Facet analysis: a technique to combine structure and content in the study of social representations'

1
Facet analysis a technique to combine structure
and content in the study of social
representations.

• Fabrice Buschini
• University of Geneva

2
Outline

• Introduction content and processes in the
  different approaches in the study of SRs.
• What is facet analysis
• basic principle
• facets and mapping sentence
• The Smallest Space Analysis
• principle
• distinctive features
• basic structures
• Interpretation of SSA
• Relevance of facet analysis for the study of SRs
• checking or re-evaluation existing models
• locating position of different groups

3
Processes and content

• This distinction is very common and heuristic in
  science.
• Focusing on the processes, one hopes to
  understand the principles of phenomena in a kind
  of universalistic dream.
• On the contrary, the content in the flesh, which
  makes the phenomenon alive, particular and
  inserted in social world
• If the focus is maintained only on one side, it
  can be prejudicial to the comprehension of the
  phenomenon
• it is why, from the beginning, the SR theory
  insists on the necessity to investigate both
  content and processes in the study of SRs

4
Three Schools in the study of SRs

• Following the seminal work of Serge Moscovici on
  the SR of psychoanalysis (1961), three schools
  were developed for the study of SR
• School of Paris
• anthropological approach mainly qualitative based
  on observation, interviews and questionnaires
• School of Aix (central core theory)
• structural approach mainly quantitative based on
  interviews, questionnaires and above all words
  associations
• School of Geneva (organizing principles theory)
• mainly quantitative, it aims to study the common
  principles organizing a SR and the individual
  position relating to those principles
• In theory, those school attach the same
  importance to the processes and the content. In
  practice, however, one of those two elements is
  more or less favoured
• Facet analysis could help to find a better
  balance between processes and content.

5
Facet analysis Basic principle

• Resulting from the work of Louis Guttman (1944,
  1950, 1977, 1994), facet analysis is a technique
  based on a method (the smallest space analysis).
• It aims to control the adequacy between the
  theoretical level and the empirical level in a
  research.
• The premise is that a good theory or a good model
  should be reflected in the empirical data.
• The theoretical structure build by the researcher
  should be found in the structure of the data
• The graphical representation of data should fit
  the structure of the model hypothesized by the
  researcher.

6
Defining the facets

• The first step is to identify the relevant
  concepts or dimensions involved in the model
• they will constitute the facets of the model
• the elements of each facet should be then listed
• The second step consists in combining those
  facets and their elements in a mapping sentence
  that
• allows to check facets are independent and not
  overlapped
• gives a convenient tool in the research planning
  and the elaboration of the tools for data
  collection.

7
Example of a mapping sentence

• From a research program by Lorenzi-Cioldi and
  Buschini (2005) on affirmative action
• representation of personality traits that male
  and female executives can possess.

8
The smallest space analysis (SSA)

• It is a statistical tool that provides a
  graphical representation of data matrix in a
  space as small as possible (Guttman, 1968, 1982).
• It belongs to the Multidimensional scaling (MDS)
  family
• but its specificity comes from the coefficient
  used to measure the relations between variables
  (elements of the representation)
• The Guttmans coefficient of monotonicity
• Contrary to the factor analysis, all the variance
  of the data is represented in the space.
• Of course, the reduction of the dimensions
  introduces a distortion
• coefficient of alienation (stress in MDS) should
  be lesser than 0.2

9
A monotonic relation

• It is a relation in which the variables go on the
  whole in the same way, but not necessary in a
  linear way.

10
Basic structures that can be found in SSA

• Two prototypical forms can be identified in a
  SSA
• they can combine together to generate other forms
  of structure.
• The first form, called simplex, corresponds to
  variables that can be ranked according to their
  proximity on a continuum
• The more the variables are close, the more they
  are linked.
• The two extreme variables on the continuum are
  the less linked
• This structure can be represented as a line,
  which is perpendicularly shared by the facet
  elements

11
Representation of a Simplex
12
Basic structures that can be found in SSA (2)

• The second form appears with variables that can
  also be ranked according to their proximity.
• However, the proximity is not on a continuum,
• unless if a continuum without end is considered
• it is a step-by-step proximity.
• The structure takes a circular form along which
  variables are linked step by step.
• The facet elements can then be delimited by
  circles

13
Representation of a Circumplex
14
Basic structures that can be found in SSA (3)

• Combination of two simplexes gives a duplex.
• Combination of a simplex and a circumplex
  produces a radex.
• If data are represented in a three-dimensional
  space,
• a duplex can be combined with a simplex to give a
  triplex,
• a radex can be combined with a simplex to produce
  a cylindrex.
• The interested reader can find more information
  on those basic structures in the very complete
  chapter by Brown (1985).

15
Duplex and Radex
16
Back to the example
17
Rules of interpretation

• It is essential not to interpret the SSA
  structure as a factorial structure.
• You cannot oppose can hide their feeling to
  sensitive
• It is the same nonsense than opposing Milano to
  Roma and saying that those two cities give a
  sense to a north / south direction.
• Here, the interpretation should be done in terms
  of areas.
• But this kind of interpretation is easier with if
  facets are projected on the SSA

18
Facet A gender of traits
19
Facet B valence of traits
20
Facet C Utility of traits
21
From two to three dimensions

• If we had a dimension to the SSA solution
• The coefficient of alienation goes from 0.21 to
  0.13
• The new solution will be better
• More complicated, but less distorted
• Facet A appears under a similar form in the plane
  made by dimensions 1 and 2.
• Facets B and C appear with a simpler structure in
  the plane made by dimensions 1 and 3

22
Facet A in the plane d1 x d2
23
Facet B and C in the plane d1 x d3
24
Interest for the study of SRs

• The reintroduction of the meaning, the symbolic
  into the structure
• To reunify content and processes
• It allows to check or evaluate some existing
  models
• Basic cognitive schemata (Guimelli and Rouquette,
  1992)
• Bidimensional model of SRs (Moliner, 1994, 1995)
• It allows to project different groups position
  on the whole structure
• By means of supplementary elements

25
List of BCS
26
Fictive structure of the bidimensional model
27
Conclusion

• Great interest of facet analysis for the study of
  SRs
• It takes into account both content and processes
• It respects the dialectical relation between
  semantic, symbolical, pragmatic aspects of the
  elements of the representation and their
  structure
• It allows to study both the consensual part of
  SRs and their group modulations
• Technically, it is not more complicated than
  other statistical software
• Facet analysis could be an important source of
  innovation
• If researchers are less conservative and timorous

28
References

• Abric, J.-C. Guimelli, C. (2002). La zone
  muette des représentations sociales approche
  théorique et méthodologique, 6th International
  Conference on Social Representations Thinking
  Societies Common Sense and Communication,
  University of Stirling, Scotland, 27th August -
  1st September 2002.
• Abric, J.-C. Vacherot G. (1976). Méthodologie
  et étude expérimentale des représentations
  sociales tâche, partenaire et comportement en
  situation de jeu. Bulletin de Psychologie, XXIX,
  735-746
• Borg, I. Shye, S. (1995). Facet theory Form
  and content, Thousand Oaks, Sage Publications.
• Brown, J. (1985). An introduction to the uses of
  facet theory, in D. Canter (ed.), Facet theory.
  Approaches to social research, New York,
  Springer-Verlag, p. 17-57.
• Canter, D. (ed.) (1985). Facet theory. Approaches
  to social research, New York, Springer-Verlag.

29
References (2)

• Clémence, A. (2003). Lanalyse des principes
  organisateurs des représentations sociales, in S.
  Moscovici F. Buschini (eds), Les méthodes des
  sciences humaines, Paris, PUF, p. 393-410.
• Deschamps, J.-C. Guimelli, C. (2001).
  L'organisation interne des représentations
  sociales de la sécurité/insécurité et l'hypothèse
  de la zone muette, 9ème table ronde
  Perspectives cognitives et conduites sociales,
  San Sebastian.
• Doise, W. (1986). Les représentations sociales
  définition dun concept, in W. Doise A.
  Palmonari (eds), Létude des représentations
  sociales, Neuchâtel, Delachaux et Niestlé, p.
  81-94.
• Doise, W. (2001). Droits de lhomme et force des
  idées, Paris, PUF.
• Flament, C. (1986). Lanalyse de similitude une
  technique pour les recherches sur les
  représentations sociales, in W. Doise A.
  Palmonari (eds), Létude des représentations
  sociales, Neuchâtel, Delachaux et Niestlé, p.
  139-156.

30
References (3)

• Flament, C. (1987). Pratiques et représentations
  sociales, in J. -L. Beauvois, R. -V. Joule J.
  -M. Monteil (eds), Perspectives cognitives et
  conduites sociales, 1, Théories implicites et
  conflits cognitifs, Cousset, Delval, p. 143-150.
• Flament, C. (1989). Structure et dynamique des
  représentations sociales, in D. Jodelet (ed.),
  Les représentations sociales, Paris, PUF, p.
  204-219.
• Giraud-Heraud, J. (1974). La notion de groupe
  analyse structurale d'une représentation sociale.
  Mémoire de maîtrise, Université de Provence.
• Guimelli, C. (1989). Pratiques nouvelles et
  transformation sans rupture dune représentation
  sociale la représentation de la chasse et de la
  nature, in J. -L. Beauvois, R. -V. Joule J. -M.
  Monteil (eds), Perspectives cognitives et
  conduites sociales, 2, Représentations et
  processus socio-cognitifs, Cousset, Delval.
• Guimelli, C. (1994) (ed.). Structures et
  transformations des représentations sociales,
  Neuchâtel, Delachaux et Niestlé.

31
References (4)

• Guimelli, C., Rouquette, M.-L. (1992).
  Contribution du modèle associatif des schèmes
  cognitifs de base à lanalyse structurale des
  représentations sociales, Bulletin de
  Psychologie, n Spécial Nouvelles voies en
  psychologie sociale , XIV, 405, 196-202
• Guttman, L. (1944). A basis for scaling
  qualitative data, American Sociological Review,
  9, 139-150.
• Guttman, L. (1950). The basis for scaloram
  analysis, in S. A. Stouffer, L. Guttman, E. A.
  Suchman, P. F. Lazarsfeld, S. A. Star J. A.
  Clausen, Measurement and prediction Studies in
  social psychology in World War II, vol. 4,
  Princeton, Princeton University Press, p. 60-90.
• Guttman, L. (1968). A general nonmetric technique
  for finding the smallest co-ordinate space for a
  configuration, Psychometrika, 33, 469-506.
• Guttman, L. (1977). What is not what in
  statistics, The Statistician, 26, 81-107.
• Guttman, L. (1982). Facet theory, smallest space
  analysis and factor analysis, Perceptual and
  Motor Skills, 454, 491-493.

32
References (5)

• Guttman, L. (1994). Louis Guttman on theory and
  methodology selected writings, edited by Shlomit
  Levy, Aldershot, Hants, England, Dartmouth
  Publishing Company.
• Haas, V. Jodelet, D. (1999). Pensée et mémoire
  sociales, in J. -P. Pétard (ed.), Psychologie
  sociale, Paris, Editions Bréal, p. 111-160.
• Jodelet, D. (1989). Folies et représentations
  sociales, Paris, PUF.
• Jodelet, D. (2003). Aperçus sur les méthodologies
  qualitatives, in S. Moscovici F. Buschini
  (eds), Les méthodes des sciences humaines, Paris,
  PUF, p. 139-162.
• Laurens, S. Roussiau, N. (eds) (2002). La
  mémoire sociale. Identités et représentations
  sociales. Rennes, Presses Universitaires de
  Rennes.
• Lorenzi-Cioldi, F. (2002). Les représentations
  des groupes dominants et dominés. Collections et
  agrégats. Grenoble, PUG.

33
References (6)

• Lorenzi-Cioldi, F. Buschini, F. (2002).
  Carrière au masculin et carrières au féminin, 6th
  International Conference on Social
  Representations Thinking Societies Common Sense
  and Communication, University of Stirling,
  Scotland, 27th August - 1st September 2002.
• Moliner, P. (1994). Les deux dimensions des
  représentations sociales, Revue Internationale de
  Psychologie Sociale, 7, 2, 73-86.
• Moliner, P. (1995). A two-dimensional model of
  social representations, European Journal of
  Social Psychology, 25, 1, 27-40.
• Moscovici, S. (1961). La psychanalyse, son image
  et son public. Paris, P.U.F. (nouvelle édition,
  entièrement refondue, P.U.F., 1976).
• Schwartz, S. H. (1992). Universals in the content
  and structure of values Theoretical advances and
  empirical tests in 20 countries, in M. P. Zanna
  (Ed.), Advances in Experimental Social
  Psychology, Vol. 25, New York, Academic Press,
  1-65.