1 / 11

Insights into Matched Models: Interactions, Additive Models, and Diagnostics

In this lecture, we explore advanced concepts in matched models, focusing on interactions with matching variables and the formulation of additive models. We discuss the potential pitfalls of ignoring matching effects and emphasize the importance of diagnostics for ensuring model efficiency. Additionally, we delve into the challenges of fitting additive models in software like Stata, highlighting the necessity of specialized packages for accurate statistical analysis. Understanding effect modification by matching variables is critical for robust data interpretation and model accuracy in biostatistics.

gazit
Download Presentation

Insights into Matched Models: Interactions, Additive Models, and Diagnostics

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Lecture 13 – More on Matched Models • Outline • Interactions with the matching variables • Additive models • Diagnostics • Efficiency BIOST 536 Lecture 13

  2. Leisure world data • Neither ignoring the matching nor fitting individual αj for each matched set is a good idea BIOST 536 Lecture 13

  3. Effect modification by the matching or stratifying variables BIOST 536 Lecture 13

  4. Interactions with the matching variables BIOST 536 Lecture 13

  5. Interactions with the matching variables BIOST 536 Lecture 13

  6. Additive models BIOST 536 Lecture 13

  7. Additive models • As is often the case, the interaction term allows OR ( A=1 & B=1) < OR(A=1) x OR(B=1) • Suggests an additive model without interaction might be as good • Stata cannot fit the additive model without writing a special routine; some specialized packages (Egret; Epicure) can do these model fits BIOST 536 Lecture 13

  8. Additive models • Additive models can be attractive scientifically, but in practice are difficult to fit and have poor statistical properties BIOST 536 Lecture 13

  9. Diagnostics BIOST 536 Lecture 13

  10. Diagnostics BIOST 536 Lecture 13

  11. Efficiency BIOST 536 Lecture 13

More Related