2024
Experimental Design
Name: Experimental Design
Code: MAT10167M
6 ECTS
Duration: 15 weeks/156 hours
Scientific Area:
Mathematics
Teaching languages: Portuguese
Languages of tutoring support: Portuguese
Sustainable Development Goals
Learning Goals
In this course we study the most current models considered in Experimental Design. These methods attempt to give students a broad education and current methods in Experimental Analysis.
The use of statistical software and analysis allow the treatment of databases.
Objectives: To study basic principles and concepts of experimental design and to provide students with essential statistical tools in the analysis of several experimental designs.
Competences:
To develop student?s capacity to selected and organize information, in a critical way.
To understand some fundamental statistical tools of experimental designs.
To select the appropriate regression model and to know how to validate the model.
To validate different parametric approaches based on their objectives.
To recognize and to know how and when to apply nonparametric approaches.
To stimulate autonomous learning and adaptation to new situations.
To properly use the software.
The use of statistical software and analysis allow the treatment of databases.
Objectives: To study basic principles and concepts of experimental design and to provide students with essential statistical tools in the analysis of several experimental designs.
Competences:
To develop student?s capacity to selected and organize information, in a critical way.
To understand some fundamental statistical tools of experimental designs.
To select the appropriate regression model and to know how to validate the model.
To validate different parametric approaches based on their objectives.
To recognize and to know how and when to apply nonparametric approaches.
To stimulate autonomous learning and adaptation to new situations.
To properly use the software.
Contents
Scientific method and experimental design.
Analysis of variance models: fixed effects (single and multiple factor), random effects (single and multiple factor) and mixed effects.
Split-plot and nested designs.
Multiple comparisons.
Complete and incomplete block designs. Latin square designs.
Non-parametric approaches.
Simple linear regression model and multiple regression model (estimation, inference, prediction, model adequacy and validation). Diagnostics for influence points, outliers,
multicollinearity and autocorrelation. Model selection.
Analysis of Covariance.
Nonlinear Regression.
Analysis of variance models: fixed effects (single and multiple factor), random effects (single and multiple factor) and mixed effects.
Split-plot and nested designs.
Multiple comparisons.
Complete and incomplete block designs. Latin square designs.
Non-parametric approaches.
Simple linear regression model and multiple regression model (estimation, inference, prediction, model adequacy and validation). Diagnostics for influence points, outliers,
multicollinearity and autocorrelation. Model selection.
Analysis of Covariance.
Nonlinear Regression.
Teaching Methods
Theoretical-practical lessons mainly lectured with a blackboard, with e-learning tools, and transparencies.
Introduction to theoretical concepts appealing to different areas of applications to illustrate the importance of course contents. Exercises with emphasis in the resolution of real problems, to motivate interest in the course and to demonstrate its utility.
To stimulate individual and group participation within the classroom and at home.
To emphasize the critical analysis and interpretation of data, appealing to software outputs as much as possible.
Introduction to theoretical concepts appealing to different areas of applications to illustrate the importance of course contents. Exercises with emphasis in the resolution of real problems, to motivate interest in the course and to demonstrate its utility.
To stimulate individual and group participation within the classroom and at home.
To emphasize the critical analysis and interpretation of data, appealing to software outputs as much as possible.
Assessment
To privilege continued evaluation carrying out one test (50%) plus two individual/group homework projects (50%). If continuous evaluation is not feasible for the student, a final examination (75%) is possible, but the individual / group project is still required although with lesser weight for final grade (25%). In the test, the student must have a grade greater than 8.
Teaching Staff (2023/2024 )
- Dulce Gamito Santinhos Pereira [responsible]
