Panel data analysis using Stata
[3 half days] - [English]
Are you conducting a longitudinal survey? In other words, have you collected data at different points in time, hence called longitudinal or panel data?
The mixed models are a family of models with both fixed and random effects. They make it possible to take into account the correlation between several observations, in particular the longitudinal aspect of the data.
Training aims
This course provides an introduction to the analysis of panel-data. At the end of the training, participants will be able to use Stata effectively for manipulating and analyzing panel-datasets, to understand the error components model and the issue of unobserved heterogeneity, and finally to implement most commonly used linear panel-data estimators.
Prerequisites
By registering for this training, you commit to a level of knowledge equivalent to the following training(s):
Content
- Working with panel-data: data management tools, summary statistics and dynamics.
- Linear panel-data models: pooled model, random-effects model and fixed-effects model.
- Post-estimation: tests for valid inference, comparison of estimators, attrition bias.
Rate
This training course is recognized by the IABE, enabling participants to earn CPD points.
(Note that this is true for all SMCS courses.)
Tools used during training
Stata
Methods and method families discussed
Data processing
Data set cleaning and processing for statistical analysis
Descriptive statistics
Descriptive statistics
Summary tables
Regression model
Multiple linear regression
Mixed models
Linear mixed model
Please log in to register as interested in this training.
It is necessary to create an account for a first use of our services.