeed, retrospective studies are somewhat limited, both because the
researcher must necessarily reconstruct the past in a simpliﬁed manner and, above
all, because of the memory lapses, or distortions, interviewees are subject to when
remembering past events. The new techniques are trying to overcome these limits
and, today, are able to offer a more and more articulated picture of life histories,
family and individual strategies, the concatenation of events, the impact of the
duration over time of certain situations and so on.
Longitudinal analysis is a luxury in two ways. Literally, because gathering and
interpreting longitudinal data is extremely expensive: especially if one is seeking
reliable information about the behaviour of a large population – sub-divided into
groups with different habits/customs – over a fairly long time period within which
both persistence over time and substitution of cohorts from generation to gener-
ation are important if one is to be able to understand the phenomena of social
change better. However, apart from the direct economic costs, which are always
high, above all for panel studies, longitudinal analysis is a luxury also because of
the cost of the humans involved in interpreting the results. Longitudinal studies are
multipliers of information in the sense that variables assume different meanings at
different times, which are contemporaneously historical time periods – hence, peri-
ods during which all subjects have the same experiences: of war, peace, changes of
government, economic boom or slump, etc. – and moments in the lives of subjects
with different ages (see the contribution by De Graaf). Furthermore, in many cases,
the length of the experience itself is the most important factor. Basically, one should
not be content with knowing, for example, that the subject is unemployed, but