Date: February 18-22, 2019
Every year Metrika organizes a Stata Winter School in Stockholm. These winter schools represent a unique opportunity for students, academics, and professionals to expand their skills in data management and data analysis and to learn how these skills can be applied to their own fields. All courses combine teaching and problem solving, and there are ample opportunities for participants to ask questions and to receive individualized guidance.
The 2019 Winter School is organized by Metrika Consulting providing an excellent set of courses taught by experienced Stata users and effective teachers of statistical methods:
- Stata Programming. Associate Professor Nicola Orsini, Karolinska Institutet (February 18)
- Regression Spline Analysis. Associate Professor Nicola Orsini, Karolinska Institutet (February 19 - 20)
- Econometric Methods of Causal Inference. Christopher F Baum, Professor of Economics and Social Work, Boston College (February 21 - 22)
Please click here to sign up for the Stata Winter School.
Stata Programming (February 18)
Stata is programmable and this means that you can easily write your own Stata programs and commands to share with others or to simplify your work using Stata's do-files, ado-files, and matrix-language program, Mata. Aim of this short course is to introduce you to the fundamentals of developing new Stata commands. It will demonstrate ways to program new simple procedural commands and more complex estimation commands.
Course outline
- Extending Stata
- From do-files to ado-files
- Background for ado-files
- Writing an ado-file
- Working with the syntax command
- Documenting a new command
- Writing your own help file
- Sharing programs and data
- Subroutines
- Programs in Stata
- Creating new commands
- Returning values from programs
- Programming estimation commands
- Programming your own estimation command
- Sharing your new command
- Standardizing and returning results
- Postestimation commands
- Using Mata in programs
- What is Mata?
- Examples of using Mata with maximum likelihood
- Linear/logistic regression using Mata
Instructor
Nicola Orsini is Associate professor of Medical Statistics at the Department of Public Health Sciences, Karolinska Institutet. Dr. Orsini has developed several Stata commands and contributed with several articles to the Stata Journal. He is a 2018 Highly Cited Researcher as ranking in the top 1% by citations worldwide. Dr. Orsini has over 15 years teaching experience of statistical courses and Stata software.
Regression Spline Analysis (February 19-20)
Aim of this course is to provide the fundamentals of spline analysis to model quantitative predictors (exposures, confounders, time) in relation to different kinds of outcome data (continuous, binary, counts) in common study designs (experimental, quasi-experimental, observational) based on either individual data or aggregated data (time-series, quantitative review, pooling project).
Participants will learn how to generate splines in line with specified research questions; how to use these splines in widely used regression models; and how to graph the results in a form suitable for publication in high impact factor scientific journals.
Course outline
- Moving beyond the linearity assumption for quantitative predictors
- Defining splines of different degrees
- Choosing number and location of knots
- Increasing stability using restrictions
- Fitting additive and multiplicative models (i.e. generalized linear models, quantile models, survival models) with regression splines
- Making statistical inference (confidence intervals, tests of hypotheses) for predicted responses or differences in predicted responses (i.e. mean differences, percentile differences, odds ratios, hazard ratios)
- Visualizing the spline analysis in high quality graphs suitable for publication
- Modelling multiple predictors with regression splines
- Interacting regression splines with other predictors
Instructor
Econometric Methods of Causal Inference (February 21-22)
In this two-day course you will learn how to estimate treatment effects using observational or non-experimental data. You will learn how and when to use Statas treatment-effects estimators and selected community-contributed commands to estimate average treatment effects (ATEs), treatment effects on the treated (ATET) and local average treatment effects (LATE). We will cover the theoretical basis of treatment effects estimators as well as many examples using Stata.
Topics covered
- Fundamental problem of causal inference
- Identifcation under random assignment
- Selection on observables and unobservables
- Regression adjustment
- Matching methods
- Propensity score matching
- Coarsened exact matching
- Reweighting on the propensity score
- Instrumental variables methods
- Difference-in-difference models
- LATE estimation
- Regression discontinuity designs
Prerequisites
Instructor
Christopher F Baum is Professor of Economics and Social Work at Boston College, and the author of An Introduction to Modern Econometrics using Stata and An Introduction to Stata Programming, Second Edition. He is the coauthor of several community-contributed Stata commands, such as ivreg2. Baum has lectured annually at the International Monetary Fund, several central banks and a number of universities on applied econometrics with Stata and Stata programming. He maintains the SSC Archive of community-contributed Stata packages. His current research has been focused on public health, including analysis of policies in?uencing tobacco and marijuana use and maternal mortality.
Logistics
The Winter School is held at Hotel Birger Jarl and they offer discounted accommodation for all course participants (please contact us for further details).
Please register for the courses you want to attend by sending us an email.
Attendance is limited and places are allocated on a first come, first serve basis. Please register long in advance to guarantee your place.
Stata 15 software is provided free of charge to all participants during the courses but participants are assumed to bring their own laptops.
Prices
Stata Programming
- Academic and student 435 USD
- Non-academic 600 USD
Regression Spline Analysis or Econometric Methods of Causal Inference
- Academic and student 870 USD
- Non-academic 1450 USD
Programming Stata and (Regression Spline Analysis or Econometric Methods of Causal Inference)
- Academic and student 1150 USD
- Non-academic 1730 USD
Regression Spline Analysis and Econometric Methods of Causal Inference
- Academic and student 1459 USD
- Non-academic 2020 USD
Programming Stata and Regression Spline Analysis and Econometric Methods of Causal Inference
- Academic and student 1900 USD
- Non-academic 2700 USD
Please observe
- All courses start at 9:00 and end at 17:00.
- These prices do not include Swedish VAT/moms and they will be charged in SEK as of the day's currency exchange rate according to Sveriges Riskbank.
- Faculty members and students must provide a proof of their current university affiliation at the time of booking (valid university email address for academics, and valid student ID card or official letter of enrollment for students)
- The price includes course materials, lunch and refreshments (you will need to notify us well in advance in case of special dietary requirements)
- These courses will be given in English.
Terms & conditions
- Only paid participants are guaranteed places in the courses.
- 100% of the fee is returned for cancellations made over six weeks prior to start of the course.
- 50% of the fee is returned for cancellations made three weeks prior to the start of the course.
- No fee is returned for cancellations made less than three weeks prior to the start of the course.