CONCOURS IEH01 : Chargé d`études administratives

INRA-DRH | numconcours externes 2011
Page 1 of 2
IR | Ingénieur de Recherche en production et analyse de données
profil n° :
IR11-SAE2-1
emploi-type n°:
D1D24
concours :
IRD01
Lieu du concours (Centre Organisateur):
Paris
Paris
::::::::::::::::::::::::::::: FICHE DE PROFIL || Concours Externes INRA 2011 :::::::::::::::::::::::::::::
Dans le cadre de la problématique d'un projet de recherche, ou en réponse à une demande externe l'ingénieur de recherche conçoit et pilote,
en expert, des dispositifs d'enquêtes de terrain et conduit l'analyse et le traitement des données recueillies. Il garantit la qualité et la validité
des données produites et des traitements réalisés.
.......................................................................................................................................................................................................
Domaine professionnel : Sciences humaines et sociales/ BAP D | Sciences humaines et sociales
Département : Dpt Sc Soc Agricult Alimentation Esp Environm
unité n° 1303 ALISS | Alimentation et Sciences Sociales
personne(s) à contacter :
SOLER LouisGeorges
01.49.59.69.82
[email protected]
.......................................................................................................................................................................................................
Environnement :
La personne recrutée exercera son activité au sein d'une unité de recherche de l'INRA spécialisée dans l'analyse économique et sociologique
de l'alimentation dans sa double dimension comportement du consommateur et stratégie des firmes agroalimentaires et de la distribution. Il
(elle) aura pour tâche principale de piloter les travaux que mène l'INRA dans le cadre de l'Observatoire de l'alimentation (OQALI) et qui ont
vocation à recueillir, assembler, traiter et analyser les données concernant les caractéristiques des produits alimentaires.
Tout en étant un expert de la recherche et de la production de données, il(elle) s'investira dans les études et expertises qui les mobilisent et
sera la cheville ouvrière des liens entre les données produites par l'OQALI et les activités de recherche menées dans l'unité.
Il(elle) assurera ainsi les liens entres les unités de recherche, les lieux de conservation de données (organismes de production et de mise à
disposition de données, observatoire), les milieux sociaux divers en fonction du terrain étudié et les partenaires extérieurs impliqués dans
l'opération.
Des déplacements sur le territoire national et à l'étranger sont nécessaires et les horaires de travail peuvent varier en fonction des partenaires
extérieurs.
Adresse WEB de l'unité : http://www.paris.inra.fr/aliss
Activités :
Concevoir un dispositif de recherche en réponse à un questionnement d'une équipe de recherche ou d'acteurs sociaux et d'institutions
externes
Connaissance approfondie des techniques et méthodes en SHS
Connaissance générale des champs de recherche complémentaires au champ étudié
Connaissance générale des outils informatisés de traitement de données
Connaître les principes éthiques et réglementation relative aux enquêtes
Connaître la communauté scientifique et les réseaux professionnels
Connaître les circuits de diffusion scientifique et technique du domaine de spécialité
Connaître les dispositifs nationaux de financement de la recherche et des règles de soumissionnement
Elaborer un dispositif de recherche en réponse à un appel d'offre ou à une demande spécifique
Elaborer des conventions de partenariats ou formaliser des contrats
Mobiliser et intégrer les apports de plusieurs disciplines pour traiter une question
Mettre en oeuvre une démarche qualité
Encadrer une équipe, fixer des objectifs et procéder à des évaluations
Utiliser les supports et registres de communication adaptés à différents publics (étudiants, stagiaires, institutions, entreprises)
Maîtrise de l'anglais oral et écrit
Elaborer un cahier des charges pour choisir des prestataires
Compétences :
Connaissance approfondie en sciences humaines et sociales et de la spécialité de l'unité de recherche
Coordonner les moyens humains, techniques et financiers nécessaires à la réalisation du projet de recherche, piloter des études
Organiser et animer des groupes de partenaires de la recherche
Valoriser les réalisations et les résultats sous forme de publications, de présentations orales auprès des communautés scientifiques
concernées et des partenaires extérieurs
Développer les outils de diffusion des savoirs et méthodes dans un objectif d'appropriation par différents partenaires (entreprises, collectivités
territoriales, administration)
Concevoir et réaliser une veille scientifique et méthodologique dans son domaine de spécialité
Organiser des manifestations scientifiques, nationales ou internationales. Assurer le secrétariat de rédaction d'une revue scientifique
http://www.inra.fr/drh/ce2011/bdd/profil.php?numprofil=IR11-SAE2-1
09/02/2011
INRA-DRH | numconcours externes 2011
Page 2 of 2
Gérer un collectif de travail
Actualiser ses connaissances disciplinaires et thématiques
Capacités personnelles :
Doctorat, Ingénieur gde Ecoles ou assimil.
Rigueur, Capacités relationnelles et d'adaptation sont indispensables pour ce poste.
2011 | conception et réalisation : DRH - DSI
http://www.inra.fr/drh/ce2011/bdd/profil.php?numprofil=IR11-SAE2-1
09/02/2011
DRH - SRM
C O N C O U R S
E X T E R N E S
I T
S E S S I O N
2 0 1 1
CONCOURS N° IRD01
INGENIEUR DE RECHERCHES
Ingénieur de Recherche en production et analyse de
données
EPREUVE ECRITE D’ADMISSION : DOSSIER TECHNIQUE
Date : 07/06/2011
Durée : 4 heures
Coefficient : 3
______________________________________________________________________________
Consignes générales : A lire attentivement



L’épreuve étant anonyme, aucun signe distinctif ne devra être porté sur les copies d’examen
(signature, inscription de votre nom et votre prénom ailleurs que dans le cadre réservé à cet
effet, etc…).
Veuillez vérifier que votre questionnaire est complet.
L’usage de la calculatrice est autorisé.
Composition de l’épreuve :
Question 1 ……………………………………………………….………………………… 30 points
Question 2 ……………………………………………………….………………………… 10 points
Question 3 ……………………………………………………….………………………… 10 points
Question 4 ……………………………………………………….………………………… 10 points
TOTAL……………... 60 points
EPREUVE ECRITE D’ADMISSION : DOSSIER TECHNIQUE
Question 1 (30 points)
Résumez en deux pages maximum le texte ci-joint « Social Norms of Body Shape and
Nutritional Policies » en faisant ressortir les idées dominantes.
Question 2 (10 points)
Que peut-on retenir des résultats présentés dans l’article pour les politiques publiques de santé ?
Question 3 (10 points)
Qu’elles sont les différentes disciplines pouvant être concernées par la thématique de recherche de
l’article. En quoi sont-elles complémentaires ? (10 points)
Question 4 – Questions techniques (10 points)
4.1 - Qui sont les principaux fournisseurs de données des thématiques de recherche portant
sur l’alimentation humaine ? (2 point)
4.2 - Définissez les termes suivants, indiquez en les avantages et limites : Données
d’enquête ; données de panel ; pseudo-panel. (3 points)
4.3 - Définissez la notion d’élasticité en économie. (5 points)
1/6
N° 1 –February 2008
RESEARCH IN ECONOMICS AND RURAL SOCIOLOGY
Social Norms of Body Shape and Nutritional Policies
The implementation of a nutritional policy by the recent “Plan National Nutrition Santé” (PNNS - National Nutrition and
Health Programme) raises the question of the relative effectiveness of policy tools, such as price interventions and information
campaigns. However, social interactions between consumers are likely to complicate policy evaluation. More precisely,
suppose that a public policy has a direct effect on consumption behaviours and that, by aggregation, individual behavioural
changes alter certain social norms. These norms may in turn influence individual behaviours. The overall effect of the public
policy is the sum of a direct and immediate effect on individuals, and an indirect effect through social norms. With this in
mind, we here present the results of a quantitative study, which identifies the impact of social norms of body shape on
individual representations of ideal body and the effect of the latter on nutritional attitudes. Hence, the research asks whether
collective representations of the ideal body are a parameter that must be taken into account in the economic analysis of
nutritional health policies.
Introduction
The reduction of overweight and obesity in adults and
children is one of the nine targets of the PNNS. An adult is
overweight if her/his body mass index (BMI: weight in
kilograms divided by height squared) is over 25 and obese
if his BMI is over 30 (see frame 1). According to this
indicator, between 1990 and 2002 the prevalence of
overweight in the French adult population went from
29.7% to 37.5% (OECD Health Data, 2005). This physical
condition is a risk factor for many chronic illnesses
(diabetes, mobility disorders, heart diseases). As a
consequence, the medical overcosts of obesity are
growing, and will be even more important in the future as
young generations are more concerned than the older ones.
To illustrate this point, in 1990 the additional medical cost
generated by an obese individual was between 150 and
350 euros. Since, in France, all individuals are covered by
the Social Security, whatever their efforts to stay fit, there
is a problem of “ex-ante moral hasard”. A solution would
be to modulate social security contributions according to
the efforts of prevention observed among the insured. As
the Social Security’s founding principles forbid this kind
of discriminatory measure, other policies are proposed or
implemented: disseminating targeted or generic
information, food price interventions, banning or
regulating sales of certain food products, and so on.
Assessing their expected benefits requires, among other
things, that social interactions be considered, as they may
multiply or nullify their effects. In this perspective, we
here present some research results on the ways social
norms of body shape are formed and their link with
consumers’ nutritional attitudes (Etilé, 2007).
Social interactions and social norms in economics
For economists, social interactions between rational
individuals belong to one or another of these three
categories: constraint interactions (seen in competition for
a limited resource, for instance); information interactions
(cf. the informational cascades, that may explain herd
behaviours on financial markets); and preference
interactions, wherein individual preferences depend on the
distribution of actions and/or preferences within a
reference social group (Manski, 2000). Regarding
preference interactions, theoretical models usually assume
that the deviation between the action selected by the agent
(here, his BMI) and the average decision of members of
the reference group (the mean BMI in the group) affects
her/his well-being. If well-being decreases with the
deviation from the average, then the agent has a preference
for conformism (or an aversion for stigmatisation) which
can be interpreted as a social norm effect. Conversely, if
well-being increases with the deviation, then the agent has
a preference for distinction (Akerlof, 1997).
However, this approach to norms and distinction raises
two problems:
•
•
It gives no indication on the choice of the
reference group. Many quantitative studies mimic
the sampling design of the survey they use to
form ad hoc reference groups (for instance, the
school in a survey on teenagers snacking). But as
recommended by Manski (1993), it is more
appropriate to define these reference groups with
information obtained from qualitative, socioethnographic, studies.
Although reducing a social norm to a behavioural
regularity eases the construction of an empirical
measure (means or medians can be easily
computed from well-defined variables), the
2/6
system of values and collective expectations,
which underlie social norms, are forgotten. For
example, the norm for murder is “thou shall not
kill” rather than the prevalence of manslaughter in
the population.
Against this reductionism, Akerlof and Kranton (2000)
suggest an alternative approach to social norms, that they
call prescriptions, which is grounded on two assumptions.
The social reference group is the group to which the
individual is exogenously assigned (because of his age,
gender and his financial or human capital). Second,
prescriptions are shared expectations on how the group’s
members must behave or on the attributes they must have
(here, a specific body shape). This is the approach we
intended to implement.
Social norms and individual representations of the
ideal body
The research is based on the French Institute of Statistics’
survey on households’ living conditions “Enquête sur les
conditions de vie des ménages” (INSEE, 2001), which
collects information on actual and ideal body weights,
actual height and a number of nutritional attitudes for more
than 5,000 individuals (see frame 2). It is then possible to
construct measures of actual and ideal BMIs, which are
supposed to summarise respectively the individual’s actual
and ideal body shapes.
Several works in social psychology have showed that,
from a phenomenological point of view, it is difficult to
distinguish the attributes that the individuals would like to
possess (the ideal attributes) from those they think they
ought to have in order to conform to the expectations
produced within their assigned social category. For
instance, following Bourdieu (1979), working class
members are expected to be strong and, in some sense,
“large”, while executives and white collars ought to be
slim and smart. As a consequence, each individual’s ideal
BMI must contain information about the body shape
prescriptions prevailing in her/his social group. One way
to extract this information is to compute the average ideal
BMI in this group. This yields a measure of the social
norm of body shape. Yet, the definition of the social group
of assignation has to be precised. We initially relied on
Bourdieu (1979)’s study of the role played by body shape
in the distinction between socio-professional categories
(SPC), and on a sociological work, which has shown that,
in France, ideal BMI is lower for executives and
intermediate professions than for workers and employees
(Régnier, 2006). The assignation group was then initially
defined by crossing age (individuals with a maximum 5
year difference in age), sex and the socio-professional
category (SPC in 12 categories). Including social
stratification in the definition of the social group modified
the results slightly and weakened their robustness in a
significant way. The results presented here, the most
robust ones we have, are therefore based on assignation
groups only defined from the age and sex criteria.
Nearly 40% of individuals consider that their ideal weight
is their actual weight, which is interpreted by assuming
that individual representations of ideal body shape are
produced, on the one hand, by social norms and, on the
other hand, by the application of a principle of reality - “to
be satisfied with what we have”-, which is a cognitive
strategy that helps to cope with life and preserve wellbeing. Finally, the model underlying the empirical analysis
supposes that social norms and habitual body shape jointly
determine individual representations of ideal body shape
(ideal BMI), which in turn affect food behaviours. Actual
body shape (that is to say BMI) at the moment of the
survey is an approximate measure of habitual body shape.
Figure 1 summarises the model, while Frame 3 presents
the estimation method
Table 1 reports the main results, by gender, for those men
and women who want to lose weight (ideal BMI lower
than actual BMI). It is found that for women, a 1%
increase in the social norm (the average ideal BMI in their
assignation group) induces a rise of nearly 0.5% in ideal
BMI. For men, a similar variation of social norm would
have no effect on ideal BMI. This gender difference is
consistent with results of the international literature on the
impact of beauty stereotypes, which is much mode
important on women. A 1% variation in habitual BMI
produces a 0.5% rise in ideal BMI of women and a 0.8%
rise for men.
For instance, a woman whose social norm prescribes her a
weight of 60 kg would declare an ideal weight 450g higher
when the norm goes up to 61 kg. In the same way, a price
policy, which would reduce usual BMI by 1% would
directly lowers the ideal BMI of those women who want to
slim by a minimum of 0.49%. As all group members
simultaneously modify their representations, the average
ideal BMI of the group - the social norm - would also
move downward. In fine, a second -0.54% drop in each
individual’s ideal BMI, indirectly caused by a modification
of the norm, would reinforce the (-0.49%) direct effect.
This second effect would not be observed in men.
Weight, ideal weight and nutritional attitudes
We can now turn to the predictive value of the variables
we use, and ask whether individual food behaviours are
affected by the deviation of their usual BMI from their
ideal BMI. Although the survey does not give precise
information on these behaviours, it informs on some
nutritional attitudes that measure somehow the healthiness
of the individual’s diet. Then, we can estimate the
correlations (everything else equal) between these
attittudes and the deviation between actual and ideal BMIs.
Table 2 gives the sign of the coefficients that are
significant at the 5% level.
The more weight an individual would like to lose, (BMIhBMI* increases), the less balanced he estimates that his
food diet is, and the more likely he will eat low-fat and
low-sugar products and will avoid excessively rich food.
Conversely, individuals who would like to put on weight
(BMI* -BMIh increases) declare less often that they pay
attention to fat and sugar. Furthermore, in these estimates,
social norms have little effect on nutritional attitudes (just
one significant coefficient out of four). In other words, the
impact of social norms is only indirect, and works through
a modification of individual representations of ideal body
shape.
3/6
In the end, while for women the effects of social norms on
individual representations of ideal body shape are
significant, this is not the case for men. This gender
difference could partly explain why female obesity has
increased more quickly than male obesity (+64% between
1997 and 2006 against 40% for men): for women, the
impact of environmental changes (fall in prices,
advertisement pressure, and so on) is all the stronger that
they induce modifications of social norms, which amplify
their effects. This analysis sheds light on the effect of
global norms specific to large social groups. It could be
extended to analysis of the impact of norms produced by
media or norms produced more locally at the
neighbourhood scale or school scale.
Fabrice Etilé, INRA, UR 1303 Alimentation et Sciences Sociales ALISS, Ivry France
[email protected]
For further information
Etilé, F. (2007). Social Norms, Ideal Body Weight and Food Attitudes, Health Economics, 16, 945-966.
Régnier, F. (2006). Obésité, corpulence et souci de minceur: inégalités sociales en France et aux Etats-Unis, Cahiers de
Nutrition et Diététique, 41, 97-103, 383-413.
Akerlof, G.A. (1997). Social Distance and Social Decisions, Econometrica, 65, 1005-1027.
Akerlof, G.A. and Kranton, R.E. (2000). Economics and Identity, Quarterly Journal of Economics, 115, 715-754.
Bourdieu, P. (1979). La distinction, Paris : Editions de Minuit.
Manski, C.F. (1993). Identification of Endogenous Social Effects: The Reflection Problem, Review of Economic Studies, 60,
531-542.
Manski, C.F. (2000). Economic Analysis of Social Interactions, Journal of Economic Perspectives, 14, 115-136.
____________________________________________________________________________________________________
Published by the INRA Social Sciences, Agriculture and Food, Space and Environment
Publishing unit: Editorial Director: Bertrand Schmitt– Editor: Didier Aubert (Chief Editor),
Translation and composition: Ariel Gille
Parts of reproduction may be used only with mention of origin
Copyright: 1st term 2008 - ISSN 1778-4379
4/6
Figure 1: The model structure
Individual representation of ideal
corpulence: BMI*
Social norm: E{BMI*|g}
Nutritional attitudes
Usual corpulence: BMIh
Note: the red continuous arrows represent the relationships that the research intends to identify. A bijective arrow indicates a simple correlation, while a single
arrow indicates a potentially causal relationship. BMI* is ideal BMI. BMIh is habitual BMI, approximated in the estimation by actual BMI. E{BMI*|g} is the
mean ideal BMI in the social group g.
Table 1: Main results
Sample
Women who wants to lose
weight (N=1341)
Dependent variable
Men who wants to lose
weight (N=767)
Logarithm of ideal BMI (log(BMI*))
Logarithm of the social norm (log(E{BMI*|g})
0.524*
-0.022
Logarithm of the usual BMI (log(BMIh))
0.490*
0.784*
Hansen test (p-value)
0.197
0.313
Stock et Yogo test: relative Bias IV versus LOS
5-10%
20-30%
Note: Coefficients are to be interpreted as elasticities, everything else equal (income, family situation, and living place). *=significant at the 1% level. The
Hansen statistic p-value shows that the choice of the instrumental variables (IV) is correct. For women, the relative bias in the estimates (difference between
estimated coefficients and their «true» values) is at most 10% of the bias that would be produced by the ordinary least squares (OLS). In other words,
according to this criterion, the instrumental method reduces the instrumentation bias by a factor 10 relatively to OLS (3 for the men estimates).
Table 2: Correlations between nutritional attitudes and representations of body shape
Dependant variable
Thinks has a
balanced food diet
Consumption of lowfat products
Consumption of
low-sugar products
Avoids excessively
sweet and fat
products
Satisfied or wants to gain weight
(BMI* -BMIh if BMI*≥ BMIh, 0
otherwise)
-
-
0
-
Wants to lose weight (BMIh- BMI* if
BMI*< BMIh, 0 otherwise)
-
+
+
+
Social norms (log(E{BMI*|g})
0
0
-
0
Usual BMI (log(BMIh))
0
+
+
0
Note: Correlations everything else equal (age, sex, income, SPC, family situation, living place).
Frame 1 - BMI and Body Shape
In a very simplistic way (but in this matter, data are scarce), we here summarise body shape by the body
mass index (BMI), obtained by dividing weight in kilograms by height squared. According to the World
Health Organization ( WHO) medical norms, an individual is overweight when her/his BMI is over 25
kg/m2, and obese if her/his BMI is over 30 kg/m2. BMI is a good predictor of overweight-related
morbidity, and health risks significantly increase when it goes beyond the threshold of 27 (and not 25).
However, the BMI does not take into account the distribution of weight in muscle and fat and cannot be a
good summary of some individuals’ body shape (in particular, or those who do “physical” jobs), even
though it was demonstrated that the correlation between BMI and the distribution of weight between fat
mass and muscular mass is quite high, whatever the age, sex, ethnic origin or physical activity level
.
Social Norms of Corpulence and Nutritional policies
Fabrice Etilé, INRA, UR 1303 Alimentation et Sciences Sociales ALISS, Ivry France
[email protected]
ISSN 1778-4379 N°1 - February 2008
5/6
Frame 2 - Survey on households’ living conditions 2001
In 2001, in addition to usual information on socio-demographic characteristics and living conditions, the
INSEE survey on households’ living conditions contained detailed information about the nutritional
attitudes and health behaviours and status of 5,198 individuals belonging to as many households
representative of the whole population. Individuals were asked in particular to declare their height and the
weight they “would like to have”. This last variable is interpreted as a measure of ideal weight and,
divided by height in meters squared, provides a measure of ideal body shape (ideal BMI).
Frame 3: The econometric analysis
The econometric analysis focusses on the estimation of the equation that describes the formation of
individual representations of ideal body shape. Let respectively BMI* and BMIh be ideal and habitual
BMI. The latter is proxied by actual BMI. E{BMI*|g} is the mean ideal BMI in the individual’s
assignation group g, in other words, our measure of the social norm. The equation to be estimated is as
follows:
BMI*=α E{BMI*|g}+β BMIh+δX+η
where X is a set of control variables and η an error term. Manski (1993) showed that estimating this
equation by ordinary least squares (OLS) does not identify α, because the correlation between BMI* and
E{BMI*|g} is the product of three effects: (1) a direct influence effect of social norms (direct causality);
(2) a reflection effect due to the fact that the BMI* individual representations of each group member g
and then, their average, depend on each individual’s ideal BMI (reverse causality); (3) a contextual
effect due to the fact that group members are influenced by similar information flows about the ideal body
shape (same medical information on what is a “healthy” weight, same standards of beauty proposed by
the media, and so on, or any other third variable observable or not). To identify the first effect, which is
the focus of the research, E{IMC*|g} is instrumented by a set of instrumental variables (
IV), whose validity is assessed by a Hansen-Sargan test (orthogonality of the instruments with the error
term η) and a Stock and Yogo test (which examines whether instruments correctly predict the
instrumented variable).
Last, the joint endogenous nature of social norms and habitual BMI renders impossible the identification
of a system of equations that would describe the formation of individual representations and the effect of
these representations on actual BMI (because instruments lack). This is the reason why, more modestly,
we just have a look on the relationships between individual representations and nutritional attitudes.
Social Norms of Corpulence and Nutritional policies
Fabrice Etilé, INRA, UR 1303 Alimentation et Sciences Sociales ALISS, Ivry France
[email protected]
ISSN 1778-4379 N°1 - February 2008
6/6
DRH - SRM
C O N C O U R S
E X T E R N E S
I T
S E S S I O N
2 0 1 1
CONCOURS N° IRD01
INGENIEUR DE RECHERCHES
Ingénieur de Recherche en production et analyse de
données
EPREUVE ORALE D’ADMISSION
Temps de préparation : 15 minutes
Durée : 35 minutes coefficient : 3
Question à préparer par les candidats :
Compte tenu de votre expérience, quel serait votre apport au poste auquel vous postulez.
Temps de l’exposé devant le jury : 5 minutes maximum (inclus dans les 35 minutes)