The course aims to present the basic tools for analysing statistical data relating to the most common experimental designs in the field of Natural Sciences. The fundamental purpose is to allow the student to carry out autonomous calculations and to correctly interpret and present the results of these elaborations. At the end of the course the student is expected to know: - the main types of statistical variables and how to suitably represent them (frequency tables, mean, median, mode, variance, standard deviation, standard error); - the main statistical reference distributions (Gauss, Poisson, binomial, Bernoulli); - the assumptions and the meaning of a statistical test (hypothesis test, null hypothesis, significance); - the main parametric statistical tests with one or two variables (Chi-square, t-test, F-test, binomial test); - the theoretical basis of linear models (ANOVA, ANCOVA, simple regression, multiple regression); - the basic syntax of programming in R environment (script). And the student should be able to: - identify the variables associated with a simple experimental design; - represent graphically and through adequate descriptive statistics a sample variable in the R environment; - identify the correct statistical analysis to be applied to a specific case study and perform it in the R environment; - report and interpret the results of a statistical analysis.
Course Prerequisites
- Basic knowledge of mathematics (numerical sets, elements of literal calculus, linear algebra, mathematical analysis, elements of probability calculus) - Basic skills in using a personal computer (file manager, internet browser, text editor, application installation)
Teaching Methods
- Frontal lessons with the aid of slides and shared writing through the panel board of the R scripts associated with the lesson; - Laboratory lessons aimed at solving problems based on simulated cases and of increasing difficulty throughout the course; - Exercises on the KIRO platform with requests and commented solutions that retrace the fundamental stages of the course; - Guided classroom exercises to carry out exam topics; - Use of the KIRO platform for sharing all the reference material used and/or produced during the course.
Assessment Methods
Practical exam (3 hours) based on problem solving: some selected case studies are proposed to each candidate, in order to assess the ability to correctly choose and use the adequate statistical techniques.
Texts
Soliani L., Siri E. e Sartore F., 2005. Manuale di statistica per la ricerca e la professione: statistica univariata e bivariata, parametrica e non parametrica per le discipline ambientali - Edizione Aprile 2005. Uninova, Parma. Freely available at http://www.dsa.unipr.it/soliani/soliani.html Crawley MJ. 2007. The R Book. 1st Edition. John Wiley & Sons, Ltd. ISBN 9780470510247 Zuur AF, Ieno EN & Meesters E. 2009. A Beginner’s Guide to {R}. Springer. ISBN 9780387938363 Fowler, Cohen 2010. Statistica per ornitologi e naturalisti. Franco Muzzio Editore. ISBN 9788874132225
Contents
The course will focus on a small set of techniques, but widely used in practice. In particular, the synthesis methodologies of a variable (frequency distributions, averages, variability indices), the study of the correlation between two quantitative variables, the fit of a regression line and linear models will be examined. The logical foundations and the cognitive purposes of each technique will be illustrated, while the technical details and mathematical derivations will be placed in the background. The course will be based on the use of the R statistical software.