Half the village

In 1947 India became independent and at the same time lost a large part of its territory to the newly created state of Pakistan. This new political arrangement was achieved at a terrible cost in terms of human life and suffering. The years up to and immediately following 1947 provoked a debate about national and human identity. Pakistan was founded on the grounds that Muslims could not be safe or prosperous under majority Hindu rule in newly independent India. The new India insisted that she was a secular democracy in which all religious communities could enjoy equal status. The debate is reflected in a number of Hindi‐language novels which deal with this period. The present article is about one of them, written by a member of India's small Shi'ite Muslim community who was also a Marxist. The debate raises issues about human identity which we can now see as being tragically and importantly relevant for contemporary Europe. These issues are not only political but also religious in that they go to the heart of our understanding of what it means to be human beings. The article points to some of these issues and at the same time argues that the novel as a literary form provides a valuable and significant vehicle for their discussion.     

Paradoxical Results and Item Bundles

Hooker, Finkelman, and Schwartzman (Psychometrika, 2009, in press) defined a paradoxical result as the attainment of a higher test score by changing answers from correct to incorrect and demonstrated that such results are unavoidable for maximum likelihood estimates in multidimensional item response theory. The potential for these results to occur leads to the undesirable possibility of a subject’s best answer being detrimental to them. This paper considers the existence of paradoxical results in tests composed of item bundles when compensatory models are used. We demonstrate that paradoxical results can occur when bundle effects are modeled as nuisance parameters for each subject. However, when these nuisance parameters are modeled as random effects, or used in a Bayesian analysis, it is possible to design tests comprised of many short bundles that avoid paradoxical results and we provide an algorithm for doing so. We also examine alternative models for handling dependence between item bundles and show that using fixed dependency effects is always guaranteed to avoid paradoxical results.