USES OF SYMBOLIC RESOURCES IN YOUTH : MOVING FROM QUALITATIVE TO QUANTITATIVE APPROACH

Youth is a period of intense changes during which young people engage in various transitions resulting in theirs acquisition of a longer time perspective and a system of orientation, enabling to set priorities and values, and to guide their actions accordingly. In a socio-cultural theoretical background, both the establishment of values and the ability to think time require some psychological distancing from the here and now, distancing which is foremost enabled by semiotic mediation. In our former studies on youth transitions, we observed that young people may use songs, movies, arts, or novels as symbolic resources, that is, as external mediators that seem to support these developmental processes. Through an abductive process linking qualitative, ideographic data and theoretical elaboration, we proposed a theoretical 7 dimensional model for analyzing people’s uses of symbolic resources. This model was then turned to a first, provisional questionnaire aiming at testing the model, whose items were extracted from the first empirical investigation. In this paper, we attempted to tests this questionnaire on a population of young people in Serbia. The symbolic resources questionnaire was tested on a sample of young people (N=475). A SEM analysis was used to test the model. At large, the theoretical model is verified. However, an unexpected, very strong correlation between the dimensions had to be explained. We finally propose a further adaptation to the questionnaire.

Contemporary youth in western societies is living in increasingly complex and quickly changing cultural environments.If social structures change, some basic psychological needs remain.People need to find, to some extent, a feeling of continuity through time beyond constant changes, and of self integrity through a multiplicity of social insertions (Erikson, 1968).As young adults, they not only need to have some orientation toward the future that are anchored in the past and to construct a hierarchy of values and orientations, but also to confer some sense to their experience and find means to deal with uncertainty and ambiguity (Mørch, 2006;Zittoun, 2007).Hence, youth can be said as the period of emergence and stabilization of a system of orientation based on two dimensions: a hierarchical system of values, consistent with possibilities of action in given settings, and the construction of a time perspective (see also Brandtstädter & Lerner 1999).

Symbolic resources in youth transitions
One of the basic process standing at the core of the developmental task of young people is that of the construction of meaning.Adopting a socio-cultural perspective (Bruner, 1990;Valsiner & Rosa, 2007;Vygotsky, 1934), we will question the way in which young people can confer sense to their experience.Within this framework, one key phenomenon is process by which culture mediates thinking and action, or in other words, how the person comes to think and act with the mediation of the cultural means at her disposal.This leads us to examine how a person confers sense to her experience (and on this basis, how she shapes and canalizes action and thinking) thanks to the semiotic elements that culture provides her with.
We have called cultural elements distinctive artefacts, made out of various semiotic modes, and have especially studied these that enable imaginary experiences, such as movies, novels, paintings, music, and so on.On the basis of indepth analysis of a series of case studies, we have shown that people seem to support their processes of sense making through their relations to such elements.We have called such deliberate uses of cultural elements to support such changing processes, uses of symbolic resources (Zittoun, Duveen, Gillespie, Ivinson, & Psaltis, 2003;Zittoun, 2004Zittoun, , 2005)).Based on careful analyses of in-depth interviews and diary material (Zittoun 2006(Zittoun , 2007(Zittoun , 2008)), we have proposed a model to account for people's uses of symbolic resources.This model is organized around three dimensions: aboutness of the use, level of distantiation of the use and time orientation of use.
The aboutness refers to the idea that a symbolic resource is a mediating tool.One can thus differentiate uses which are mediating one's relationship to oneself, mediating one's relationship to another person, or mediating her relationship to the world (Rabardel, 1999;Vygotsky, 1934).
Uses of symbolic resource as mediation tools can then imply more or less distantiation from the here and now of one's experience.One can thus distinguish different "levels" of distantiation on a continuum that goes from total identification to the here and now of the flow of experience, to a very generalized, detached experience.We identified four levels of distantiation, partially following Valsiner (2005).A person can thus feel that her readings make her a more optimistic person, which is a very general value (highest level of distantiation based on symbolic mediation -Level 4); she can rather think that it helps her to distinguish among sorts of people, which is a category (Level 3); it can guide her behaviour in specific situation, for example when she has to take a decision (Level 2); and finally, it can mediate very undefined, embodied feeling and experience, such as when a painting seems to capture one's own sadness (lowest level of mediation -Level 1).
These uses can be in the immediate present, but they can also be means to connect to the past (e.g., when a picture makes us recall some past events) or to the future (e.g., when seeing a movie about university life helps to project oneself in one's future).
Finally, we stated that people could be more or less reflective about their uses of symbolic resources.Some people could simply observe the mediational effects of cultural experiences, while other could articulate them or could be actively looking for them.It is possible to locate most of people's uses of symbolic resources collected through case studies in a space organized around these three dimensions.

Construction of a questionnaire
To progress in the study of uses of symbolic resources, one of us has attempted to develop a questionnaire that could capture the modalities of uses of symbolic resources by people in a more systematic way.To construct such a questionnaire raises particular problem: how to develop a tool that could be applied to a wide number of participants while referring to one's person unique, significant cultural experiences and symbolic resources?To solve this issue, we adopted a methodological strategy consisting in combining an ideographic phase and a nomothetic one (Riediger, 2001;Scheibe, 2005;Wrosch & Heckhausen, 2002).
In a first part of the questionnaire participants are asked to explore personal experiences corresponding to a certain type.Hence, the first part of the questionnaire explains what a "cultural experience" is (such as reading a novel, listening to music, watching a film), and when they can take place; it also explains that a cultural experience always requires a "symbolic object" such as a book, a film, an image, etc.It guides the person through remembering and identifying such experiences.It then offers some space to write down a list of such experiences.The second part is an attempt to evaluate systematically the (mediated) experiences evoked by the person.The questions are standardised, but each participant is asked systematically to keep in mind the experiences he or she has identified in the first part of the questionnaire.
Among the dimensions of uses of symbolic resources, we decided to ignore the "aboutness" of uses and to focus on the: a) time span of the use of the resource, b) the level of elaboration at which it is used (with four possible levels: emotion regulation, local conduct, world & self categories, commitments and values), and c) whether the use is reflective or not.In order to define items for these dimensions, we used examples spontaneously proposed by young people interviewed in the UK (Zittoun, 2006).As mentioned below, these interviews also provided the data leading to the proposition of a theoretical model of uses of symbolic resources.Hence one has to question how accurate the operational model is.Does it enable to account for more general uses of symbolic resources?Are the items extracted from our qualitative study strong enough to capture the diversity of people's uses of resources?
In this paper, we report about an adaptation of this questionnaire in Serbian which aims at testing this model.This implied, first, to translate and adapt the items for the purpose of a Serbian study; second; to treat the categories identified above as independent variables; third, to collect data; fourth, to analyse the data so as to identify emerging dimensions; and finally to verify whether these are related, and related so as initially theoretically designed.

METHOD
The aim of the analysis was to examine to what degree the proposed theoretical model fits data collected in the empirical study, so as to revise the initial model according to recommendations emerging from the analysis.Besides that, examining the answers obtained from a large sample of participants should help us to choose the items that are better representatives of the dimensions under scrutiny in order to get a shorter and more valid version of the questionnaire.

Sample
The final version of the Serbian Symbolic Resource Questionnaire was administered to 475 participants, aged 17-23 (293 senior high school students, and 182 university students).The number of males (N = 121) and females (N = 354) was not balanced.However, as there is no theoretically or empirically based reason for us to believe that general socio-demographic variables modify the use of symbolic resources in a systematic way, we will not consider the lack of equality of case distribution between gender categories as a problem.

Instrument
On the basis of the theoretically defined dimensions and the suggested items, we had to adapt the questionnaire to the specificities of the Serbian speaking community, in cultural-linguistic and practical terms.
The first issue we encountered during the adaptation process was related to the question of translating the original items to Serbian, which was performed by two independent translators.It turned out that, in a few points, it has been very difficult to hold on to the literal meaning of some phrases and idioms.Hence, the main concept of "cultural experience", to which the items refer, does not correspond to any of the usual utterances in Serbian language.Even after instruction and explanation of the phrase, participants cannot avoid ambiguous and vague understanding of it.We can look for an explanation of this observation in the cultural specific connotations and associations that both words from the expression "cultural experience" have in a Serbian speaking community.Firstly, the term "experience" is more commonly associated with the notion of accumulation of the knowledge or skill that results from participation in some activities, than it is understood as direct observation or involvement in an event that entails emotional sensation or particular state of mind.Secondly, and more important, the attribute "cultural", when it is translated with the most frequent utterance, literally means adherent to social norms, manners and accepted usages and behaviours (i.e.civilized, cultivated, polite).Even if we can avoid this specific interpretation, the most we can attain is relating "cultural" with distinctive ways of living, values and shared knowledge of ours, or foreign, social groups (i.e.ethnical).Therefore, in order to guide our participants towards a mobilisation of experiences related to their encounter with some specific cultural artefact, we decided to replace the expression "cultural experience" used in the English version of the questionnaire by a description and explanation3 .Hence, in the Serbian version, the instructions at the beginning of the questionnaire mention people's "experience with an artistic piece" or "experience with a cultural artefact", and in the main part of the questionnaire cultural artefacts that can be used as symbolic resources were concretized, and every item referred to "reading a novel, listening to music, watching a film" instead to the general expression.This was aimed at gaining shared understanding of the items and easing participants the process of linking the item content to their personal experiences.The choice of these specific artefacts (books, films, music) was justified by data collected in previous case study (English sample) and in sequel preliminary inquiry (Serbian sample), where these three categories appeared as the most frequent and were considered as the most significant.
Finally, the questionnaire was constituted of 64 items, to be evaluated by a five points Likert scale (1 -never applies to me, 5 -always applies to me)4 .The items were considered to be indicators of latent dimensions of uses of symbolic resources (see Table 1): Time orientation of the use of the resource was limited to uses oriented towards past vs.towards future (4 items each) Levels of mediation of experience (i.e.levels of differentiation that goes from total identification to the here and now of the flow of experience, to a very generalized, detached experience) Level 1: Emotion regulation (6 items) Level 2: Local conduct (7 items) Level 3: World & Self categories (6 + 7 items) Level 4: Commitments and values (7 items) Expertise of use (Reflexivity) -we can differentiate here whether the person is reflective when it comes to engage into some cultural experience and recognizing specific affect by it (Reflective I -6 items); and whether the choice and use of symbolic resource is expert and deliberate (Reflective II -8 items) Real/imagination boundaries -this additional dimension was designed to capture parasite phenomena.It refers to a person's tendency and eagerness to suspense real/imagination boundaries and freely involves himself in spheres of imaginary experience enabled by cultural artefact, which is supposed to have facilitating effect on its internalization and later mobilization as symbolic resource.Further, according to psychoanalytically oriented theory, the ability to maintain a distinction between reality and fiction might be a precondition of the ability of using cultural experiences as semiotic mediator to reflect upon reality (Winnicott, 1971).We thus designed 9 items to capture possible confusion between real/imagination boundaries.6,32,34,36,11,12,24,29,41,64 F5 Commitments & Values 4,15,44,45,46,55,61 F6 Expertise of use Reflective 8,25,26,43,14,20,23,27,30,37,59 F7 Real/imagination bound. 13,16,18,22,31,35,49,56,62 It was also supposed that the proposed dimensions are substantive and firmly inter-related.Although answers on some items could be affected by more than one dimension, we assume in the initial model that each item indicates one specific dimension at the first place.
Pilot research.Before administering the questionnaire to a large sample of adolescents, a pilot research was performed.The motivation for that step came from the socio-cultural nature of the phenomenon in question.Even though it can be assumed that socio-economical origins will not make a difference in the use of symbolic resources (Zittoun, 2004(Zittoun, , 2006)), the way young people express and describe their cultural experiences, and the words and phrases they use to capture personal meanings, are very likely to be culture-specific.Adoption of a research technique such as questionnaires makes this issue especially relevant, since the researcher has no opportunity to negotiate meanings and explore associations with the participants.So, the aim of the preliminary study was to reach for spontaneous verbal formulations of Serbian participants through individual interviews and to compare categories gained this way with the ones based on the previous research, now constituting the core of the questionnaire items.The second task was to explore possible meanings, associations and connotations that arise when confronted with the content of the items.This served to modify the first translated version of the questionnaire in order to make it more apprehensible and closer to participants from Serbia.The preliminary research consisted of five semi-structured interviews with Serbian adolescents that were carefully selected to be actively interested in cultural experiences and eloquent enough when it comes to talking about their reflections.

Procedure
University students filled in the questionnaire in large groups, while sitting in classrooms, either before the beginning or after their regular lectures.High school students filled in the questionnaires during the regular school classes (it took approximately 25 to 30 minutes).

RESULTS AND DISCUSSION
The statistical technique we used for analyzing the research data is a confirmatory factor analysis -a structural equation modelling (SEM) -which allows us to test hypotheses about latent dimensions and their inter-relations by providing fit measurements indicating compatibility between the theoretical model and empirical data.The initial model was modified according to the first results from the analysis in order to match the data better, of course only when that action was reasonable and coherent with the theoretical assumptions.
The initial model tested is covariant (all latent factors are of the same level, i.e. no causal effect between them is assumed), has 64 observed endogenous variables (answers on survey items) and 71 unobserved exogenous variables (64 residual or error variances and 7 latent factors).Its structure is presented in the Table 1.

Evaluating the initial model
The null hypothesis under chi-square test, which is the test of absolute fit, is that the model fits the data.This analysis shows that absolute fit is statistically significant (χ 2 = 4458, df = 1931, p = .000).This result is expected considering the fact that model involve big number of variables.In that case, only small deviations of expected parameter values from empirical measures obtained generate false negatives on the test of absolute fit.Besides, the χ 2 test of absolute model fit is sensitive to insufficient sample size and non-normality in the underlying distribution of the input variables and it is expected to get convergence failures and underestimated parameters, that is, false negatives.In that case it is advisable to turn to other, relative fit statistics to assess the overall fit of the model to the data.
Commonly reported relative fit statistics are Cmin/df (the minimum discrepancy, i.e. χ 2 statistic, divided by its degrees of freedom) and the RMSEA (root mean square error of approximation).The absolute and relative fit statistics for the initial model are presented in the table 2. There are border values of these statistics proposed by statisticians that show if the tested model is acceptable as an adequate description of the structure of the empirical data.Thus, a Cmin/df ratio lower than 3 (or even 5 in some opinions) is indicative of an acceptable fit between the hypothetical model and the sample data (Carmines & McIver, 1981); ours is 2.39.Regarding RMSEA, which is the only measure that favours big complex models like this presented here, a value lower than .08 is considered to be acceptable if we want to make positive conclusions about the fitness of the model (Browne & Cudeck, 1993); ours is .05.Thus, it could be stated that the initial model satisfactorily fits the data, but in order to meet our other research goals, namely to offer a shorter and more valid version of the questionnaire, we made some modifications.

Modifications of the initial model
There were two steps in the redefinition of the model.First, some items were moved from dimension they have originally been attached to and considered to be indicators of other latent dimensions, if their regression weights were low and if that was consistent with theoretical assumptions of the questionnaire.
For example, the next two items -Q15 "I try to find a "moral" in every cultural experience" and Q21 "I keep pictures, tapes, books, from travels or from my past" -were moved from dimensions Commitments & Values and Time orientation, respectively, to dimension Expertise of use, because they both refer to manifestations of deliberate and reflective uses of symbolic resources for some personally relevant purposes5 .
Second, some items were excluded because they had insignificant or low regression weights on all latent dimensions they were connected to6 .These items might be indicators of some other dimensions that have been bypassed during the construction of the instrument (e.g.social dimension which denotes interpersonal uses of cultural experience).Alternatively, the content of the items is vaguely formulated, or for some other reason unclear to the participants.
To summarize, there are four reattached items and eight more that have been excluded from further analyses (the modified model is presented in the Figure 1).
The modified model presented has 56 observed and 63 unobserved variables7 .Absolute and relative fit measures for the modified model are presented in the table 3. The change in absolute fit indices is statistically significant (χ 2 is decreased for 970 for change in degrees of freedom of 468) while relative fit measures stay almost the same.

Latent correlations
The table 4 below shows that all latent dimensions are significantly positively correlated and that some correlations are very high.The theory suggests that the dimensions of use of symbolic resources are distinct, but strongly connected and in continuing interaction, and these results could be a statistical confirmation of the stated idea.Dimensions that have the closest inter-relationships are Local conduct, World & self categories and Commitments & Values (latent correlation is above .940),that refer to three upper levels of mediation of experience (i.e.levels of differentiation).On the other side, their latent correlations to the first level of mediation of experience (Emotion regulation) is significantly lower (r is under .750)although still very high.
One more thing is evident from the table above -the dimension Expertise of use is relatively strongly connected with all other dimensions (latent correlations are between .824 and .922),except dimension Real/imagination boundaries (r =.563), which is on the contrary weakly connected with all the others (r is between .563 and .749).We think that this is also expected.Indeed, in order to answer to items about personal uses of cultural elements, a person needs to be reflective.This is particularly true when it comes to uses on higher level of semiotic mediations, and great number of items refers precisely to these aspects of personal experience.On the other hand, we have one parasite dimension, which does not originate from the theoretical model of uses of symbolic resources (Zittoun, 2006), and which describes involvement in spheres of imaginary experience enabled by cultural artefact.It is more related to the possibility of providing a suitable environment for the internalization and later exploitation of symbolic resources, than to the aspects of the use itself.Some other possible interpretations of these correlations will be discussed later.

DISCUSSION
The final model suggested is some kind of compromise between the theoretical assumptions about the role of symbolic mediation that underline the construction of the questionnaire, and the outcomes of the analysis of empirically produced data.
The explanation of the fact that some items are very and difficult to consider as clear indicators of just one dimension, could be found in their specific origin.Namely, all items are adopted from interviews, so they present spontaneous verbalizations coming from participants, and denote reflections of their personal experiences with symbolic resources.We cannot expect to find in their personal narratives some pure analytical categories, especially when we are aware of the fact that aspects of symbolic mediation are closely associated.It should be noted that, even though the modifications of the initial model following the results of analysis did not get significantly better correspondence to the data, it fulfilled the practical goal by offering shorter and improved version of the questionnaire, which could be further validated and used in future researches.
We consider one more issue to be relevant for a reconsideration of the questionnaire and its further exploitation.We have considered the high correlation indices between the dimensions of use of symbolic resources as a strong confirmation of the factual close inter-connection between different aspects of use, which were initially distinguished primarily for analytical purposes.However, there are other possible interpretations of the results we presented here.One potential explanation could be that our participants were not able to differentiate sufficiently their experiences, either because of a lack of reflexivity, or because of the lack of proper means for expressing their thoughts, which results in a description of cultural experience which is not articulated enough.From the perspective of our subjects, different uses of cultural artefacts are embodied in one holistic picture, or general impression.
A second interpretation might be that there is something specifically culturallylinguistic reflected in these results.Although we did intend to simply test our questionnaire and did not plan to engage in an intercultural comparison, this first series of issues we were confronted with when trying to translate the term "cultural experience" is in itself revealing the diversity of the nature of phenomenon under investigation itself.For if a language does not have a generic term for a series of experiences involving an imaginary experience, that is, does not have a concept to organise them in a hierarchical way, then it might be that people's phenomenological sense of these and their links are quite different than the sense that people speaking a language that propose such concepts.
A third interpretation for these strong correlations can be found in the solutions proposed to the previous problems at the level of the construction of the questionnaire itself.In the instructions, at the beginning of the questionnaire, participants are guided to identify their own experiences with cultural artefacts and to write them down.Then, we expect them to keep in mind these experiences they have written as they answer to the various questions.The problem here is that different personal experiences might imply different uses of symbolic resources and lead person to different answers on the same item.In this situation, participants might tend to generalize these different experience with different symbolic resources and to give answer based on the generalized experience.We suppose that at this general level, items and latent dimensions might be correlated more than at the level of specific experience with some book, movie or song.
How can we overcome the methodological problems identified so far?Our suggestion would be to change the instruction given to the participants.In order to gain more differentiated uses of symbolic resources, we should ask them to concentrate on one single concrete cultural experience that was very important to them for some reason.In addition, the items in the main part of the questionnaire should refer only to the symbolic resource that the particular subject used in the experience he or she emphasized.Maybe then we could hope to get individualized answers, which would better reflect someone's personal uses of symbolic resources.

CONCLUSIONS
The analysis has shown that, in general terms, our theoretical model is an adequate representation of the data structure, and, consequently, of the phenomena of uses of symbolic resources.The proposed aspects of the uses of symbolic resources act as coherent, distinct and strongly inter-related dimensions.In addition, we met both intentions set at the beginning -proposed model is modified on the basis of outcomes emerging from the analysis, which also allowed us to offer a less extensive, but in the same time more coherent and better grounded version of the questionnaire, that can be used in future studies of the similar phenomena.Finally, we underlined some methodological issues potentially relevant for the interpretation of the results and proposed directions of further adaptation of the questionnaire, which could overcome some problems identified so far.

I. Thinking about cultural experiences
We would like you to think about cultural experiences that you had.Cultural experiences are experiences with objects that are part of our culture, such as reading novels, biographies, or comic books.Cultural experiences are also experiences of watching films, TV programs, and movies; experiences like seeing paintings, posters, or graphics; or experiences like listening to music, at home, in the street, in the car, or at a concert place or a festival.Cultural experiences can also be experiences of reading, interpreting, or meditating on the Bible or other religious texts, and celebrating religious events.
You will be asked to concentrate on such experiences that you have had and that were somehow important for you.Of course, cultural experiences that you find important might be not so important to someone else.That's fine.It is about cultural experiences that you had.Now, how can you be sure that an experience is a cultural experience?A cultural experience always requires the use of a "symbolic object": a book, a painting, a disk, a picture, notes, a film, a tape, etc.And how do you know that a cultural experience has been important for you?You can know that it was important, perhaps because you were thinking about that cultural experience (that book, that song, that movie, etc...), or maybe you discussed it with friends or other people.Maybe you started to read or to look for information about it.You might have dreamed about it, or have had "flash backs".Maybe you just tried to have the same experience again (see the same movie again, read the same pages, listen to a song that you know), may be you advised friends to have (or not to have!) the same experience.Or you started to do some things (sing songs, telling jokes), or buy some objects (clothes, instruments, etc...) related to the cultural experience.Maybe also you wrote something about it.Or maybe you just know it is a cultural experience that was interesting or moving for you.
Please take some time to think about cultural experiences that you had.Now, please list a few cultural experiences that you had (between 2 and 10).You can think about cultural experiences that you just had, or that you had during the past few months, or the past two years.If you can't remember any important cultural experiences during the past two years, think about cultural experiences that you had during your adolescence, or during your childhood.You can also just mention a few cultural experiences that you had, even if you are not sure that they were important to you.
Please keep these cultural experiences in your mind when you answer to the following questions.

II. The role of cultural experiences in your life
In what follows, you will be asked to evaluate how much each statement applies to describe your cultural experiences or your memory of these.Please read carefully each statement and try to connect it with your own cultural experiences.You can think about cultural experiences that you have listed on the previous page.You can also think about any other cultural experiences you had.
Remember: A cultural experience is an experience linked to listening to music or songs; watching movies or TV programs; looking at paintings, sculptures, or pictures; reading or meditating on religious texts… For each statement, please indicate on a scale from 1 to 5 how much it applies to you.

Figure
Figure 1.Structure of the modified model