ANALYSIS OF ECOLOGICAL DATA
cod. 18333

Academic year 2022/23
1° year of course - First semester
Professor
- Stefano LEONARDI
Academic discipline
Statistica per la ricerca sperimentale e tecnologica (SECS-S/02)
Field
Discipline del settore biomedico
Type of training activity
Characterising
48 hours
of face-to-face activities
6 credits
hub: PARMA
course unit
in ITALIAN

Learning objectives

The main objective of the course is to provide students with the
theoretical and practical knowledge to apply the correct scientific
method in ecology. Students learn to test various types of
models preparing a experiment to effectively analyzing and presenting
the data collected.

The interpretation emphasis of is always on the data in order to develop a critical concept of the ecological science, free from any bias.

The student will learn the theory and practical skills to analyze data
("Knowledge and understanding"), learns to apply them to new cases,
first during class and then independently ("applied knowledge and
understanding"), learns to draw his own statistically correct
conclusions based on data and assumptions and tested models ("Making
judgments") and finally learns how to graphically present the results
in an effective manner ("Communication skills"). Student's progression
towards autonomy also promotes the development of so-called "Ability
to learn."

Prerequisites

Students should have passed an introductory statistic course or a
elementary probability course during the previous degree.

Course unit content

# "Classic" Statistical tests with R

* The student's t
* Non-parametric tests
* Chi-square
* A permutation tests

# Analysis of variance in ecology

* Snova as interpretative model of ecological processes
* Preparation of the data matrix
* The function lm R
* Analysis and interpretation of statistical interaction
* Checking the assumptions
# The linear regression with matrices

* The model Y = Xb + e
* "Normal" equations
* Applications with R

# The ANOVA with matrices

* Applications with R
* Experimental Design

# Nonlinear fitting

* The nls function of R
* Choice of three different models

# Generalilzed Linear Models (GLM)
* Logist regression
* Poisson regression
* How to deal with over-dispersion

# Multivariate Statistics
* Principal components analysis
* Non Metric Multidimensional Scaling – nMDS
* Cluster Analysis


Full programme

See Contents

Bibliography

Online Lecture notes written by the teacher are available.

Teaching methods

The course has a strong practical content. All students will have
a computer and each lesson consists of a short
theoretical introduction followed by practical analysis of
ecological data guided by the teacher or a teaching assistant. Some
topics will be presented with a deductive approach to learn the
general law from computer simulations of different peculiar cases.

Assessment methods and criteria

The final grade is the result of the average of the score obtained by each
student with two different modes

- During the course, at the end of each lesson, the teacher assigns
homework to be done independently. Students
apply the notions to new situations in the classroom. By
specific questions, the student's critical thinking,
correct statistical interpretation of data, data support to the hypothesis
and the testing of assumptions are assessed.

- At the end of the course a practical examination is carried out in
student total verified autonomy. A simulation of a session analysis of data
obtained from one or more ecological experiments is carried out and students
present the results graphically in a correct and effective way.

Other information

- - -