Instructor: Levente Littvay (TK PTI)
Regression
Gain a deeper understanding of regression analysis through this intensive workshop covering the foundations: correlations and simple linear regression, multiple regression, and advanced regression techniques. We will build regression analysis from the ground up through hands-on calculation of correlations and bivariate regressions to ensure a strong foundation. The assumptions underlying regression modeling, including linearity, normality, homoscedasticity, and lack of multicollinearity (things you may be conveniently ignoring when running regression models) will be thoroughly examined to equip participants to apply and interpret results properly. Advanced topics such as general linear models, weights, and models beyond regression basics will be covered to provide you with the skills needed to tackle complex research problems. Join us for this comprehensive workshop and master regression analysis for advanced research and data science applications.
Some modeling examples will be provided in R. No pre-requisite R knowledge is required, although if you want to use what you learned, it is useful if you know how to work with data in R.
Sessions will be in English.
Pre-requisites: The workshop is targeting people who have some statistical foundations. Either you are using statistical techniques but feel you are missing important foundations. Or you had an introductory course that has been too long ago and need a review. It is possible to come to this session and learn everything from scratch, but if you are in this situation, email me beforehand as soon as possible, and I will help you get the most out of the workshop. Otherwise, all I will ask of you is to remember basic algebra, addition, subtraction, multiplication, and division, know the order of operations, and put up with the occasional square root.
-
Day 1: Regression foundations
Summary: On this day, we will review the foundations necessary for regression. We discuss the most important aspects of statistical inference, sampling, and central limit theorem; we learn what is behind correlation and regressions.
Who is this session for? If you have been using statistics but feel you are missing important foundations. If you work with correlations and regressions but never calculated them by hand, reviewing the basics would be a very good idea. Don’t worry, I will go easy on you. If you know how to add, subtract, multiply, and divide, you know everything you need to know for this session. (OK, we may utilize the square root once or twice as well).
Learning outcome: After this session, you will have a solid understanding of what this p-value thing is and what is behind the numbers when you are looking at a correlation or a regression.
-
Day 2: Multiple Regression and Regression Assumptions. Regression foundations
Summary: On this day, we review regression assumptions. Thinking through this will not only help you do regressions right, it will help you become a better researcher, be more thoughtful about your research designs, and recognize interesting patterns in your data going beyond relationships typically found with regressions.
Who is this session for? If you have been running regressions but feel you may have been ignoring things you know you shouldn’t, this session is for you.
Learning outcome: After this session, you will know the assumptions you make when running regressions and help you diagnose if you are meeting these assumptions.
-
Day 3: Advanced Topics in Regression Analysis
Summary: In this session, I plan to introduce general linear models and regressions with dependent variables that are not normally distributed and continuous. Beyond that, I plan to tailor the session to the interests of the audience. Topics will include introductions of more advanced techniques and solutions to problems associated with assumption violations from day 2.
Who is this session for? If you want to learn more about regression.
Learning outcome: After this session, you will know exactly what tools to reach for and potentially learn as you explore more advanced topics in inferential statistics appropriate for your analyses.
The event is available via the link!