Pitfalls in (Multiple) Regression Analysis

Jos W. R. Twisk
Additional contact information
Jos W. R. Twisk: Amsterdam UMC, Department of Epidemiology and Data Science

Chapter Chapter 11 in Basic Principles of Applied Medical Statistics, 2025, pp 235-246 from Springer

Abstract: Abstract In this chapter, two important pitfalls in (multiple) regression analysis are discussed. Collinearity occurs when two independent variables are highly correlated with each other. This can lead to huge standard errors and sometimes to a so-called sign switch. Mediation analysis is an important tool to investigate causal mechanisms. Within mediation analysis, a distinction is made between the total effect, the direct effect and the indirect effect. The problem with mediation analysis is that it is basically the same as the analysis of confounding. The only difference is the interpretation of the results of the analysis. It is also shown, that there is a pitfall within the pitfall of mediation analysis, i.e. the estimation of the standard error of the indirect effect.

Date: 2025
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:sprchp:978-3-031-86278-6_11

Ordering information: This item can be ordered from
http://www.springer.com/9783031862786

DOI: 10.1007/978-3-031-86278-6_11

Access Statistics for this chapter

More chapters in Springer Books from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().