Helsinki, 2-6 November 2009

Dear All,

Here you can continue the discussions you started at the workshop 2009. Just add a “comment”.



Lectures on Social representations

Here you can watch videos of the lectures on social representations presented during the Helsinki workshop (by clicking the links below – or if you are viewing this post in the categories section, first click the title of this post in order to get at the links).

You can find instructions for watching the videos from the web pages of the Department of Social Psychology (Helsinki). You might find the instructions useful, even though part of the instructions might be relevant only to the University of Helsinki students and staff.

Paula Castro

Klaus Helkama

Elisabeth Lage

Anna-Maija Pirttilä-Backman

Hannu Räty

Christian Staerklé

Toshio Sugiman

Wolfgang Wagner

How to discuss in this blog

Here we can continue the discussions started in our workshop.

If you want to comment something that is already in the blog, click the “Comments” on the left and write your comment.

In order to write comments this blog system requires that you type in your name and e-mail address. Your name will appear in context of your comment (so you might want to consider which name you want to type in, e.g. your whole name / only first name / first name and initial letter of the last name etc.). Your e-mail address will never appear on the site.

You can write your comment in the “Discussions” category by clicking “Comments” on the left of the post. You can also write your comment on whichever post on Information board and in other categories.

You can see all the comments and follow the discussions by clicking “All comments”.

If you want to post new information to the information board or start a new discussion, please send your contribution to Antero (antero.olakivi[at], who will put it on the blog. (Sorry that you cannot submit new entries by yourself – dictates of the programme.)

You can see all posts by clicking “All posts”.

For additional information about this blog, please see “About this blog” on the top of the front page.

More interesting information related to power and trust

Pierre Rosanvallon, La contre-démocratie, Paris, Seuil, 2006, p. 18.

Material related to prof. Lage’s presentation

Here you can find Elisabeth Lage’s powerpoint presentation in pdf format:


And other material related to Elisabeth Lage’s presentation:




(if you are viewing this post in the categories section, first click the title of this post in order to get at the pdfs)

More examples of how to analyze word associations

Hovardas, T. & Stamou, G. P. (2006). Structural and Narrative Reconstruction of Representations of ‘‘Environment,’’ ‘‘Nature,’’ and ‘‘Ecotourism’’. Society and Natural Resources, 19, 225–237.

Hovardas, T. & Korfiatis, K. J. (2006). Word associations as a tool for assessing conceptual change in science education. Learning and Instruction 16, 416-432.

It would be great to learn more about how to analyze word associations. Jaana-Piia

Dear Jaana-Pia,


so here I would like to give you some hints concerning word associations

in SR research. In general, my comments are very much based on the

“Genevan” approach to SR, initiated by Doise, the main reference of which

is Doise, Clémence, Lorenzi-Cioldi (1993). The quantitative analysis of SR.



Although the book contains detailed descriptions of many multivariate

techniques (Cluster, Factor, Correspondence, Discriminant in particular)

and their relationship with SR theory, it is also a bit outdated as it

lacks discussions about both experimental and more recent techniques

(structural equations, multilevel). Anna-Maija should have it.


In general, as you can imagine, I would rather recommend quantitative

analyses on word associations as such, that is, words produced without

sentences and hence arguments. I don’t really see how you can do

qualitative analyses on words without context and rhetorics / arguments.

Maybe in terms of meaningful category construction (groupings of

words), but even that seems a bit flimsy. I think the fact that word

associations allow statistical treatment is precisely their advantage

over qualitative, argumenative analyses. quantitative analyses open up

the possibility for group comparisons and individual difference

analyses, both of which are difficult to do properly in a purely

qualitative approach. The main problem with word associations, however,

is the difficulty to accurately assess the meaning of the produced

words, often ambiguous, especially for general and abstract terms which

are precisely those which often come to mind most easily (e.g., liberty,

equality, justice, intelligence, beauty, …). Without qualifiers, it is

hard to know what people put behind such concepts, but we need to live

with that.


As you know, the main quantitative analysis on word associations,

except, of course, simple descriptive statistics of mean occurrences, is

correspondence analysis (available also in SPSS). It gives you the

structure of the representational space along with its organising

principles (e.g., dimensions 1 & 2). Groups can easily be integrated in

this analysis, by using them as supplementary variables (in which case

they do not contribute to the structuring of the factorial space, and

are only “projected” on it), or as active variables (in case you

hypothesise that group positions are co-determinant of the factorial

space). Furthermore, you can run the analysis on individual or on

aggregate (group-level) answers, depending mostly on whether you are

more interested in individual or in group level variation.


But the group comparison thing is tricky. The question is the

theoretical significance of group comparisons: what are they supposed to

tell you other than “gee, that’s interesting”. For one, you can put

forward hypothesis tests of group differences, that is, you expect two

groups to differ on a given dimension, based on theoretical assumptions,

and you test this prediction with an appropriate statistical test,

anova, log-linear, and the like. With natural groups, and in particular

with national groups, this is often difficult, because the groups differ

on 1001 different dimensions, and you cannot know which one is

responsible for your observed group differences. If you don’t have a

theory on the groups and their differences, you cannot “compare” them

stricto sensu. And one should probably be careful with the term

“comparison” anyway, since it would imply formal comparison on

predefined and measurable dimensions.


The second option is more exploratory and illustrative; you use

different groups to differentiate the sources of representations and

thus to maximise their variation. In this case, it may be appropriate to

include other groups as well in your analysis. If, for example, your

data come from Italy, Denmark and Finland, then you could also use other

groups differentating your sample, for example, gender, age, social

class, etc.,. These groups then allow you to evidence social anchoring

patterns. This can easily be done for example in correspondence analyses

(which obviously does not test hypotheses).


Another technique for studying group differences is discriminant

analysis, which is basically a mixture of Manova and principal

components analysis. It requires scale data and gives you the factorial

dimensionS on which predefined groups differ. In a way it is the most

sophisticated group difference analysis, but word association data would

need to be converted into scales (for example with correspondence



Finally, you can use cluster techniques to create groups on the basis of

your data (numerical or textual), that is, similar individuals will be

grouped together which basically gives you a typology. You can then

compare this typology with predefined groups, e.g., national groups.


So these are some preliminary indications on words and groups. An

example of a paper using some of these techniques is


Doise, W., Staerklé, C. & Clémence, A., & Savory, F. (1998). Human

Rights and Genevan Youth : A developmental study of social

representations. Swiss Journal of Psychology, 57, 86 – 100.


Unfortunately, I only have the manuscript version of this paper

(attached). You might also want to look at the text by Lorenzi-Cioldi

and Clémence which was suggested for the workshop. In case you read

French, I’ve also attached a chapter on experimental group differences

in a SR perspective. These should give you an idea.


good luck with your work,



Pictures from the Helsinki Workshop


Anna Huotilainen’s PhD thesis

Dear All,

as discussed in the Helsinki Workshop, here’s the link to
the summary of Anna Huotilainen’s PhD thesis

(if you are viewing this post in the categories section, first click the title of this post in order to get at the link)

The format is so called article thesis, meaning that in this case the thesis consists of five articles (the articles are now all published) and the summary.
The articles can be found in the electronic databases, they are not reproduced here.

In the workshop it was mentioned that in her summary Anna refers to (with adequate references) a few well established ways to use the process terms of the SR theory.
Best regards,