The benefit to memory of spacing presentations of material is well established but lacks an adequate explanation and is rarely applied in education. This paper presents three experiments that examined the spacing effect and its application to education. Experiment 1 demonstrated that spacing repeated presentations of items is equally beneficial to memory for a wide range of ages, contrary to some theories. Experiment 2 introduced 'clustered' presentations as a more relevant control than massed, reflecting the fact that massed presentation of material is uncommon in education. The scheduling of clustered presentations was intermediate between massed and distributed, yet recall was no different than for massed. Experiment 3, a classroom-based study, demonstrated the benefit of distributed over clustered teaching of reading through modification of the scheduling of everyday lessons. Thus, the effectiveness of teaching may be improved by increasing the degree to which lessons are distributed.The benefit for memory of distributed (over massed) presentation of to-be-remembered material is robust. Spaced presentation of material leads to better retentions than does blocked presentation. However the principle of distributing learning has not been widely applied in education. Dempster (1988) argued that a major factor limiting such application is the lack of direct classroom-relevant demonstrations of its efficacy. Indeed, it is questionable whether or not the contrast between distributed and massed presentation of material is relevant to education at all. Massed presentation, in the sense of repeated presentations of to-be-learnt stimuli, or more generally, concentrated teaching of a topic in a single session, is rarely found in the classroom. Of more interest is the question of whether different degrees of distribution are influential in learning. The experiments reported in this paper begin to address the need to bridge laboratory studies (Experiments 1 and 2) and classroom-based studies (Experiment 3). At the same time a more important comparison than between massed and distributed presentation is investigated, namely between different degrees of distribution.

The revised measure corrects a basic error of failing to divide total income for the UK by number of people in the population. This alters the measure but is more correct. Further improvements result in a measure that correlates more closely with UK alcohol consumption over the last decade.

At least one generic self-discrepancy is related specifically to eating disorder rather than depressive symptoms. Implications of the findings for theory and clinical practice are briefly discussed.

The teaching of statistics to psychology undergraduates has traditionally included formulae and methods for conducting statistical tests without the aid of a computer. Does this requirement to master the formulae encourage deeper understanding or merely provoke anxiety? The relationship between competence in calculation and statistical thinking was examined using questions from an exam set at the end of a compulsory first year module. In addition, scores on a premodule test and coursework grades were recorded, and used to control for prior ability and learning effort. The control variables jointly accounted for 26% of the variance in statistical thinking, whereas calculation accounted for only a further 2%. The contribution made by competence in calculation, though statistically significant, was not large.

PerspectiveT he relationship between alcohol sales, alcohol consumption patterns, and levels of violence is well established. In a meta-analysis of data from seven countries, Jason Bond and colleagues estimated that the fraction of violence-related injuries attributable to alcohol is between 28% and 43% [1]. There is a stronger link between alcohol impairment and being a victim of violence than between alcohol impairment and suffering from accidental injuries [2].