This talk will present the results of the direct assessment of students' numeracy when they enter university in order to discuss the effects of poor basic numerical and graphical skills on their statistical literacy. The context for this study is a Level-1 Statistics course for students in the Faculty of Psychology. The research findings being reported are part of a Master in Science carried out at the University of Glasgow in the academic year 2001/2002. Two antecedents justify this study. One is the growing interest, from the middle of the last century, in numeracy as a fundamental proficiency - together with literacy - which implies an effective use of mathematics to meet the general demands of life at home (Steen, 2001). The second antecedent of interest is statistical literacy, as an ability to interpret and critically evaluate statistical information and data-related arguments which students might encounter in diverse real contexts, to discuss their understanding of the meaning of the information and their opinion regarding the acceptability of the given conclusions (Gal, 2002). The development of statistically literate citizens is one of the main focuses of most current debates on the teaching of statistics. The recent discussion points to the need for education that is useful, enabling students to develop the ability to transfer their knowledge outside their school or university context. Towards that end, instruction in 'statistical literacy' is considered fundamental to prepare students to interact daily with quantitative situations, including many where quantitative information is embedded in text. The Sixth International Conference on Teaching Statistics (ICOTS6, South Africa, July 2002) was entirely devoted to a discussion on 'Developing a Statistically Literate Society'. What is common in all the topics of debate is the understanding that statistics is integral to an extensive range of functions of society and to many disciplines, and that society's statistical health depends on facilitating development of, and access to, a full and diverse range of statistical capabilities. Two Baseline tests, the Baseline Tests A and B, were administered at the beginning of the course to 171 and 194 students, respectively. With these tests we assessed students' numeracy, their confidence with fractions, proportions and percentages, and their ability to interpret simple tables and graphs. Two other tests - the Follow-up tests A and B - were administered near the end of the course to the same classes and students - respectively 139 and 141. With these tests we re- tested students' numeracy, we assessed students' mastery of statistical concepts, their ability to evaluate critically data-related arguments (e.g. newspaper reports) and to discuss their understanding of the meaning of the given information. Difficulty indices and discrimination rates were calculated for all test items; examples of easy and difficult items are given. The impact of numeracy (Baseline tests) and other background variables on the subsequent level of statistical literacy (Follow-up tests) has been explored through stepwise regression. Details of these analyses are discussed. This investigation suggests that statistical literacy is a hierarchical skill which builds upon a broad range of interlinked mathematical, statistical and probabilistic concepts. In accordance with a very recently published work on this subject (Watson & Callingham, 2003), this study confirms that the interaction of mathematical skills with data-related messages creates situations that only few students at the highest ability level can interpret successfully. The implications for future research and the opportunity for interventions which would make statistics more accessible and responsive to other disciplines and the needs of a broader citizenship are discussed. References Gal, I. (2002), "Adults' Statistical Literacy: Meanings, Components, Responsibilities," International Statistical Review, 70, 1-51. Steen, L.A. (Ed.) (2001), "Mathematics and Democracy: The Case for Quantitative Literacy," Washington, DC: National Council on Education and Disciplines. Watson, J., and Callingham R., "Statistical Literacy: A Complex Hierarchical Construct," Statistics Education Research Journal 2(2), International Association for Statistical Education, November, 2003.