The ubiquitous angle

Summary.  Previously we used the geometry of n -dimensional space to derive the paired samples t -test and its p -value. In the present paper we describe the 'ubiquitous' application of these results to single degree of freedom linear model hypothesis tests. As examples, we derive the p - and t -values for the independent samples t -test, for testing a contrast in an analysis of variance and for testing the slope in a simple linear regression analysis. An angle θ in n -dimensional space is again pivotal in the development of the ideas. The relationships between p , t , θ , F and the correlation coefficient are also described by using a 'statistical triangle'.     

Partner Violence and Mental Health Outcomes in a New Zealand Birth Cohort

This study examines the prevalence and extent of domestic violence and the consequences of domestic violence for mental health outcomes in a birth cohort of New Zealand young adults studied at age 25 years. A total of 828 young people (437 women and 391 men) were interviewed about the domestic violence victimization and violence perpetration in their current or most recent partner relationship. Key findings of the study were (a) domestic conflict was present in 70% of relationships, with this conflict ranging from minor psychological abuse to severe assault; (b) men and women reported similar experiences of victimization and perpetration of domestic violence; and (c) exposure to domestic violence was significantly related to increased risks of major depression (p < .05) and suicidal ideation (p < .005) even after extensive control for covariates.     

A scientist's viewpoint on promoting effective use of experimental design: Ten things a scientist wants to know about experimental design

This paper tackles the practical application of statistical tools. Chemical development is a rich area for collaboration between statisticians and chemists, but most statisticians encounter difficulties in promoting the tools to this client group. This paper, presented from a chemist's perspective, is intended to help bridge the gap between the two professions. The paper explores differences in uptake and attitudes towards adopting these techniques and suggests what has hindered or helped the process of effective application of experimental design. Copyright © 2003 John Wiley & Sons, Ltd.     

Structuralist Marxism and the labor process: Where have the dialectics gone?

Theory and Society -     

The bureaucratic side of computers: Memory,evocation and management information

Management information systems fail to support management because they are not flexible enough to keep pace with the unstructured nature of decision making processes. Observations made in four welfare management information systems suggest that computerization improves evocation—the means by which decision makers pool data from the collective memory, only to the extent that choices and the relevancy of data are predetermined. Facing rigid procedures and irresponsiveness to irregular data needs, managers of all ranks duplicate the computerized memory with manual records and files so as to retain access to their memory under their terms.     

Effects of alcohol on the nonverbal communication of anxiety: The impact of beliefs on nonverbal behavior

Three male and three female raters were shown videotapes of 29 male subjects who were speaking to a silent female confederate. Raters were asked to make 10 judgments about the subjects' behavior. The subjects had been assigned to one of two expectancy conditions in which they were led to believe the drink they consumed prior to meeting the confederate contained alcohol and tonic or tonic only. Actually, half of the subjects in each expectancy condition received alcohol and half received tonic only. Judgments of the female raters viewing the videotapes were congruent with physiological measures of the subjects' anxiety. Those subjects who believed they received alcohol were perceived by female raters as more relaxed, less anxious, less inhibited, and more dominant than subjects who believed they received tonic. The actual content of the drink had no significant effects on the raters' judgements of the subjects' behavior. Male raters were unable to discriminate among the experimental conditions. Implications for possible gender of rater effects in behavioral assessment procedures as well as for the impact of beliefs on nonverbal communication are discussed.