Timeline for Number of contrasts in an orthogonal contrast test
Current License: CC BY-SA 4.0
9 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 20, 2022 at 18:10 | comment | added | Russ Lenth | Please note that the lsmeans package has been superseded by emmeans. lsmeans is just a front end for emmeans. Even the lsmeans() function itself is in emmeans. | |
| Feb 11, 2022 at 23:17 | history | edited | Raymond | CC BY-SA 4.0 |
added 8 characters in body
|
| Feb 11, 2022 at 22:30 | history | edited | Raymond | CC BY-SA 4.0 |
added 1702 characters in body
|
| Feb 11, 2022 at 22:00 | comment | added | Raymond | a "n_i" before c_i*d_i makes things more complicated. | |
| Feb 11, 2022 at 21:57 | review | Late answers | |||
| Feb 11, 2022 at 22:01 | |||||
| Feb 11, 2022 at 21:55 | comment | added | Raymond | Thanks, @ttnphns. A lot of helpful discussions there. I'm also not sure about how various R packages deal with unbalanced sample sizes in this scenario. In the "Montgomery-Design-and-Analysis-of-Experiments" book, page 94, chapter 3.5.5, when the sample sizes are not equal, Two contrasts with coefficients {ci} and {di} are orthogonal if:sum(n_ic_id_i)==0. | |
| Feb 11, 2022 at 21:40 | comment | added | ttnphns | Search also "contrast coefficient matrix" on this site. This one is a nice thread stats.stackexchange.com/q/78354/3277 | |
| S Feb 11, 2022 at 21:35 | review | First answers | |||
| Feb 12, 2022 at 4:27 | |||||
| S Feb 11, 2022 at 21:35 | history | answered | Raymond | CC BY-SA 4.0 |