/explore

Click through on any links that interest you or select the planets on the right to continue exploring the Outer Web.
You are here

 statsandr.com
Multiple linear regression made simple - Stats and R
| | freerangestats.info
Stepwise selection of variables in regression is Evil.
3.4 parsecs away

Travel
| | Stepwise variable selection is bad and dangerous, and you shouldn't do it. It increases false positives. It drops variables that should be in the model. It gives biased estimates for regression coefficients. The problems are worse for smaller samples; higher correlation between the X variables; and models with weaker explanatory power for the y (i.e. lower R-squared).
| | aosmith.rbind.io
Simulate! Simulate! - Part 1: A linear model
2.9 parsecs away

Travel
| | Where I discuss simulations, why I love them, and get started on a simulation series with a simple two-group linear model simulation.
| | entropicthoughts.com
The Surprising Richness of Correlations
3.2 parsecs away

Travel
| | [AI summary] The text explores various statistical concepts, including correlation, variance, and partial correlations, using examples like income and room count, word count predictions by AI models, and programming language commit sizes. It discusses methods for calculating and interpreting these correlations, as well as the importance of considering degrees of freedom in statistical analysis. The article also references historical statistical methods and emphasizes the value of multiple perspectives in understanding statistical relationships.
| | freerangestats.info
Suppressed data (left-censored counts)
27.0 parsecs away

Travel
| I experiment with some different ways of handling counts in tables that have been suppressed for confidentiality, and come up in favour of multiple imputation. The mice R package helpfully lets you define your own imputation algorithm.