Summer School in Advanced Methods for Data Analysis

Summer School in Advanced Methods for Data Analysis

Local: 
Salas Polivalente e 3
Date: 
09/07/2019 to 26/07/2019
Horário: 
09:30 - 17:30
Idioma: 
Portuguese, with possibility of using English
Credits: 
5 ECTS

This year we offer three modules that aim to maintain the pedagogical standard of previous years. In this way, the modules will have a theoretical component accompanied by a strong practical one in which the participants will have the opportunity to develop skills in data analysis using MPlus and R project.

Participants are invited to work on their research problems using their own data and to develop their own analytical models.

 

Intended for

The Summer School in Advanced Methods is intended for all those interested in updating and / or deepening their knowledge in quantitative methods applied to the social sciences. Thus, the courses we propose this year are intended for postgraduate students, researchers, university professors, and public and private administration staff.

Course I | Multilevel Analysis with MPlus | 9th to the 12th July Multipurpose Room

Cicero Pereira (Federal University of Paraíba) | PhD in Social Psychology by the University Institute of Lisbon (ISCTE-IUL). His thesis elaborated on the relationship between prejudice and discrimination in anti-prejudice contexts. He developed postdoctoral studies at the Institute of Social Sciences of the University of Lisbon, where he was until February 2015, as an assistant researcher. He taught statistics at the Faculty of Psychology at UL. He is currently Professor of Methods of Social Psychology and Advanced Methods of Data Analysis at the Federal University of Paraíba. It has several publications using multilevel analysis.

Course II | Multivariate Analysis (with R) 16 to 19 July Room 1

Sérgio Moreira (FPCE-UL) | PhD in Social Psychology by ISCTE-IUL. Invited statistics professor at the Faculty of Psychology of UL and private consultant in methods of investigation and analysis of data in social sciences. Currently, his professional activity is divided between university education, coordination, consulting projects and participation as a researcher in academic projects.

Course III | Big Data (with R) | 23 to 26 July Room 3

Cláudia Abreu Lopes (POLIS, University of Cambridge) | Social Psychologist, Doctorate in Research Methods, London School of Economics. She lectures Research Methods and Statistics at the Department of Politics and International Relations at the University of Cambridge and is Director of Research at the Africa's Voices Foundation start-up. Associate Researcher at ICS.

Course I | MultiLevel Analysis with MPlus | 9th to 12th July Multipurpose Room

Day 1

Introduction to Data Analysis

Challenges in the analysis of hierarchical data;

Basic types of hierarchically structured data;

Regression models (ANOVA and ANCOVA) and Multilevel Models.

Introduction to data analysis in Mplus

Mplus environment and language;

Specification and estimation of regression models.

Output Interpretation.

Day 2

The meaning and centrality of the intercept

The Null Model: Specification and estimation methods;

The interpretation of the intercept;

Intraclass correlation (ICC);

Output Interpretation

The Meaning of regression coefficient

Multilevel analysis with individual predictors;

Multilevel analysis with contextual predictors;

Estimation of variance explained at each level of analysis;

Output Interpretation.

Day 3

The meaning of the interaction effect

Analysis of multilevel interaction effects;

interaction interpretation: simple slopes and conditional effects;

Interpretation of outputs.

Analysis of multilevel interaction effects;

The Multilevel Approach in Repeated Measure Drawings

The structure of the database;

The specification and estimation of the model;

Output Interpretation .

Day 4

Discussion of models prepared by participants

Model Specs

Outputs interpretation

Alternative analysis

Course II | Multivariate Analysis (with R) 16 to 19 July Room 1

Day 1

The comparison model approach

Revision of basic R. Data, model and error statistical concepts. Model I: continuous DV and dichotomous IV (linear regression with a dummy and t-test)

Creation and management of a database in R. Numerical and visual descriptive statistics with R. Execution of the model I.

Day 2

Models with one IV

Model II: no VDs (correlations) and with continuous DVs and IVs (linear regression). Model III: continuous DV and categorical IV (linear regression with two dummies and one-way Anova)

Execution of models II and III and their graphical representation.

Day 3

Models with more than one IV

Model IV: continuous DV and two categorical IVs (multiple regression with interaction terms, contrasts, simple slopes and Anova factorial).

Model V: continuous DV, a categorical IV and a continuous IV (multiple regression with terms of interaction, contrasts, simple slopes and Ancova factorial)

Execution of model IV and V and their graphical representation.

Day 4

Parametric analysis 101

Extreme cases (outliers). Parameters normality and errors. variances Homogeneity. Multicolinerity.

Diagnostics and graphical representation.

Course III | Big Data (with R) | 23 to 26 July Room 3

Day 1

Conceptual and ethical issues in big data applications

What is big data; types of big data; what is it for? examples of applications; epistemological and ethical issues.

Collection and manipulation of digital data (Twitter and SMS)

Day 2

Introduction to big data preparation and visualization

Steps for processing big data; procedures for checking missings and detecting outliers; coding text using regular expressions; stringr R introduction.

Large data processing and visualization using Tableau and R (ggplot and ggraph).

Day 3

Big data analysis techniques: statistical / machine learning

Problems in grouping and classifying; introduction to R Data Miner.

Big data analysis through machine learning methods. Group project Preparation.

Day 4

Database Integration (e.g., surveys and big data)

Applications and examples of organic databases integration with surveys.

Presentation of the participants group projects.

 

Requirements

The admission requirements depend on the attended module, but in case it is an advanced course, a prior knowledge of the basic statistical principles for the social sciences is recommended.

The modules are independent and each student can attend any of them. Each course lasts for one week, from Tuesday to Friday.

Applications

Tuition:

Registration until 05/16/19: 1 course € 125; 2 or more courses 100 € per course
Registration after 05/16/19: 1 course € 150; 2 or more courses € 125 per course

Registration fee: € 30 (deducted from tuition)
The ICS-ULisboa Community: 50% off Tuition

The fee paid upon registration is non-refundable

The school will only be held if the necessary quorum is reached.

Coordenador