Henri Leridon
Henri Leridon, scientific integrity referent for INED, presents INED’s proposed charter on survey data accessibility (open science).
(Interview conducted in May 2021)
What does the data accessibility principle laid down in France’s October 2016 “Digital Republic law” cover?
The text specifies that all data collected as part of either French- or European-funded projects that are “not protected by a specific right or regulation” and “have been made public by the researcher or research institution” can be “freely reused.” The principle of making data widely available has therefore been clearly stated: the only requirement is for the data to have already been made public. But the text does not specify implementation modes.
The INED charter thus presents a framework for data accessibility. The prerequisite is a digital file in an accessible format containing data that have been checked and documented. Unless those criteria are met, the data do not really exist. Next an embargo period has to be agreed on during which the data will be accessible only to the given survey’s designers and operators, who can legitimately claim priority access because without their work, often conducted over several years, those data would not have existed. Obviously the embargo period has to be limited; the INED charter specifies a maximum of three years after the completion of data collection, though exemptions to this rule will be possible.
What is at stake in making data like those collected by INED—i.e., quantitative social science data on individuals—accessible?
As a research institution, INED has always been and continues to be a remarkable data collector; the methodologies used are constantly being improved. This is what makes INED one of the finest social science survey institutes. Its researchers often work in association with specialists from other institutions, giving them access to excellent materials for push research forward in their fields—as long as they have the resources to do so. It is important that the mass of data thus created be accessible to the entire French and international researcher community so that it can be used as intensively as possible. Therein lies the Institute’s second responsibility—and it is of course in survey designers’ interest to have their work used and built on! All of this requires the data to be described and documented in detail, to preclude erroneous interpretations by secondary users—specifically, inaccurate use of weighting.
And in practice?
For all survey proposals, authors have to submit a “data management plan” explaining how the file of collected data will be developed and stored—for example, on France’s Quetelet-Progedo platform. They also have to meet the EU’s General Data Protection Regulation (GDPR), aimed among other things to guarantee respondents’ consent to answer questions and the preservation of respondent anonymity after data collection. Last, they must reach an agreement with the head of the institution on the length of the embargo period—an agreement that takes into account the specificities of the given survey as planned. This will allow them to use their data without too much pressure during the embargo period while complying with “open science” objectives.