Interview with 2024 ELIPSS module laureate Constance FROHLY (PhD student)

Hello ! My name is Constance Frohly and I'm a doctoral student at Sciences Po, researching for my doctoral thesis under the supervision of Pierre Cahuc and Emeric Henry. I'm working on the choice of schools made by families at the junior high school level (college).

In particular,  I'm interested in what information families have about public and private secondary schools when they must choose a high school for their children when the latter are moving on from primary school education to secondary school education. 

The academic literature shows that families from wealthier backgrounds are more represented in private schools. This could be due to any  number of factors, both in terms of family choices (preferences, financial constraints) and schools (selection). The first step then so as to disentangle these differentiating mechanisms is to ask parents what they know about state and private secondary schools, what they prefer, and why.

I mainly use quantitative data, i.e. I compare what participants answer on average, depending on their socio-demographic characteristics, their location, their experiences at junior high school, etc. This method is based the power of statistics - the more participants there are, the more reliable the result - hence the advantage of the ELIPSS panel! 

In addition to the ELIPSS panel, I have developed an experimental approach, comparing, for example, the responses of participants who are given information about public or private colleges.  The objective is to understand whether more information contributes to people's decision-making or attitudes. The results of this experiment could, for example, inform public debate or help model a public information policy.

I responded to the call for doctoral projects because the question of what average information about colleges has been poorly documented in the past: little information exists on the quality and sources of information for families. To better inform my research hypothesis, the variables of interest with respect to parents, and the mechanisms at work, I needed to have a better idea of what the public actually thinks about state and private secondary schools and, in particular, particularly with respect to certain characteristics such as the Brevet pass rate and social position.

My initial analysis shows that the criteria that matter most when parents choose between state and private secondary schools are similar for all social classes. The most important criteria are the number of non-education related incidents (theft, harassment, etc.) per establishment and the number of teacher absences, followed by geographical distance and the number of pupils in the classes, the child's wishes, the options available and the school's added value (i.e. the school's contribution to the pupils' academic performance).

The results also show that, on average, individuals overestimate the number of pupils per class in state secondary schools, and conversely underestimate the number of pupils per class in private secondary schools. In fact, there are, on average, more pupils per class in private colleges (27.3) than in state colleges (24.7). What's more, the success rate for the Brevet is generally higher than people think, in both state and public schools.

According to initial analyses, the different social classes seem to choose between state and private secondary schools based on the same criteria and therefore have similar preferences. Surprisingly, however, the least privileged social categories declare a stronger preference for private schools. This could perhaps be due to the fact that they live geographically closer to public schools that are worse than average, but this remains to be seen...

It was a great opportunity that I would absolutely recommend! It has enabled me to obtain valuable, practical information in the field, which is otherwise very expensive and complicated for PhD students to obtain. In addition, the opportunity to discuss the proper wording of survey questions so as to ensure their understanding by participants with the people in charge of the ELIPSS panel was also a very enriching experience.

When submitting your application, I would advise you to concentrate more on the objective of the questionnaire and on the way the questions are linked together rather than on the precise wording of each question, as they can still be modified a little later. Don't hesitate to ask people around you to test your questionnaire too!