A dialogue of the deaf in the statistical theater? Adressing structural effects within a geometric data analysis framework


Since their introduction in the late 1960s, the “moderate”, and moreover “metrological” and “hypermetrological” uses of regression models quickly became the dominant quantitative approach in the Anglo-Saxon social sciences. This “sociology of the variables” has been the subject of many critical insights, with little impact on its dominance. By contrast, the French situation is quite different, mainly because of the strong association between Pierre Bourdieu’s research program and the correspondence analysis methods. In this context, the relationship between geometric data analysis and regression models has turned into a “dialogue of the deaf”. Complementarity is sometimes emphasized, correspondence analysis being associated with exploration and description of the data, and regressions being used to explain, reject or confirm assumptions. But regression models may also be used in order to analyze structural effects within a framework of geometrical data analysis, e.g. by visualizing graphically the results of a regression (Rouanet et al. in Math Sci Hum 160:13–46, 2002; Lebaron 2013). We propose a new multi-step approach, “Standardized Factor Analysis”, which relies on geometric analysis and uses linear regression in a second stage in order to uncover structural effects in the original space. We illustrate it with data about tastes for cinema in France. We conclude by raising a more general set of questions about causality: social determinisms, even well established, are partial in the sense that they produce their effects only when associated with each other.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4


  1. 1.

    It is technically possible to add interaction effects between explanatory variables in a regression model but besides that it is a relatively uncommon practice, the multiplication of combinations quickly becomes unworkable. An interesting approach is developed by Judea Pearl in the framework of analytical sociology, based on structural causality (Pearl 2009).

  2. 2.

    There is already criticism of this assumption in Halbwachs’ The causes of suicide (2002 [1930]), about the relationship between religion and urbanization.

  3. 3.

    This argument is also very present in the epistemological reflections of Passeron (1991).

  4. 4.

    We cite here some of the main criticisms, leaving aside—among others—those of Abbott about temporality issues (2001).

  5. 5.

    One may ask if we would not understand better some scientific debates by analyzing these kinds of competitions on the "size" and "robustness" of effects as a form of male competition.

  6. 6.

    This is the same logic that explains the recent success, within state organizations or NGOs, of the "randomized evaluation" promoted by Esther Duflo (2010) to test the effectiveness of development assistance programs.

  7. 7.

    «This method [the elimination of structural effects] leads to the study and the comparison between the behaviors of a reindeer in the Sahara and of a camel at the North Pole» (cited in Maurice Halbwachs, «La statistique en sociologie», in Centre internationale de synthèse, La statistique. Ses applications. Les problèmes qu’elles soulèvent, Paris, PUF, pp. 113–160).

  8. 8.

    For a study of tastes for cinema in France, see Duval (2011).

  9. 9.

    We use “eta²” indicator here, which is similar to R² in linear regression analysis (Le Roux and Rouanet 2004).

  10. 10.

    In demography, one of the most widely used indicators to study mortality is the crude death rate. When comparing Sweden and Mexico, for example, we see that the crude death rate in Sweden is higher than that of Mexico, which may seem counterintuitive in that health conditions are better and life expectancy is higher in Sweden. This is due to the fact that the crude death rate is influenced by the age structure of the population under study: Sweden has an older population than Mexico, and the elderly have higher death rates than other categories of age, which leads to a relatively high crude death rate. Demographers thus use statistical techniques of "standardization" to neutralize the effect of the age structure, i.e. to calculate the mortality indicators that are net of this structural effect. More generally, SFA is applied to MCA in this article but it could apply to any other factor analysis technique, such as principal component analysis (PCA).

  11. 11.

    At this stage, one can choose to retain only the first principal components, those that contain the most information.

  12. 12.

    To the extent that the PCA is not performed directly on the variables of the original MCA, one can’t obtain the contributions of these variables. However, all the other tools for the interpretation of supplementary variables may be used (cos², v test, eta²).


  1. Abbott, A.: Time Matters. On Theory and Method. The University of Chicago Press, Chicago (2001)

    Google Scholar 

  2. Bourdieu, P.: Les Structures Sociales de L’économie. Seuil, Paris (2000)

    Google Scholar 

  3. Bourdieu, P., Darbel, A.: La fin d’un malthusianisme. In: Darras (eds) Le Partage des Bénéfices, pp.135–154. Minuit, Paris (1966)

  4. Bourdieu, P., de Saint-Martin, M., .: Anatomie du goût. Actes Recherche Sci. Soc. 2(5), 2–81 (1976)

    Google Scholar 

  5. Deauvieau, J.: Est-il possible et souhaitable de traduire sous forme de probabilités un coefficient logit ? Réponse aux remarques formulées par Marion Selz à propos de mon article paru dans le BMS en 2010. Bull. Méthodol. Sociol. 112(1), 32–42 (2011)

    Article  Google Scholar 

  6. Des Nétumières, F.: Méthodes de régression et analyse factorielle. Hist. Mesure 12(3–4), 272–297 (1997)

    Google Scholar 

  7. Desrosières, A.: Entre réalisme métrologique et conventions d’équivalence : les ambiguïtés de la sociologie quantitative. Genèses 43, 112–127 (2001)

    Google Scholar 

  8. Desrosières, A: Bourdieu et les statisticiens. Une rencontre improbable et ses deux héritages. In: P. Encrevé, P., Lagrave, R.M. (eds) Travailler avec Bourdieu. Flammarion, Paris (2003)

  9. Donnat, O.: Les pratiques culturelles des Français à l’ère numérique. Enquête 2008. La Découverte, Paris (2009)

  10. Duflot, E.: Le Développement humain. Lutter contre la pauvreté (I). Le Seuil/République des idées, Paris (2010)

  11. Duval, J.: L’offre et les goûts cinématographiques en France. Sociologie 2(1), 1–18 (2011)

    Article  Google Scholar 

  12. Halbwachs, M.: Les causes du suicide. PUF, Paris (2002 [1930])

  13. Le Roux, B., Rouanet, H.: Geometric Data Analysis: From Correspondence Analysis to Structured Data Analysis. Springer, New York (2004)

    Google Scholar 

  14. Lebaron, F.: L’analyse géométrique des données dans un programme de recherche sociologique : le cas de la sociologie de Bourdieu. Modulad 42, 102–109 (2010)

    Google Scholar 

  15. Lebaron, F.: La régression peut-elle faire progresser? Approches et usages critiques des modèles de régression. Working Paper for: Sociologie quantitative et sociologie de la quantification. PRINTEMPS, University of Versailles-Saint-Quentin-en-Yvelines (22 March 2013)

  16. Lebart, L., Morineau, A., Piron, M.: Statistique Exploratoire Multidimensionnelle. Dunod, Paris (2000)

    Google Scholar 

  17. Léridon, H., Toulemon, L.: Démographie.Approche Statistique et Dynamique des Populations. Economica, Paris (1997)

    Google Scholar 

  18. Lieberson, S.: Making It Count. The Improvement of Social Research and Theory. University of California Press, Berkeley (1985)

    Google Scholar 

  19. Ministère de la Culture et de la Communication: Les résultats complets de l’enquête 2008. http://www.pratiquesculturelles.culture.gouv.fr/08resultat.php (2008). Accessed 20 Sept 2014

  20. Ollion, E.: De la sociologie en Amérique. Éléments pour une sociologie de la sociologie étasunienne contemporaine. Sociologie 2(3), 277–294 (2011)

    Article  Google Scholar 

  21. Passeron, J.-C.: Le raisonnement sociologique, L’espace non-poppérien du raisonnement naturel. Nathan, Paris (1991)

    Google Scholar 

  22. Pearl, J.: Causal inferences in statistics: an overview. Stat. Surv. 3, 96–146 (2009)

    Article  Google Scholar 

  23. Pénissat, E.: Quantifier l’effet “pur” de l’action publique: entre luttes scientifiques et redéfinition des politiques d’emploi en France. Sociol. Soc. 43(2), 223–247 (2011)

    Article  Google Scholar 

  24. Ragin, C.: The Comparative Method: Moving Beyond Qualitative and Quantitative Strategies. University of California Press, Berkeley (1987)

    Google Scholar 

  25. Rouanet, H., Ackermann, W., Le Roux, B.: The geometric analysis of questionnaires: the lesson of Bourdieu’s La distinction. Bull. Méthodol. Sociol. 65(1), 5–18 (2000)

    Article  Google Scholar 

  26. Rouanet H., Lebaron F.: La preuve statistique: regard critique sur la régression. Working Paper presented at: Qu’est-ce que Faire preuve?, University of Amiens (2006)

  27. Rouanet, H., Lebaron, F., Hay, V.L., Ackermann, W., Le Roux, B.: Régression et analyse géométrique des données: réflexions et suggestions. Math. Sci. Hum. 160, 13–46 (2002)

    Google Scholar 

  28. Vallet, L.-A.: A propos d’un ouvrage peu connu dans la sociologie française: making it count. The improvement of social research and theory, de Stanley Lieberson. Revue Eur. Sci. Soc, XLII(129), 341–348 (2004)

    Google Scholar 

  29. Vallet, L.-A., Caille, J.-P.: Les carrières scolaires au collège des élèves étrangers ou issus de l’immigration. Éduc. Form. 40, 5–14 (1995)

    Google Scholar 

Download references


The authors gratefully acknowledge Gabriel Abend, Jérôme Deauvieau and Frédéric Lebaron as well as the anonymous reviewers for their proofreading, discussions and advice.

Author information



Corresponding author

Correspondence to Olivier Roueff.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Bry, X., Robette, N. & Roueff, O. A dialogue of the deaf in the statistical theater? Adressing structural effects within a geometric data analysis framework. Qual Quant 50, 1009–1020 (2016). https://doi.org/10.1007/s11135-015-0187-z

Download citation


  • Correspondence analysis
  • Regression models
  • Analysis of structural effects
  • Combination of geometrical and inferential statistics
  • Standardized factor analysis (SFA)