Approximate Studentization with marginal and conditional inference

Many inference problems lead naturally to a marginal or conditional measure of departure that depends on a nuisance parameter. As a device for first-order elimination of the nuisance parameter, we suggest averaging with respect to an exact or approximate confidence distribution function. It is shown that for many standard problems where an exact answer is available by other methods, the averaging method reproduces the exact answer. Moreover, for the gamma-mean problem, where the exact answer is not explicitly available, the averaging method gives results that agree closely with those obtained from higher-order asymptotic methods. Examples are discussed; detailed asymptotic calculations will be examined elsewhere.     

A taxonomy of all ordinal 2 × 2 games

The 2 × 2 game is the simplest and most commonly employed representation of strategic conflict. The 78 strict ordinal 2 × 2 games have been used as conflict models extensively, and have been related in several different taxonomies. However, interest has recently focussed on the full set of 726 general ordinal games, in which one or both players may have equal preferences for two or more outcomes. This paper describes the development of a practical taxonomy of all 726 ordinal 2 × 2 games. The taxonomy provides for rapid identification of particular games, gives a convenient ordering, is as consistent as possible with previous work, and yet is not tied to any specific solution concepts. As well, definitions of several significant game properties are developed or extended to general ordinal games and applied in conjunction with the taxonomy.     

Narrating Love and Abuse in Intimate Relationships

via email: heather.fraser{at}rmit.edu.au Summary In this article, I use discourse analysis to explore the relationshipbetween love and abuse. I argue that Anglo-American social workhas been reticent to theorize love; and that when it does, theborders separating love from abuse are usually assumed to berelatively stable and readily apparent. After nominating someof the reasons for and repercussions of dichotomizing love andabuse, I suggest that knowledge about intimate abuse will expandif more attention is given to the study of love. I promote theuse of feminist discourses because they provide valuable insightsabout the operations of power in love relationships. Drawingideas from both structural feminism and post-structural feminism,I encourage social workers to move their analytical attentionback and forth between the micro-politics of intimate relationshipsto the wider cultural contexts within which they are staged.I conclude with the assertion that social workers will be betterplaced to understand the dilemmas that many ‘ordinary’women confront in their everyday lives, if they critically analysethe intersections of love and abuse.     

Conical tests: Observed levels of significance and confidence regions

Tests and confidence regions for a p-parameter nonnormal model can require integration in p dimensions, for example with a Bayesian or structural model. For small p, the computer integrations are manageable, see for example Fraser [3], Naylor and Smith [6], but for p beyond 5, 6, or 7, the integrations become unfeasible. This paper proposes conical tests of significance which involve manageable computer calculations. The conical tests also provide confidence regions, giving the confidence bound at a chosen level of significance in any direction from a central 0-confidence point.     

BUCKLEY'S SCHEME OF SCHEMES AS A FOUNDATION FOR TEACHING FAMILY SYSTEMS THEORY

In recent training literature, a consistent recommendation emerges concerning the need for an emphasis on the development of a strong system theory base upon which future practitioners and researchers can build. This paper advocates the use of Buckley's hierarchical scheme as a teaching tool to provide an understandable, yet broadly applicable, theory base for graduate students in an introductory family therapy course. A method of teaching Buckley's scheme is outlined that not only incorporates the recommendations of the literature, but also provides an umbrella model of systems for future application and theory building.     

STRUCTURAL AND STRATEGIC FAMILY THERAPY: A BASIS FOR MARRIAGE,OR GROUNDS FOR DIVORCE?*

In reaction to a current move toward eclecticism in family therapy, the Structural and Strategic schools of therapy are examined. Both a conceptual and practical framework for relating these two, often contradictory, schools to one another is presented through the use of some of Gregory Bateson's recently presented views. The therapist's dilemma in trying to integrate these two approaches is discussed in the light of self-fulfilling prophecies and the necessity for second-order change as related to the therapist's actions. A conclusion is reached that conceptual integration of the approaches may be possible at a higher level, yet practical integration of the two by the practitioner may be difficult, if not impossible, due to the different units of analysis and subsequent action commanded by each approach.     

Approaches to conflict modeling: A study of a possible USA-USSR nuclear confrontation

A possible nuclear confrontation between the USA and USSR is analyzed using five different models from game theory in order to compare the various approaches that can be taken in the analysis of conflicts. Of particular importance is an improved metagame analysis model that provides a comprehensive procedure for analyzing not only the nuclear conflict but also any other type of real world conflict that can arise in practice. When the improved metagame analysis model is linked with a new state transition model, the dynamics of the game can be thoroughly studied.