The methodology of the study presented above was characterised by outsideness (or other ness) and by a search for regularities instead of unique and personal life histories and interpre tations.
What mattered was whether there are general and generalisable effects of contextual factors on young people’s lives. This kind of information is important, if the practical or po litical importance of the results is to be assessed. The search for regularities does not imply determinism, however. The regularities that were found are seen as stochastic processes, which cause certain probabilities to occur. Therefore, individual variation and unique expla nations are acknowledged and much remains to be explained in the educational careers of young people.
Because there is nothing in the results which discloses information on individual (and possibly rec ognisable) young people, and since the data set was constructed so as to not make personal identi fication possible, many ethical concerns are avoided. However, the applied methodology may be faulted on ethical grounds for not allowing the young people have their say on the subject. It is clear that this study could not discover all there is to neighbour hood effects on young people’s edu cational careers, partly because of this limitation. On
the other hand, given the possibility of an epistemological fallacy – failure to realise the effects of structural factors when there are such ef fects (Karvonen & Rahkonen 2000) – research into the importance of contextual factors cannot be limited to analysing the young people’s own views and interpretations.
The main ethical concern in the study was the possible stigmatisation of certain neighbourhoods. Because spatial patterns were presented as maps and some specific neighbourhoods were mentioned in the text, this could affect the neighbourhoods’ image.
However, I avoided putting too much em phasis on individual neighbourhoods and concen trated instead on the more general results. Addi tionally, since I did not find neighbourhoods at the lowest socio-economical level to have any det rimental effect on educational careers the results should not, in my opinion, stigmatise these neighbour hoods. They will more likely point out affluent neighbourhoods, which were shown to have positive effects. This may be considered a lesser ethical concern.
notes
1 It must be emphasised that these conclusions regarding the existence of neighbourhood effects are made on the supposition that the individual-level background variables adequately control for selection into neighbourhoods.
2 Empirical-Bayes residuals may be seen as predictions rather than observations.
Informa tion across all neighbourhoods is “lent” when calculating the estimate for each neighbour hood, and in this calculation, especially the estimates for neighbourhoods with only few ob servations are “shrunk” towards zero (no deviation from the average). This improves the overall precision of the estimation but also makes it “conservative”.
3 The “empty” model actually included one explanatory variable in this case:
a cohort vari able, which indicated the year of turning 16, was included in all models order to control for temporal differences in educational, employment and welfare policies and employment opportunities.
4 The exact meaning of the odds ratios may be difficult to grasp, and therefore in my thesis I illus trated the most important effects by transforming the odds ratios back into probabilities. The problem with that approach is that in a model
with several categorical control variables it is dif ficult to present probabilities that would refer to an “average” boy or girl. Instead, the probabili ties that I presented referred to a “relatively average” group, whose values for all the explana tory variables were the reference categories (which were “relatively average” in their effects).
references
Bryman, A. (1988). Quantity and Quality in Social Research. London:
Routledge.
Buck, N. (2001). Identifying Neighbourhood Effects on Social Exclusion, Urban Studies 38(12): 2251–2275.
Clayton, D. & Hills, M. (1993). Statistical Models in Epidemiology. Oxford:
Oxford University Press.
Dietz, R. D. (2002). The Estimation of Neighborhood Effects in the Social Sciences:
An Interdisciplinary Approach, Social Science Research 31: 539–575.
Diez-Roux, A. V. (1998). Bringing Context Back into Epidemiology: Variables and Fallacies in Multilevel Analysis, American Journal of Public Health 88:
216–222.
Duncan, G. J. & Raudenbush, S. W. (1999). Assessing the Effects of Context in Studies of Child and Youth Development, Educational Psychologist 34(1):
29–41.
Elliott, D. S., Wilson, W. J., Huizinga, D., Sampson, R. J., Elliott, A. & Rankin, B.
(1996). The Effects of Neighborhood Disadvantage on Adolescent Development, Journal of Research in Crime and Delinquency 33(4): 389–426.
Furlong, A. (1996). Opportunity structures and the management of transitions, in H.
Helve & J. Bynner (eds.), Youth and Life Management. Research Perspectives, 79–99. Helsinki: Yliopistopaino.
Goldstein, H. (1995). Multilevel Statistical Models. (2nd Edition). London: Edward Arnold.
Goldstein, H., Browne, W. & Rasbash, J. (2002). Partitioning Variation in Multilevel Models, Understanding Statistics 1(4): 223–231.
Greenland, S. (2000). Principles of multilevel modelling, International Journal of Epidemiology, 29: 158–167.
Halpern-Felsher, B. L., Connell, J. P., Spencer, M. B., Aber, J. L., Duncan, G. J., Clifford, E., Crichlow, W. E., Usinger, P. A., Cole, S. P., Allen, L. & Seidman, E. (1997). Neighborhood and Family Factors Predicting Educational Risk
and Attainment in African American and White Children and Adolescents, in J. Brooks-Gunn, G. J. Duncan & J. L. Aber (eds.), Neighborhood Poverty.
Volume I. Context and Consequences for Children, 146–173. New York:
Russell Sage Foundation.
Jencks, C. & Mayer, S. E. (1990). The Social Consequences of Growing Up in a Poor Neighborhood, in L. E. Lynn & M. G. H. McGeary (eds.), Inner-City Poverty in the United States, 111–186. Washington, D.C.: National Academy Press.
Jones, K. & Duncan, C. (1998). Modelling context and heterogeneity: applying multilevel models. In E. Scarbrough & E. Tanenbaum (eds.), Research Strategies in the Social Sciences, pp. 93–123. Oxford: Oxford University Press.
Karvinen, M. (1998). Huono-osaisuus ja disorganisaatio kaupungeissa.
Modernisaation, muuttoliikkeen ja paikallisten olosuhteiden vaikutusten jäljillä. (“Disadvantage and disorganisation in the city; tracking the effects of modernisation, migration and local conditions”). Helsinki: Suomen Kuntaliitto.
Karvonen, S. & Rahkonen, O. (2000). Nuorten elämäntyylin ja terveyskäyttäytymisen kulttuurinen vaihtelu (“Cultural variations in young people’s lifestyles and health practices”), Yhteiskuntapolitiikka 65(2): 135–145.
Kauppinen, T. M. (2004). Asuinalueen ja perhetaustan vaikutukset helsinkiläisnuorten keskiasteen tutkintojen suorittamiseen (“Effects of neighbourhood and family background on the secondary educational achievements of young people in Helsinki”). Tutkimuksia 2004:6. Helsinki: Helsingin kaupungin tietokeskus.
Lieberson, S. (1985). Making It Count. The Improvement of Social Research and Theory. Berkeley and Los Angeles: University of California Press.
Oakes, J. M. (2004). The (Mis)estimation of Neighborhood Effects: Causal Inference for a Practicable Social Epidemiology, Social Science & Medicine 58(10): 1929–1952.
Peng, C., Ying, J. & Tak-Shing, H. S. (2002). Logistic Regression Analysis and Reporting: A Primer, Understanding Statistics 1(1): 31–70.
Puuronen, V. (1997). Johdatus nuorisotutkimukseen (“Introduction to youth research”). Tampere: Vastapaino.
Rankin, B. H. & Quane, J. M. (2002). Social Contexts and Urban Adolescent Outcomes: The Interrelated Effects of Neighborhoods, Families and Peers on African-American Youth, Social Problems 49(1): 79–100.
Raudenbush, S. W. & Bryk, A. S. (2002). Hierarchical Linear Models. Applications and Data Analysis Methods (2nd Edition). Advanced Quantitative Techniques in the Social Sciences Series 1. Thousand Oaks: Sage Publications.
Riley, M. W. (1963). Sociological Research. I. A Case Approach. New York:
Harcourt, Brace & World.
Sampson, R. J., Morenoff, J. D. & Gannon-Rowley, T. (2002). Assessing
“Neighborhood Effects”: Social Processes and New Directions in Research, Annual Review of Sociology 28: 443–478.
Snijders, T. & Bosker, R. (1999). Multilevel Analysis. London: Sage.
South, S. J. (2003). Interpreting Community Effects on Youth Educational Attainment, Youth & Society 35(1): 3–36.
Subramanian, S. V. (2004). The Relevance of Multilevel Statistical Methods for Identifying Causal Neighborhood Effects, Social Science & Medicine 58:
1961–1967.
Tabachnick, B. G. & Fidell, L. S. (2001). Using Multivariate Statistics. (4th Edition).
Boston: Allyn and Bacon.
Teachman, J. & Crowder, K. (2002). Multilevel Models in Family Research: Some Conceptual and Methodological Issues, Journal of Marriage and Family 64:
280–294.
Wolfinger, R. D. (1999). Fitting Non-linear Mixed Models with the New NLMIXED Procedure. SUGI 24, paper 287. Cary, NC.: SAS Institute.
towArds An
understAndIng of VAlue orIentAtIons: tHe CAse of estonIAn YoutH
by Andu Rämmer
Introduction
The issue of beliefs has attracted philosophers and theologians since the formation of primi tive societies. Besides theoretical speculations, an increasing number of empirical studies have been carried out in modern societies. At the beginning of the next chapter, I will give a brief overview of how researchers’ focus of interest shifted from opinions to values. Special attention will be paid to the interpretation of value orientations through different factor analy sis techniques.
The advancement of data analysis techniques made it possible to study various value patterns, namely value orientations that consist of aggregated value notions.
opinions
Empirical studies of social issues and public opinion imprinted the development of social psychology in the first half of the past century.
For a long time, psychology was seen as the science of attitudes (Moscovici 1963). In the first half of the twentieth century quite many re searchers were convinced that opinions and attitudes could be measured effectively with a single question in a questionnaire.
Attitudes were measured through their verbal expressions, opinions.
Subjects’ attitudes were measured through their acceptance or rejection of the opinions presented to them. Different measurement instruments were developed for that pur pose. These were not only analysing techniques that emerged during the twentieth century however. The concept of attitude itself was also transformed. Thomas and Znaniecki intro duced attitudes as an explanatory concept in social sciences, explaining attitudes as links between individuals and groups to which they belong. Augoustinos and Walker (1995, 30) note that mainstream social psychology has increasingly individualised the attitude construct. Already in the thirties, Gordon Allport defined attitude as a global stimulus-response disposi tion for the explanation of behavioural differences in similar situations. In that vein, attitudes and their verbal expressions – opinions – ceased to be used for the measurement of social phenomena.
Values
During the second half of the past century, values became the major focus of research. There are several reasons for that. As attitudes are highly context-dependent, there arose a need for a new concept that would be more comprehensive in the explanation of social issues.
Al though there is little coherence in the nature of value research (Hitilin & Piliavin 2004), there are some aspects of conceptualisation on which researchers agree. Values are about modes, means or ends with reference to the desired goals. According to Smith and Schwartz (1997, 80), researchers agree that values refer to major life goals and general modes of con duct that promote these goals;
values transcend specific actions and situations (but this does not mean that they are context-independent). Values guide cognition,
evaluation, and behav iour.
An important step in the belief studies was the introduction of ordinal scales. Rosenberg (1957) was one of the first to use ordinal scales in a large empirical value study. Many re searchers maintain that beliefs are complex phenomena that cannot be captured through sin gle values. For example, Babbie (1998, 167) noted that complex concepts like orientations re quire the researcher to use several questionnaire items to measure them adequately. A single variable might not provide the desired range of variation, whereas aggregate variables would give a more comprehensive and more accurate indication. The development of data analysing techniques made it possible not only to reduce the observed ratings to a smaller number of categories, but also to detect meaningful patterns of values. I will call these patterns value orientations, as they are based on different values and refer simultaneously to certain complex societal aims that can be achieved through work. One of the popular data reduction tech niques is known as factor analysis.
In the following text I will compare possibilities of different data reduction techniques using the example of post-Soviet youth values. Previous empirical studies of work values revealed similar work orientations both in Estonia (Titma 1979; Saarniit 1995; Titma
& Helemäe 1996; Titma & Rämmer 2002) and in USA (Johnson 2001). Most researchers agree that work orientations form on the basis of desired outcomes, which could be achieved through the working process. For example, Titma and Rämmer (2002) revealed that three substantial orientations emerged among the school-leavers in the 1980s: self-expression, social recogni tion and career, which all remained remarkably stable throughout one’s working life. It has been of interest to see to what extent these orientations re-emerge in the wake of the new millennium.