BIOSTATISTICS SCOPE: The course introduces statistical concepts and methods that are commonly applied in medical research. It aims to provide students with the necessary skills to a) interpret and critically appraise the results of data analyses published in the biomedical literature and b) perform standard statistical analyses. Each session consists of a lecture followed by a practical in the computer lab. Pen & paper exercises are given during each session. The concepts introduced in each lecture are illustrated in the corresponding practical using the statistical package SPSS and certain web Applets. The focus is on communicating the knowledge necessary to decide on the appropriateness of particular statistical tests under different circumstances e.g. with non-normality, small samples, paired data etc. Standard statistical topics such as types and presentation of data, point and interval estimation, hypothesis testing, sample size estimation, correlation & regression will be covered. Multiple linear, logistic and Cox regression techniques, and thus the concepts of statistical confounding and interaction, will be discussed. READING MATERIAL: 1.Notes & references provided by the lecturer. 2. "Medical Statistics At a Glance 2nd ed" Α. Petrie & C Sabin. Blackwell Publishing, Oxford, 2005. (short textbook & accompanying MC-type exercises & solutions on http://www.medstatsaag.com) 3. "Statistical Methods in Medical Research (4th edition)". Armitage P, Berry G, Matthews JNS. Oxford: Blackwell Science 2002. EXAM: Written exam (multiple choice and questions with short answers) TOPICS: Descriptive statistics: summarizing & presenting data. An introduction to estimation and hypothesis testing. Univariate numerical data analysis: t-tests for independent and paired samples, nonparametric tests, ANOVA. Univariate categorical data analysis: comparing two proportions, the chi-squared test, Fisher’s exact test, McNemar’s test for paired proportions, odds ratios. Estimating relationships between two characteristics: parametric and nonparametric correlation analysis, simple linear regression. Extension to >2 variables: multiple linear regression & logistic regression. Survival analysis: Kaplan-Meier survival curves, log-rank test, Cox regression. 1