Abstract: Hierarchically structured data are common in many areas of scientific research. Such data are characterized by nested membership relations among the units of observation. Multilevel analysis is a class of methods that explicitly takes the hierarchical structure into account. Repeated measures data can be considered as having a hierarchical structure as well: measurements are nested within, for instance, individuals. In this paper, an overview is given of the multilevel analysis approach to repeated measures data. A simple application to growth curves is provided as an illustration. It is argued that multilevel analysis of repeated measures data is a powerful and attractive approach for several reasons, such as flexibility, and the emphasis on individual development.

Abstract: There has been considerable debate on how important goodness of fit is as a tool in regression analysis, especially with regard to the controversy on R 2 in linear regression. This article reviews some of the arguments of this debate and its relationship to other goodness of fit measures. It attempts to clarify the distinction between goodness of fit measures and other model evaluation tools as well as the distinction between model test statistics and descriptive measures used to make decisions on the agreement between models and data. It also argues that the utility of goodness of fit measures depends on whether the analysis focuses on explaining the outcome (model orientation) or explaining the effect(s) of some regressor(s) on the outcome (factor orientation). In some situations a decisive goodness of fit test statistic exists and is a central tool in the analysis. In other situations, where the goodness of fit measure is not a test statistic but a descripitive measure, it can be used as a heuristic device along with other evidence whenever appropriate. The availability of goodness of fit test statistics depends on whether the variability in the observations is restricted, as in table analysis, or whether it is unrestricted, as in OLS and logistic regression on individual data. Hence, G 2 is a decisive tool for measuring goodness of fit, whereas R 2 and SEE are heuristic tools.

Abstract: Most scholars doubt that voters are able to explain their own vote. We argue that introspective questions whereby respondents are invited to tell, in their own words, the reasons why they vote the way they do, provide useful information on which considerations are most salient in their voting decisions. We show that open-ended questions about reasons for voting Yes or No in the 1995 Quebec referendum on sovereignty help us to sort out subgroups of voters for whom a given consideration is more salient.

Abstract: In this paper we use log-linear models with latent variables to estimate and compare the structure of job characteristics and their relation to positions in the Goldthorpe class schema for male and female employees. We proceed by estimating latent class models of job attributes and then evaluate the degree of similarity in these models between men and women. Sex differences in the patterns of association between the various latent types of job attribute and between those attributes and the Goldthorpe class schema are then assessed. The structure of job attributes is found to be generally similar for men and women; and the Goldthorpe schema is shown to be a similarly effective predictor of these attributes for both sexes. The class structure of employment relations is therefore argued to be equivalent across the sexes, so confirming the generality of class relations as conceptualised by Goldthorpe and his colleagues.

Abstract: This research paper has explored the image of China portrayed in the New York Times from 1949 through 1988. The study has shown that the image of China has changed in accordance with the changes in the U.S. foreign policy toward China over the past forty years. The changes were discussed in the historical context of Sino-U.S. relations. In addition, this paper has explored some methodological issues relating to Q-analysis. This methodology, different from Q-sort technique, is proven particularly powerful in this study for describing semantical structures in a cluster of news stories.

Abstract: When handling missing data, a researcher should be aware of the mechanism underlying the missingness. In the presence of non-randomly missing data, a model of the missing data mechanism should be included in the analyses to prevent the analyses based on the data from becoming biased. Modeling the missing data mechanism, however, is a difficult task. One way in which knowledge about the missing data mechanism may be obtained is by collecting additional data from non-respondents. In this paper the method of re-approaching respondents who did not answer all questions of a questionnaire is described. New answers were obtained from a sample of these non-respondents and the reason(s) for skipping questions was (were) probed for. The additional data resulted in a larger sample and was used to investigate the differences between respondents and non-respondents, whereas probing for the causes of missingness resulted in more knowledge about the nature of the missing data patterns.

Abstract: This article has two aims: first, to define and analyse the theoretical implications of the concepts of social urban identity and symbolic urban space (Valera, 1993, 1996, 1997; Valera and Pol, 1994) from within the integrated perspective provided by Environmental Psychology and Social Psychology; and second, to present an empirical study of these concepts based on nonquantitative data collection and analysis. Texts recorded in discussion groups were studied using Contextual Content Analysis (McTavish and Pirro, 1990) to determine the social identity of the inhabitants of a neighbourhood within the city of Barcelona and the meaning given to those spaces which they consider to be symbolic or representative of the neighbourhood. Finally, the results, as well as the advantages of this type of analysis, which are rarely used in Environmental Psychology, are discussed.

Abstract: A look at comparative political science shows that many contributions to the field do compare, but do not reflect the methods used. This paper discusses the analytical differences between the comparative method, the statistical method and experiment. An overview of the methodological discussion shows that a case study and namely the analysis of deviant cases has its place in comparative analysis. The last part of this paper consequently deals with different forms of case study analysis and the underlying research interest. The discussion emphasizes that in-depth analyses of a limited number of cases is the core of comparative political science with a good ratio between methodological input and analytical results.

Abstract: Q-analysis has proven a useful research methodology in a number of social science research areas for exploration of structures in human and social phenomena. Unrelated to either Q-mode factor analysis or the Q-sort technique, Q-analysis combines an anti-reductionist ethos with algebraic description. This paper addresses Q-analysis's potential contribution to content analysis of print communication. Twenty-seven news stories covering the “drug war” from the New York Times are analyzed to illustrate the Q-analysis approach and conceptual comparisons with traditional content analysis methodology are drawn. Advantages and disadvantages are evaluated. The report concludes that Q-analysis is a promising approach to content research that makes possible the mathematical expression of content properties beyond the reach of traditional content analysis methods.

Abstract: Young's C statistic (1941) makes it possible to compare the randomization of a set of sequentially organized data and constitutes an alternative of appropriate analysis in short time series designs. On the other hand, models based on the randomization of stimuli are also very important within the behavioral content applied. For this reason, a comparison is established between the C statistic and the Edgington model. The data analyzed in the comparative study have been obtained from graphs in studies published in behavioral journals. According to the results obtained, it is concluded that the Edgington model in experimental designs AB involves many measurements while the C statistic requires fewer observations to reach the conventional significance level.

Abstract: The purpose of the paper is to compare results of estimation and inference concerning covariate effects as obtained from two approaches to the analysis of survival data with multiple causes of failure. The first approach involves a dynamic model for the cause-specific hazard rate. The second is based on a static logistic regression model for the conditional probability of having had an event of interest. The influence of sociodemographic characteristics on the rate of family initiation and, more importantly, on the choice between marriage and cohabitation as a first union, is examined. We found that results, generally, are similar across the methods considered. Some issues in relation to censoring mechanisms and independence among causes of failure are discussed.

Abstract: A frequent variety of typology is that which results when a given population is classified simultaneously by several criteria, the categories of each criterion being ordered in a sense common to all the criteria. Such simultaneous orderings automatically define a partial order typology.The empirical problem is to ascertain the dimensionality and the substantive meaning of the partial order. This is illustrated here by characterization of the Jewish population of Israel by the frequencies of use of a variety of drugs. The data analysis approach adopted is that of Partial Order Structuple (Scalogram) Analysis (POSA). The particular computer program used is POSAC (POSA with Base Coordinates). The analysis revealed that it is possible to portray the distinct profiles (structuples) of the respondents according to drug use in a two-dimensional space, with a substantive regularity relating both to legal (pain relievers) and illegal drugs (marijuana, methadon). This space also serves as a basis for discriminant analysis for age as an external background variable.

Abstract: This article explores the proposition that the appropriate use of suitable forms of graphic communication can improve the formulation and presentation of hypotheses in quantitative social science research. The creative nature of scientific diagrams is discussed and the technological advances in computer graphic media are seen as part of a ‘visual revolution’ which is markedly changing not only the way we see things but also the way we think and do things today. Brief historical views on the use of hypotheses and diagrammatic languages in science are given. The restricted use of graphic communication tools in social research academic documents is discussed and the importance of using well-designed data graphics in the production and transmission of scientific knowledge is highlighted. Hypotheses are conceptualised and their importance within social research is emphasised. A methodological approach for formulating hypotheses graphically is proposed based on the use of three types of language: notation, statement (ordinary language) and diagram. Some criteria are suggested for the selection of diagram type dependent on the related variables. Several examples are given covering the different models proposed.

Abstract: This paper deals with problems encountered in analyzing how an individual is deterred from committing a crime by the severity and probability of punishment. It is argued that it might be advantageous to base such an analysis on a model of maximization of expected utility. According to this model, the attractiveness of committing a crime is strongly affected by the product of the relative utility of committing the crime and being punished as compared with the utility of not committing the crime times the probability of the punishment. This implies that the bivariate linear relationships of choices to commit or not to commit crime and the severity and probability factors are dependent on the variations in both these factors and their mean values. In this paper, these bivariate relationships are analyzed in two ways – formally-algebraically and by numerical examples.

Abstract: The demonstration of non-spuriousness is both critical in making causal statements and extremely difficult Unfortunately this issue is often summarily dealt with in methodological treatments in which spuriousness is reduced to instances of “common cause” It is argued here, among other things, that extraneous variables need not be prior to the independent nor necessarily causally related to either independent or dependent variable This being so, spuriousness is a far more common theoretical problem than is often currently acknowledged, and we might do well today to listen again to the advice of an earlier generation of sociologists whose work on the topic deserves more careful attention than it appears to have received

Abstract: Although the weights in a discriminant function (both linear and quadratic) are independent of group prior probabilities, the performance of the classifier (on the training and validation data) is sensitively dependent on these often unknown probabilities. After reviewing some defects of a popular measure of performance in the situation where the group sizes are naturally disproportionate, three alternative measures of performance (or association) are considered and it is shown that the behavior of the measures as a function of group prior probability is different between measures. Consequently, the optimum choice of the group prior probability depends on the specific measure of performance. Among the measures considered, only two measures - the index of mean square contingency and the Heidke Skill Statistic - are found to be well defined in the disparate-group size situation, and are, therefore, recommended. An empirical data set, dealing with delinquency among high school students is employed to illustrate all of the findings.

Abstract: In some publications the mean is identified with the constant of a Box–Jenkins time series model. In this paper the relation between both terms is demonstrated. Furthermore, by means of an example, the errors which may be made if one does not use each term adequately are shown.

Abstract: In this paper it is said that cultural evolution consists of four phases which are denoted by the words nature, mythology I, mythology II and history. Pictograms, ideograms and logograms signify two phases fundamental for the later emergence of the linear writing and time as history and thought. The nowaday operational structures of cognition appear as rooted in prehistoric thought meaning vision, ideas – images seen with the inner eye – and logos as orientation of logograms. Thus it is assumed that logic in the form of the elementary first order propositional calculus should be derived from the orientation symmetry of space. The author gives an exact proof in terms of the Clifford algebra that this is indeed the case. Four basic representations in Cl2, Cl3, Cl3 and Cl3,1 of the 16 expressions of second order logic relations are given. The proposition of transformational self-reference is defined. The most basic self-referent representation of Peirce's box-X icons (the 16 binary logic connectives) is derivated from the spin-representation δD2d of the spatial congruence group of the square or the mandala of the sunwheel. Finally it is shown that the representation in terms of the idempotents of the Clifford algebra Cl3,1 is isomorphic with Zellweger's logic garnet.

Abstract: This article presents the global-mutable-immutable scheme. Globals are purely macro properties of societies, while immutables are purely micro properties of individuals. Mutables are the micro-macro bridging elements, having both an aggregated macro (distributional) form, as well as a micro form. The importance of this distinction for the analysis of norm, status and role is shown. The article concludes by showing how the configuration of globals, mutables and immutables affect the allocation of an individual into the five-dimensional mutable structure of his or her society at a given point in time.

Abstract: It is proposed to study the graphical representation of the parametric space of maximumlikelihood function of two parameters logistic functions, as is used in Item Response Theory. Thisproposal is made more from a point of view of understanding rather than of discovery..

Abstract: Aggregate analysis has been established as a standard method on the study of market response behavior for a long time. Aggregation has advanced our understanding of the linkages among social characteristics and aggregate response behavior. However, aggregate analysis has been hindered by fragmentary and unsystematic procedures to determine the most appropriate level of aggregation. The general objective of this paper is to provide a conceptual framework to determine the level of aggregation of variables in data analysis. In addition, statistical procedures are suggested in this framework to verify and to determine the level of aggregation represented by a variable. The conceptual framework is useful for deciding if the variables are to be analyzed from micro-analysis focus or macro-analysis focus. The statistical procedures enable the researcher to systematically identify and verify the level(s) of aggregation of variables in an existing data set.

Abstract: The essay takes the reader on a voyage of exploration with the aim of discovering the origin of Pareto-curves. It shows with thought-experiments backed up by computer-simulation the generation of log-normal curves in detail. Extending forward this conceptual trajectory, it arrives via a quasi-Newtonian “fluxion”-insight – infinitessimal differential integration – at a novel mathematical concept: Pareto-curves are simply special log-normal curves where a large number of random-factors interacted and impacted at their genesis (the author called it the “Kopp-effect”).

Abstract: This article describes a multivariate effect size measure useful in social and behavioral research, which complements existing descriptive indicators of group differences. The measure reflects probabilistic aspects of the disparity of two multivariate distributions, and its interpretation does not require advanced training in statistics. Estimation of the discussed effect size indicator is illustrated on data from a two-group cognitive intervention study (Baltes, Dittmann-Kohli, & Kliegl, 1986).

Abstract: In this article we analyse the number of “no opinion” answers to attitude items. We argue that that number can be considered as a count variable that should be analysed using Poisson regression or negative binomial regression. Since we're interested in the effect of both respondent and interviewer characteristics on the number of “no opinion”'s we use multilevel analysis that takes into account the hierarchical structure of the data. As a consequence multilevel Poisson regression and multilevel negative binomial regression are applied. Our analysis shows that answering “no opinion” is related to some sociodemographic respondent characteristics. In addition we find a significant interviewer effect, but we are not able to explain that effect in terms of interviewer variables.

Abstract: A meta-analysis of 16 educational interventions and 10 needle exchange programs was performed to estimate the effectiveness of reducing HIV risk behaviors in the injecting drug user population. Information on intervention, outcome, design and demographics was coded and analyzed for all educational and needle exchange program evaluation studies published between January 1984 and May 1995. The weighted mean effect size for the 6,251 study subjects of the 16 educational interventions was 0.749 (95% CI, 0.708 to 0.790), and the weighted mean effect size for the 1,675 study subjects of the 10 needle exchange programs was 0.279 (95% CI, 0.207 to 0.352), suggesting that both interventions had a positive impact on reducing HIV risk behaviors associated with injecting drug use. However, these results were dependent upon research design, outcome type and follow-up time.