ANOVA for unbalanced data: Use Type II instead of Type III sums of squares

By Ø. Langsrud. Statistics and computing, 2003

480 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Changes in bottlenose dolphin whistle parameters related to vessel presence, surface behaviour and group composition (2016, methodology): “Analysis of Variance type II (function 'Anova' in the ‘car’ library in R) was used to assess the significance of the predictors (Langsrud, 2003).”
Comparing Sexual Harassment Subtypes Among Black and White Women by Military Rank: Double Jeopardy, the Jezebel, and the Cult of True Womanhood (2008, methodology): “been recommended as more powerful than Type III sum of squares for ANOVAs with unbalanced data (Langsrud, 2003).”
The Geneva affective picture database (GAPED): a new 730-picture database focusing on valence and normative significance (2011, background): “biases due to unbalanced numbers of pictures in each category ( Langsrud, 2003 ).”
lmerTest Package: Tests in Linear Mixed Effects Models (2017, methodology): “We do not touch this topic here and refer to Venables (2000); Speed, Hocking, and Hackney (1978); Senn (2007); Langsrud (2003); Macnaughton (2009) for the discussions.”
Developmental variation in resistance and tolerance in a multi‐host–parasite system (2010, methodology): “(Langsrud 2003).”