Albane Gourdol
INED has been conducting surveys since its founding, as attested by surveys on the intellectual levels of children and on family living conditions and housing aspirations, two studies headed by Jean Stoetzel and Alain Girard in INED’s first years of existence, just after 1945. On Stoetzel’s return from the United States, and after he founded France’s IFOP public opinion polling institute, Stoetzel started using probabilistic sampling methods at INED, which gradually superseded the quota method. In 1954, Alain Gerard “Choosing a spouse” survey powerful impacted and shaped INED’s future survey history. The first INED surveys were quite extensive: the one on “mentally deficient” children encompassed 100,000 children. The Institute was soon pioneering surveys on subjects like family budgets, which larger research institutes would later adopt and continue.
As part of our celebration of the Institute’s 80th anniversary, Albane Gourdol, head of the Surveys support service, reiterates the importance of INED’s major surveys and the issues involved in recent technological and methodological developments.
(Interview conducted in January 2025)
What are some specificities of INED surveys?
Surveys conducted at INED are qualitative or quantitative, and they reflect the wide range of research subjects areas and themes studied at Institute. They are specific in several ways. In general, they focus on vulnerable groups that can be difficult to question (unhoused people, former prison inmates, people who have used assisted reproductive technologies, young people taken in by France’s child social welfare agency, drug users, migrants, and others). They investigate sensitive subjects, such as end of life, violence victims’ experience, life histories of women living with breast cancer. New research topics and areas are continuously being added, thereby producing singular, often exploratory, first-of-a kind surveys on new subjects that may later inspire other research institutes.
Our surveys also use highly innovative methodologies For example, “Families and Employers” is one of very few surveys that cover both households and their members’ employers. The “ChiPRe” survey (Chinese immigrants in the Île-de-France region, encompassing Paris) used a sampling method that had never been tested in France called Network Sampling with Memory, specially adapted to questioning small or difficult-to-reach populations. Moreover, in some cases our questionnaires are available in several languages to adapt to the particular set of respondents. For example, respondents in the “EXPRESS” survey (Prison experiences and post-release social reintegration), could answer in French, English, or Arabic.
INED life and event history surveys use Age-Event questionnaire forms (Ageven) to which new dimensions and themes have been added, the aim being to move beyond collecting strictly individual information and take into account observation and analysis units that are more relevant to the particular research question.
INED surveys are “custom-made,” they are crafted, in the sense they draw on particular know-how and go beyond the standard large survey protocols used in public statistics.
How has the work of producing surveys evolved and how is it being adapted to new technologies?
Several factors have recently modified survey design: a trend of diminishing survey participation rates, budget constraints, a more complex legal environment, stricter legislation on telephone surveys, and match our data with external administrative data to reduce questionnaire length.
Now that it’s possible to collect data via the internet, and at a time when survey response rates are not as strong due to the increasing difficulty of contacting households by phone, the use of multimode survey protocols has become a strong strategic orientation at INED as in public statistics institutes generally.
By combining several collection methods, we can get the best of each in connection with budget and field study constraints, particular survey subject, and targeted group or population. Since face-to-face and telephone interviews are relatively costly, they’re now reserved for issues, themes, population groups that can only be studied that way.
Although multimode survey protocols offer more ways of contacting respondents, they also complexify survey design and have to be adapted in different ways to ensure result quality. Some of the questions that have to be dealt with are the questionnaire and questionnaire length, protocol and data collection design, and post-collection statistical data-aggregation processes. Moreover, respondents answer differently depending on whether they’re looking at a screen, an interviewer, or talking to an interviewer on the phone. Statistical studies are now underway at INED, and in the world of public statistics at large, to measure the effects of data collection mode on findings.
What are the geographic scales of INED surveys, and can are they evolving toward more international themes?
The diversity of INED research means our surveys are conducted in a vast range of geographic locations: metropolitan France, French overseas departments and territories, disadvantaged neighborhoods, rural areas, and abroad. The Institute opened up to international perspectives and studies quite some time ago: in the 1950s INED was already running a survey in Canada, followed by others in Africa from the 1970s. There is one clear trend, however: an increase in surveys by European research consortiums which produces results that can be compared across European countries or at other European geographic levels.
