Geometric Data Analysis Research Articles

Political space and the space of polities: Doing politics across nations

What makes political life in the United States so different from political life in France, Hungary or Argentina? This paper considers why societies “do politics” differently. We draw on Pierre Bourdieu’s criticism of substantialist thinking in sociology and on his conceptualization of the social space to propose a new way of relating methodological to theoretical claims in comparative political sociology. We do this by exploring and constructing a “space of polities” based on data from the 2004 World Values Survey, using relational statistical techniques (e.g., geometric data analysis). The main insight is that any single form of political action (e.g., joining a voluntary association, or a demonstration, or a boycott) only takes its meaning in the context of its objective relationship to other forms of political action and non-action that have currency in each particular society. We explore the diversity of polity types that actually exist and discuss how they emerge from similar configurations in countries’ spaces of political practices. We suggest that the reason for such clustering lies in similar political–historical trajectories. We conclude by arguing for a comparative approach that is sensitive to differences in overall systems of relationships.

Citing the world: A geometric data analysis of Swedish literary scholars’ use of foreign critical resources

The academic study of literature constitutes one institutional site for the production and reproduction of conceptions of literature. In a semi-peripheral country such as Sweden, this production partly relies on foreign intellectual goods. To analyze this transnational dimension of Swedish scholarship in a period marked by increasing internationalization, a Geometric Data Analysis (GDA) (Le Roux & Rouanet, 2004) was carried out on the bibliographies of 318 PhD dissertations, defended in the period 1980–2005, at Swedish departments of literary studies (litteraturvetenskap). The analysis of citational choices showed only an insignificant increase in the reliance on foreign sources in this period. The GDA revealed how these privileged references were distributed in a tripolar opposition, reflecting fundamentally different conceptions of literature, interpreted in this study as the three poles of textual singularity, secular particularity and anthropological universality. The analysis of supplementary variables shows that these oppositions are subtended by different geolinguistic orientations and that they correlate strongly with gender, which is overwhelmingly in evidence as one moves from the male-dominated textual pole to the strongly feminist and female social pole of the first axis. The lack of increasing internationalization measured by citations is attributed to the “national cultural mission” of these departments.

Analysing the Symbolic Economy with Pierre Bourdieu: The World of Crafts

The world of crafts belongs to the ‘economy of symbolic goods’ identified by Pierre Bourdieu. The theoretical aim of this article is to show that Bourdieu’s analytical framework and methods are particularly useful to comprehend the structure of the diverse French craft economy and to understand the different ways in which symbolic and economic dimensions combine. First, the use of quantitative and qualitative methods (especially geometric data analysis) enables to highlight the structure of the world of crafts which remains divided between a traditional pole and a contemporary pole. Then, the apparent opposition between economic and symbolic aspects of crafts, represented by the two poles, is qualified by pointing out the different forms of ‘denial of the economy’ adopted by artist-craftsmen according to their economic and social characteristics. Finally, the article shows how economic and symbolic dimensions combine through the symbolic construction of the economic value of craft products.

Stability and hierarchy of quasi-stationary states: financial markets as an example

We combine geometric data analysis and stochastic modeling to describe the collective dynamics of complex systems. As an example we apply this approach to financial data and focus on the non-stationarity of the market correlation structure. We identify the dominating variable and extract its explicit stochastic model. This allows us to establish a connection between its time evolution and known historical events on the market. We discuss the dynamics, the stability and the hierarchy of the recently proposed quasi-stationary market states.

Pierre Bourdieu, Geometric Data Analysis and the Analysis of Economic Spaces and Fields

In this paper, we discuss the current relevance of geometric data analysis (GDA) in a new state of the field of economics where social science disciplines interested in economic phenomena tend to converge. We argue that this methodology can respond to the growing need to seek for more solid empirical foundations, and to adopt a multidimensional approach in social economics. We begin by setting out the basic principles of GDA methodology, which clearly distinguish descriptive and inferential moments of the research. Then, we show why and how Bourdieu used these methods in the studies he carried out in France in a variety of economic spaces and fields: lifestyle and consumption, corporate management, the individual housing market and publishing. We stress that GDA helps us to formalise the configuration and dynamics of economic institutions, groups and interests in a truly multidimensional and pluralistic way compared to standard methods used by mainstream economics.

Pattern Recognition in Narrative: Tracking Emotional Expression in Context

Using geometric data analysis, our objective is the analysis of narrative, with narrative of emotion being the focus in this work. The following two principles for analysis of emotion inform our work. Firstly, emotion is revealed not as a quality in its own right but rather through interaction. We study the 2-way relationship of Ilsa and Rick in the movie Casablanca, and the 3-way relationship of Emma, Charles and Rodolphe in the novel {\em Madame Bovary}. Secondly, emotion, that is expression of states of mind of subjects, is formed and evolves within the narrative that expresses external events and (personal, social, physical) context. In addition to the analysis methodology with key aspects that are innovative, the input data used is crucial. We use, firstly, dialogue, and secondly, broad and general description that incorporates dialogue. In a follow-on study, we apply our unsupervised narrative mapping to data streams with very low emotional expression. We map the narrative of Twitter streams. Thus we demonstrate map analysis of general narratives.

A dialogue of the deaf in the statistical theater? Adressing structural effects within a geometric data analysis framework

Since their introduction in the late 1960s, the “moderate”, and moreover “metrological” and “hypermetrological” uses of regression models quickly became the dominant quantitative approach in the Anglo-Saxon social sciences. This “sociology of the variables” has been the subject of many critical insights, with little impact on its dominance. By contrast, the French situation is quite different, mainly because of the strong association between Pierre Bourdieu’s research program and the correspondence analysis methods. In this context, the relationship between geometric data analysis and regression models has turned into a “dialogue of the deaf”. Complementarity is sometimes emphasized, correspondence analysis being associated with exploration and description of the data, and regressions being used to explain, reject or confirm assumptions. But regression models may also be used in order to analyze structural effects within a framework of geometrical data analysis, e.g. by visualizing graphically the results of a regression (Rouanet et al. in Math Sci Hum 160:13–46, 2002; Lebaron 2013). We propose a new multi-step approach, “Standardized Factor Analysis”, which relies on geometric analysis and uses linear regression in a second stage in order to uncover structural effects in the original space. We illustrate it with data about tastes for cinema in France. We conclude by raising a more general set of questions about causality: social determinisms, even well established, are partial in the sense that they produce their effects only when associated with each other.

A notion of nonpositive curvature for general metric spaces

We introduce a new definition of nonpositive curvature in metric spaces and study its relation to the existing notions of nonpositive curvature in comparison geometry. The main feature of our definition is that it applies to all metric spaces and does not rely on geodesics. Moreover, a scaled and a relaxed version of our definition are appropriate in discrete metric spaces, and are believed to be of interest in geometric data analysis.

Working in the methodological ‘outfield’: the case of Bourdieu and occupational therapy

The article reports on a study of methodological innovation involving occupational therapy (OT) students in higher education (HE). It is based on an original project which examined the experiences and outcomes of non-traditional entrants to pre-registration OT education. A feature of the original project was the application of the epistemological and methodological approach of the French social theorist Pierre Bourdieu, most noticeably in exploring the way that social back ground (habitus) interacted with the educational (field) context in terms of experience and educational outcome. Bourdieu used a ranged of techniques – both qualitative and quantitative – in collecting and analysing data. In particular, he used multiple correspondence analysis (MCA), a type of geometric data analysis recognized as a powerful tool enabling the representation of social space and the situating of individuals within it with respect to a number of variables. The article considers methodological principles in comparing ethnographic, traditional statistics and MCA. We show how the original data were reanalysed according to MCA. The article compares the original analyses and findings with those based on MCA in order to explore its strength over the previous approach and the potential it has to cast light on various issues in HE.

Performance Improvement in the Pattern Classification of Nominal Data Sets Applying Multiple Correspondence Analysis

Classification is a supervised learning problem used to discriminate data instances in different classes. The solution to this problem is obtained through algorithms (classifiers) that look for patterns of relationships between classes in known cases, using these relationships to classify unknown cases. The performance of the classifiers depends substantially of the data types. In order to give proper treatment to nominal data, this paper shows that the application of previous transformations can substantially improve the performance of classifiers, bringing significant benefits to the result of the whole process of Knowledge Discovery in Databases (KDD). This paper uses three different data sets with nominal data and two well-known classifiers: the Linear Discriminant Analysis (LDA), and the Naïve-Bayes (NB). For data transformation, the paper applies an approach called Geometric Data Analysis (GDA). The GDA techniques compared in this paper are the traditional Principal Component Analysis (PCA) and the underexplored Multiple Correspondence Analysis (MCA). The results confirm the capability of the GDA transformation to improve the classification accuracy and attest the superiority of the MCA in comparison with its precursor, the PCA, when applied to nominal data.

An analysis of complexity in primary school L2 English learning

In the context of a Complex Systems (CS) approach to language acquisition and teaching research, this paper will use Geometric Data Analysis (GDA, l’Analyse géométrique des données: Benzécri 1973; Le Roux & Rouanet 2010) to explore data collected from beginning learners of English (n = 54) in a primary school context in France. The study was initially designed along classical “group comparison” lines (six-year-olds vs. eight-year-olds), and not from the start for CS analysis; participants performed four different tasks in English at two times during their first year of English; they also completed a variety of psychometric tests and questionnaires measuring various cognitive, verbal, affective et conative characteristics. We will begin by looking at classic nonparametric comparisons of the two groups of learners and correlations between the psychometric tests and performance on the English tasks. These univariate statistics do not actually turn up very much structure in the data set, and we will use multivariate GDA for a closer look at the complex relationships between individual features and early L2 skill.

Theme Articles on Classification and Geometric Data Analysis

Theme Articles on Classification and Geometric Data Analysis

Late Jurassic-Early Cretaceous faulting in the Southeastern French basin: does it reflect a tectonic reorganization?

AbstractWe present new data constraining the Late Jurassic to Early Cretaceous evolution of the Southeastern French basin (SFB). Meso-scale faults with displacement of several decimeters to 1–2 meters were sampled for geometric and kinematic data analysis and the faulted and un-faulted sedimentary units were examined for sedimentologic and bio-stratigraphic purposes. Small-scale faults were also extensively collected to determine the local stress states during this period. Meso-scale normal faults formed during sediment deposition or before its complete lithification attest of a recurrent activity in the Late Jurassic to Early Cretaceous period in an overall extensional context. The fault network cutting the Oxfordian to Aptian sequences included normal faults of various trends, ranging between WSW-ENE and NNW-SSE. We show that the deformation mechanism in the SFB drastically changed in the Jurassic-Cretaceous transition (latermost Tithonian?), with the direction of extension rotating from a WNW-ESE to a NNE-SSW direction. Lateral thickness variation of the sequences, redistribution of sediments, faulting at various scales concur that the Early Cretaceous period marks a tectono-stratigraphic reorganization of the basin. We suggest that it traduces the rifting and later opening of the North Atlantic (main branch and bay of Biscay branch).

The Co-Production of Value in an Art Market: Exploring Service Relationships

The purpose of this research is to analyze service dominated relationships and differences in characteristics of these relationships within arts marketing. The appreciation of art is seen as a service and is intertwined with the creation of company identity and thus a subtle kind of marketing, whether this is a conscious or non-conscious strategy from a company perspective. Companies invest in art in order to enhance their image, and establish network relationship with art firms. The data is collected from the Swedish non-profit art market of public and private organizations with focus on Stockholm, an art market place as well as a prime site for company head quarters. The data is analyzed with multivariate correspondence analysis (MCA), which is also known as geometric data analysis (GDA). This method enables analysis of relationships as well as overall characteristics of the participants in these relationships: a part of service as well as a creation of company identity, thus marketing. The research has shown that the service encounter within arts is enhanced within relationships between key actors that can be considered co-producers of art and that multivariate correspondence analysis is a helpful tool to identify these relationships and company characteristics in a structured manner. The methodology is generating visual pictures of the art market that cooperates with profit making ordinary companies in different industries. The graph that is produced with MCA is a visualization of what can be regarded as a map of the companies with connections to the nonprofit art market, whether these connections are formal and a part of regular business activities or informal and relatively weak.

선형유형 구분을 통한 고속도로 사고빈도모형 개발 연구

본 연구에서는 고속도로를 대상으로 각각의 구간에 대한 선형유형을 구분하여 사고빈도모형을 개발하였다. 현재 사고빈도모형 부문의 연구는 주로 고속도로 구간 전체를 대상으로 한 연구가 대부분이기 때문에 거시적인 측면에서 사고빈도모형이 개발되었다고 할 수 있으며, 이에 따라 각각의 구간특성이 정확히 반영되지 않은 상태에서의 사고빈도를 예측하였다고 볼 수 있다. 본 연구에서는 이러한 문제점을 인식하여, 동질구간 분할법을 사용하여 고속도로의 평면선형을 직선부, 곡선부, 연속곡선부로 구분하였고, 이를 군집분석을 통하여 직선부와 곡선부의 유형을 구분하여 고속도로의 각각의 구간별 특성을 반영한 사고빈도모형을 구축하였다. 본 연구 결과는 고속도로 각 구간의 사고빈도를 예측하는데 있어 더 정확하고 합리적인 결과를 도출해 낼 것이라 판단한다. This paper presents how one can investigate the effects on crash occurrence of freeway geometric design elements including the horizontal, vertical alignment and road environment. At present, the available research results for the most part involve geometric data analysis that are obtained along a relatively long section of freeway, and, because of the long section's diverse geometric conditions, the results tend to miss the specific local geometric impacts on vehicle crashes. In this regard, this research attempts to establish vehicle crash models based on a set of freeway geometric patterns whose crash generating characteristics are identical because they are homogeneous in terms of producing the same vehicle operating speeds, and subsequently their actual relationships are described by providing statistical analysis made in this research. Also each standard is comprised of part of straight, curve and continuous curve. This research has revealed that each type of model has different relation between accident and geometry structure. This research results should be useful for doing more reasonable highway designs and safety audit analysis.

European Central Bank leaders in the global space of central bankers: A Geometric Data Analysis approach

In this article, we analyse the space of the world's central bankers since the second half of the 1990s on the basis of a collection of public biographical data. The world's central bankers compose a global ‘reference space’ for the study of ECB leaders. ECB leaders are distinctive in certain dimensions of this space, especially in relation to the Asian (Japanese and Chinese) central bankers and, although to lesser extent, in relation to the US Federal Open Market Committee members. These distinctions relate to the properties of two different sub-groups of ECB members of the Governing Council: one group composed of the members of the executive board, and another comprising the governors of the national central banks. This article stresses the methodological use of Geometric Data Analysis when applying a multi-level and sociological perspective to the central bank's decisions and discourses: in the process of examining the relevant social space, this approach allows us to study the inter-individual social variations between agents at a micro-level and to analyse institutional differences in behaviour as a probable consequence of these variations. Monetary strategies and economic discourses of central banks are analysed as the probabilistic products of a combination of individuals’ orientations, which in turn are the results of variations in their biographical experiences, based on different cultural and cognitive dispositions.

Academic formation and conceptions of accident and injury among non-English speaking students: a geometric data analysis hard look at terminologisation and injury notions

ObjectiveTo assess how the passage of Portuguese speakers through selected undergraduate university courses relates to conceptions of the words accident and injury.MethodsWeb-based survey of first and last-year students of medical,...

Advances in Studies and Applications of Centroidal Voronoi Tessellations

Centroidal Voronoi tessellations (CVTs) have become a useful tool in many applications ranging from geometric modeling, image and data analysis, and numerical partial differential equations, to problems in physics, astrophysics, chemistry, and biology. In this paper, we briefly review the CVT concept and a few of its generalizations and well-known properties. We then present an overview of recent advances in both mathematical and computational studies and in practical applications of CVTs. Whenever possible, we point out some outstanding issues that still need investigating. AMS subject classifications: 5202, 52B55, 62H30, 6502, 65D30, 65U05, 65Y25, 68U05, 68U10

Class and Cultural Division in the UK

Using data drawn from the Cultural Capital and Social Exclusion study, we examine the relationship between social class membership and cultural participation and taste in the areas of music, reading, television and film, visual arts, leisure, and eating out. Using Geometric Data Analysis, we examine the nature of the two most important axes which distinguish `the space of lifestyles'. By superimposing socio-demographic variables on this cultural map, we show that the first, most important, axis is indeed strongly associated with class. We inductively assess which kind of class boundaries can most effectively differentiate individuals within this `space of lifestyles'.The most effective model distinguishes a relatively small professional class (24%) from an intermediate class of lower managerial workers, supervisors, the self-employed, senior technicians and white collar workers (32%) and a relatively large working class which includes lower supervisors and technicians (44%).

Comparing literary worlds: An analysis of the spaces of fictional universes in the work of two US prose fiction debut cohorts, 1940 and 1955

This article presents a geometric data analysis (GDA) of the two “spaces of fictional universes” made up by the literary works of two cohorts of US prose fiction debut writers, 1940 and 1955. It is argued that the analysis reveals a common structure with significant variations that enable a comparison between two different states of a collective social imaginary, generated by 385 of the 560 authors included in the study. It is shown that the 1955 space is characterized by a greater contrast between agonistic and non-agonistic universes, and a firmer gendering of these oppositions. Other results include a confirmation of the generally accepted intuition that first novelists tend to reproduce features of their own social and geographical background in their works, but it is argued that different “site effects” enable a greater power to appropriate symbolically privileged spaces. The article concludes with a brief return to the individual authors, indicating the kind of research that the geometric data analysis makes possible.