Multivariate Data Analysis

Objectives

After this unit, students should be able to:

  • Identify characteristics of a problem to use the appropriate multivariate statistical methods;
  • Understand data limitations and verify the assumptions of statistical techniques;
  • Perform the most appropriate analysis using a statistical package (e.g SPSS) and extract the pertinent or relevant information from the output provided by the software.
  • Improve critical interpretation of statistical analysis of data in medical research.

General characterization

Code

12396

Credits

4.0

Responsible teacher

Isabel Cristina Maciel Natário

Hours

Weekly - Available soon

Total - 41

Teaching language

Português

Prerequisites

Basic notions of Algebra and Analysis and intermediate level notions of Probability and Statistics.

Bibliography

Branco, J.A. (2004) Uma Introdução à Análise de Clusters. Sociedade Portuguesa de Estatística.

Hosmer, D.W.; Lemeshow, S. (1991) Applied Logistic Regression. John Wiley & Sons. 2ª Edição.

Hair, J.F.; Anderson, R.E.; Tatham, R.L.; Black, W. C. (2005) Análise Multivariada de Dados. Bookmam. 5ª Edição.

Harrell, F. E. (2001) Regression Modeling Strategies with Applications to Linear Models, Logistic Regression, and Survival Analysis. Springer.

Maroco, J. (2014) Análise estatística com o SPSS. Edições Silabo. 6ª Edição.

Teaching method

Theoretical and practical classes will be presented using e-learning platform of IHMT and other internet resources. Practical classes and tutorials include dataset analysis, using SPSS program or similar package. Videos and written information will be provide, as well articles for discussion.

Evaluation method

The final assessment will be combined a written exam (60%) and an individual written work (40%). The exam includes different type of questions (e.g. multiple choice, true/false and essay questions). The written work is the result of the statistical analysis of a database, using a statistical package.

Subject matter

Assumptions and interpretation of parameters, assessment of model fit and validation of the fitted models. Practical considerations in health.

Simple and multiple logistic regression versus linear regression: model assumptions, interpretation of parameters and validating the fitted model.

Multidimensional contingency tables: Chi-Square test and log-linear analysis.

Analysis of variance and Linear Regression

Cluster Analysis and Factorial Analysis

Applications using SPSS Program: assumptions, model fitting, variable selection, diagnostic tools and model validation

 

Programs

Programs where the course is taught: