Thematic Area: Advanced Statistical Methodologies: Exploratory and Confirmatory Factor Analysis, Structural Equation Models, General, Mixed, and Generalized Linear Models
Duration: 20 hours of lectures
Start and End Dates: January - February 2024
Teaching Methods: Lectures and practical exercises
Instructor: Dr. Alessandro Catini
Program: This course focuses on more complex and general models used in psychometric statistical analysis. It starts with an introduction to factor analysis techniques, both exploratory and confirmatory, using IBM SPSS Statistics software. Then, with the support of SPSS, models from the GLM family (Univariate, Multivariate, and Repeated Measures) are introduced, accompanied by theoretical conceptual explanations and examples of real applications to psychometric data. The course then covers generalized linear models, with particular attention to selecting the Link function depending on the model being constructed and interpreting the statistical software output. Finally, structural equation models (SEM) are introduced through IBM’s AMOS software, covering both conceptual understanding and examples of structuring and interpreting software results. By the end of the course, students will have a solid grasp and substantial autonomy in implementing advanced statistical models suitable for various situations.
Final Assessment Method: Practical exercise
Thematic Area: Statistical Analysis Techniques via Artificial Intelligence Algorithms (e.g., Machine Learning)
Duration: 24 hours of lectures
Start and End Dates: December 2023
Teaching Methods: Lectures and practical exercises
Instructor: Dr. Merylin Monaro
Program: The course begins with an introduction to artificial intelligence and machine learning, covering the following topics: definition of machine learning, types of learning (supervised, unsupervised, reinforcement learning), types of tasks (regression, classification, clustering), metrics for evaluating algorithm performance (e.g., accuracy, AUC, sensitivity, specificity), types of classifiers (e.g., decision trees, SVM, logistic regression, KNN, etc.), feature selection techniques, and model validation techniques (e.g., k-fold cross-validation). The concepts are reinforced through practical exercises applying machine learning techniques to data collected in psychological and forensic contexts. These exercises involve learning and using the WEKA software. The second part of the course focuses on practical applications of machine learning in neuroscience, psychology, and particularly in forensics (e.g., using machine learning and natural language processing for lie detection, identifying high-risk personality profiles, brain and mind reading).
Final Assessment Method: The final assessment consists of presenting a scientific article where machine learning is applied to a research area of interest to the student, or presenting the analysis process conducted on data held by the student using these techniques.
Thematic Area: Methodologies for Conducting Meta-Analysis
Duration: 20 hours of lectures
Start and End Dates: January - February 2024
Teaching Methods: Lectures and practical exercises
Instructor: Dr. Giovanna Parmigiani
Program: The course aims to provide doctoral students with the skills to: i) understand international guidelines for producing and updating meta-analyses; ii) formulate a research question, define eligibility criteria for studies to include, design the bibliographic search, select studies, extract data, assess the quality of included studies, produce Summary of Findings tables, develop statistical analysis, and present data; iii) draft and register a meta-analysis protocol and acquire the methodological knowledge necessary to write an actual meta-analysis.
Final Assessment Method: Practical exercise