66756 - Experimental Design and Data Analysis

Learning outcomes

At the end of this course the student knows the principal concepts of experimental design to plan and to analyse results from field and laboratory experiments. The student should be able to discriminate natural variability and anthropogenic impacts on different marine populations and communities. Moreover he/she should able to manage data matrices by using uni- and multivariate statistical techniques.

Course contents

Experimental design: introduction to the scientific method; Sampling designs and methods; Components of the experimental design; Introduction to the analysis of variance as a method for a correct planning of the experimental design; Analysis of variance: one-way designs; hierarchical designs, factorial designs; designs to assess the environmental impact. Introduction to regression and correlation; Introduction to multivariate analyses; Multivariate data; distance, dissimilarity and similarity measures; Introduction to classification and ordination techniques; Statistical tests on multivariate data; Analysis of spatial and temporal patterns; Relations between biotic data matrices and environmental variables. Use of software packages for graphical analysis and statistical tests.

Prerequisites

Knowledge of basic statistics tools, with particular reference to probability and statistical inference theories. These topics will be covered in the Module 3 of the course FUNDAMENTALS OF SEA SCIENCES

Readings/Bibliography

• Copy of Power point lessons
• Experimental Design and Data Analysis for Biologists - Gerry P. Quinn and Michael J. Keough - Cambridge University Press
• Statistica di base - Lamberto Soliani. Ed PICCIN
• An Introduction to R - W. N. Venables, D. M. Smith and the R Core Team, https://cran.r-project.org/manuals.html

Teaching methods

The course consists of 6 credits, of which 4 credits of lectures and 2 credits, corresponding to 24 hours, of practical exercises in the computer lab.

The practical exercises are carried out using both the students' personal computers and the computers made available by the University. The aims of the practical lessons are to analyse data according to different theoretical approaches and to become familiar with software suitable for the analysis of different types of data and experimental designs. A specific module of 1CFU will be dedicated to the use of "R", a free software environment for statistical calculation and graphical analysis.

Both during the lectures and during the data analysis exercises the student-teacher interaction and the discussion between students is actively favoured.

All activities will be held in person according to pandemic and health situations.

Assessment methods

Verification of learning takes place through the final exam, which verifies the acquisition of the knowledge and skills expected by carrying out a written test lasting 2 hours without the help of notes or books. The exam aims to verify the achievement of the following objectives:

- in-depth knowledge of the statistical tools illustrated during the lectures

- ability to use these tools to analyse particular sampling designs

- ability to analyse both biotic and abiotic multivariate data matrices

- practical ability to use specific software packages to perform statistical analysis and graphic processing.

The exam is carried out in written form and requires an evaluation out of thirty. In addition to answering some questions concerning methods (in common language 'theory questions'), students will have to face and solve problems of setting up correct experimental designs in which they demonstrate that they know how to apply the tools acquired to estimate a particular model, verify particular assumptions and interpret the results obtained. All this in order to evaluate the acquisition of knowledge of research methodologies, analytical tools and data acquisition and analysis techniques.

Practical skills will be verified through a data analysis exercise to be carried out in R, in this case being able to use the example scripts provided in class and the system's online help. The vote of this exercise will contribute to the final vote. The grade assigned to the exam can be rejected up to a maximum of two times.

Teaching tools

Power point lessons; Practical exercises in the computer lab on real case studies in different marine habitats. The teaching material will be available in the dedicated virtual space (https://virtuale.unibo.it/).

Office hours

See the website of Marina Antonia Colangelo

See the website of Massimo Ponti