Learning objectives
The course of Statistics for experimental and technological research is designed to introduce the student to the basics of statistical thinking and its application in practice. The topics are geared to concrete problems of analysis and research and deal in particular with situations and cases drawn from the medical literature.
The course aims to give students the statistical tools needed to describe and analyze the data, extract useful information and make informed decisions.
Particolar efforts will be put on the principles of experimental design, power analysis and sample size determination.
Special emphasis will be given to statistical reasoning, interpretation and decision-making process. We will insist more on the conceptual understanding that the mechanical calculation, especially in light of the wide range of software available for analysis. The theory will be made explicit by means of practical exercises and teaching cases, therefore, the ultimate goal of the course is that the student learn "how to do" as well as knowing.
Prerequisites
- - -
Course unit content
The first part of the course will introduce the basics of statistical planning and experimental design.
Principles of probability and combinatorial analysis needed later in the course will be recalled, as well as the major probability distributions. Elementary descriptive statistics will be reviewed.
The second part of the course will review the principle of univariate statistical inference.
Sampling distribution. Type I and II errors. Power of a test and operating curve.
Parametric test : Student t-test, analysis of variance. Non-parametric test: Wilcoxon test, Mann-Whitney test, Kruskal-Wallis test, Friedman test, median test, Chi-square test, Fisher’s exact test.
The final part of the course will be an overview of multivariate statistics. Principal component analysis, Discriminant analysis, Classification and cluster analysis.
Practical sessions with the statistical software R and SPSS on medical data.
Full programme
Design and analysis of experiments.
Power analysis, effect size and sample size calculations.
Probability calculus, combinatorial analysis and special probability distributions.
Review of univariate statistics and elementary descriptive statistics.
Review of general principles of univariate statistical inference. Sampling distribution. Hypothesis and hypothesis testing. Type 1 and type 2 error. Power of a test and operating curve.
Parametric test : Student t-test, ANOVA and repeated measures ANOVA.
Non-parametric test: Wilcoxon test, Mann-Whitney test, Kruskal-Wallis test, Friedman test, median test, Chi-square test, Fisher exact test.
Multivariate Statistics
Overview of multiple regression, logistic regression, principal component analysis, classification and cluster analysis.
Practical session with the professional statistical systems "R" and "SPSS" on medical data.
Bibliography
1) Lecture notes
2) M.J. Crawley "The R book" Ed. Wiley
3) Internet resources and links
Teaching methods
During classroom lectures, the topics contained in the program of the
module will be illustrated and commented.
At the end of each topic classroom exercises explaining the application of the theory in practice will follow. The formal procedure and the step by step execution of the necessary calculations will be described. Both manual solution and computer calculation will be shown.
The students will be particularly encouraged to use the open source statistical system "R" and the free software package Epi Info.
Assessment methods and criteria
The achievement of the objectives of the module will be assessed
through a written examination, mainly consisting in open questions on the
topics of the course. This will allow to ascertain the knowledge and the
understanding of both the theoretical bases and their consequences.
The written examination will include the resolution of problems, to assess the
achievement of the ability to apply the acquired knowledge to a
simulated biological or medical situation.
All parts of the written exam will be equally weighted in the final
evaluation.
Other information
- - -
2030 agenda goals for sustainable development
- - -
