R.J. Mokken

Nonparametric Models for Dichotomous Responses

The development of nonparametric approaches to psychometric and sociometric measurement dates back to the days before the establishment of regular item response theory (IRT). It has its roots in the early manifestations of scalogram analysis (Guttman, 1950), latent structure analysis (Lazarsfeld, 1950), and latent trait theory (Lord, 1953).

Rejoinder to "The Mokken scale: a critical discussion

The nonparametric approach to constructing and evaluating tests based on binary items proposed by Mokken has been criticized by Roskam, van den Wol lenberg, and Jansen. It is contended that their argu ments misrepresent the objectives of this approach, that their criticisms of the role of the H coefficient in the procedures are irrelevant or erroneous, and that they fail to distinguish the inherent requirements (and limitations) of general nonparametric models and pro cedures from those of parametric ones. It is concluded that Mokkens procedures provide a useful tool for re searchers in the social sciences who wish to construct and evaluate tests for measuring theoretically mean ingful latent traits while avoiding the strong parametric assumptions of traditional item response theory.

Interpolation theorem for diameters of spanning trees

We give an algorithmic approach for transforming any spanning tree of a 2-connected graph into any other spanning tree of the graph. At each step of the transformation we obtain a spanning tree whose diameter differs from that of the previous tree by at most one. Thus if a 2-connected graph G has a as the minimum and b as the maximum diameter of a spanning tree, then for any integer c between a and b , graph G has a spanning tree of diameter c .

Testing Coalition Theories: The Combined Evidence

In multi-party systems, such as exist in many European countries, no single party is likely to acquire a majority on its own. For that reason parties must cooperate in coalitions that together control a majority of seats in the assembly. The question which parties are likely to find one another in such a coalition and which ones will remain outside, has occupied political scientists with some knowledge of the theory of games. A series of theories to explain parliamentary coalition formation has been proposed over the past ten years and these coalition theories have now been tested against data from European parliamentary systems.1 In one of these contributions Taylor and Laver (1973, p. 206) suggest that the other studies “… have been based on in appropriate or inadequate data or the statistical evaluation of the results has been inappropriate, incorrect or non-existent”. It is the purpose of this paper to investigate this suggestion and, more positively, to assess the combined evidence concerning the validity of particular coalition theories.

Collective judgement: combining individual value judgements

Abstract This paper addresses a fundamental problem of collective decision making: how to derive a collective value judgement from the individual value judgements of the members of a committee. Three structural conditions will be introduced, which correspond to certain consistency requirements for the collective judgement. It will be shown that the formula for the collective value judgement, based on these consistency conditions, is a quasilinear mean of the individual judgements, and moreover, that the generating function of the corresponding quasilinear mean is independent of the number of people in the committee. Some uniqueness properties are considered and, finally, it is shown that the quasilinear mean is suitable as a social choice function satisfying six Arrowian conditions.

Decision Context And Policy Effectuation : EU Structural Reform In Ireland

Abstract In 1988 the European Union reformed the allocation of its Structural Funds to effectuate broader participation of local and regional parties and institutions in the Member States. This might affect the corporative and sectoral oriented government structure in Ireland, which was chosen as a test case for this study, with decision issues for plan periods 1988–93 and 1994–99. In this paper we focus on the question whether actors with high effective power do also sustain relatively small utility losses with respect to decision outcomes. This is done in terms of a collective decision model simultaneously involving multiple actors as well as multiple issues, using linear models specifying the basic model variables of collective decision modelling. The results show that effective power does not have a direct effect in isolation, but shows strong effects in specific interaction with other factors, characterising the decision contexts as they evolve over time.

The circles of our minds. A nonparametric IRT model for the circumplex

A nonparametric item response theory (IRT) model for the circumplex is introduced, based on (1971) nonparametric IRT model for cumulative scaling, and (1984) nonparametric IRT model for unfolding. Some examples of circumplex representations in the social sciences are given. Model fit is based first on an extension of Loevinger’s coefficient of homogeneity using quadruples as elementary units of analysis. Diagnostics for the probabilistic circumplex are suggested. Assignment of (ordinal) scale values is based on the notion of item steps, as developed by (1983). The model presented here is based on dichotomous pick any/k data. Suggestions for extension to rank m/k data and to polytomous data are discussed.

European Union power and regional involvement: A Case Study Of The Political Implications Of The Reform Of The Structural Funds For Ireland

In 1988, the Reform of the Structural Funds of the European Union introduced new political structures, where one of the main objectives was to encourage participation of a much broader range of actors in the decision making process. In the present research, we limit our analysis to just one Member state, Ireland. As Ireland is a centralized, unitary state with a centralized semi-corporatist bargaining structure, we investigate how the Structural Funds Reform has been applied in this Member state. Did the Reform and to what degree, shift power away from Ireland’s traditional national centralised elite over two time periods corresponding to the two rounds of Structural Funds: 1988–1993 and 1994–1999? If so we want to know whether this has resulted in an increased involvement of a broader range of actors from the national and European union levels as well as the sub-national levels in Ireland, and under which conditions this occurred.

Density-based unsupervised classification for remote sensing

Most image classification methods are supervised and use a parametric model of the classes that have to be detected. The models of the different classes are trained by means of a set of training regions that usually have to be marked and classified by a human interpreter. Unsupervised classification methods are data-driven methods that do not use such a set of training samples. Instead, these methods look for (repeated) structures in the data.

Chapter 4 Legislative Analysis: Methodology For the Analysis of Groups and Coalitions

Publisher Summary In this chapter, a theory and methods are given that enable to analyze relations within and between groups and coalitions in decision-making bodies on the basis of actual behavior of the individual group members. The theory is suitable for the analysis of roll calls, interview questions with three response categories and preference rank orders. The methods and coefficients based on it provide large possibilities to compare the policy positions of decision-makers with respect to each other, to determine policy location and cohesion of groups of decision-makers, to indicate the location of group members relative to the group, and to search for blocs or cliques of decision-makers in terms of a certain minimal level of cohesion. In a wider context the theory and methods might be used to investigate the degree of consensus among political, social, and economic elites regarding basic values and norms; or they can be used in experimental designs to analyze the consequences of different independent variables on the cohesion of small groups.