Timeline for How to interpret type I, type II, and type III ANOVA and MANOVA?
Current License: CC BY-SA 4.0
7 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 19 at 14:03 | comment | added | svendvn | @Quertiopler, I tried answering the question, but I can't really make things add up. I am starting to doubt the orthogonality requirement of o(A,B). It turns out that the design matrix columns are not orthogonal per se, but something about the variables should be orthogonal. If you find the answer, please let me know (for example by answering your question) | |
| Apr 18 at 9:24 | comment | added | Quertiopler | I asked a new question: stats.stackexchange.com/questions/645290/… Feel free to answer | |
| Apr 18 at 8:25 | comment | added | svendvn | @Quertiopler, I feel that is too much for the comment section - I encourage you to ask a new question specifically for this | |
| Apr 17 at 12:21 | comment | added | Quertiopler | Thanks for the clarification, but I am afraid that I can't follow. Would you mind providing a (simple) example of a design matrix with factors A and B and show how to obtain o(A,B)? | |
| Apr 17 at 7:21 | comment | added | svendvn | The model line is a human-friendly interpretation of the design matrix. In the full model, the linear subspace of the design matrix spans all combinations of A and B. If the design matrix is such that some columns span the linear subspace of A, and some other columns, the linear subspace of B, and so that the remaining columns are orthogonal to both the linear subspace of A and the linear subspace of B, then those remaining columns are o(A,B). | |
| Mar 12 at 15:28 | comment | added | Quertiopler |
Could you elaborate a bit more on how the operator o(A,B) is defined?
|
|
| Mar 3, 2023 at 16:58 | history | answered | svendvn | CC BY-SA 4.0 |