Timeline for Interpretation of regression coefficients with multiple categorical predictors
Current License: CC BY-SA 4.0
25 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Oct 3, 2021 at 14:37 | answer | added | Rick Hass | timeline score: 0 | |
| Oct 3, 2021 at 14:36 | comment | added | Rick Hass | @SimonHarmel whoops, skipped right over that. See my answer | |
| Oct 1, 2021 at 21:51 | comment | added | Simon Harmel |
@RickHass, (standlrt | school) shows that standlrt is at level 1 (student-level). Can you please elaborate on what you mean?
|
|
| Oct 1, 2021 at 12:59 | comment | added | Rick Hass | What level is standlrt at? Is it a school-level variable or a student-level variable? So far the other answers are implicitly ignoring the fact that you have effects at different levels which slightly changes the interpretations | |
| Oct 1, 2021 at 0:04 | history | edited | rnorouzian | CC BY-SA 4.0 |
deleted 43 characters in body
|
| Sep 28, 2021 at 21:20 | vote | accept | Simon Harmel | ||
| Sep 28, 2021 at 21:15 | answer | added | rnorouzian | timeline score: 5 | |
| Sep 27, 2021 at 0:26 | history | edited | Simon Harmel | CC BY-SA 4.0 |
added 29 characters in body
|
| Sep 27, 2021 at 0:16 | history | edited | Simon Harmel | CC BY-SA 4.0 |
deleted 1 character in body
|
| Sep 26, 2021 at 22:43 | history | edited | Simon Harmel | CC BY-SA 4.0 |
added 1854 characters in body
|
| Sep 26, 2021 at 18:00 | history | tweeted | twitter.com/StackStats/status/1442187239673372679 | ||
| Sep 26, 2021 at 13:40 | answer | added | gung - Reinstate Monica | timeline score: 2 | |
| Sep 26, 2021 at 9:24 | comment | added | ttnphns | The specialty of your example, though, is that your design has missing cells. Cell "girls x boyonly school" is empty, likewise cell "boys x girlonly school". So I recommend you to obtain the vector of predicted values and check yourself, which differences the coefficients represent. | |
| Sep 26, 2021 at 9:24 | comment | added | ttnphns | It is not difficult to answer, keeping in mind that under dummy encoding each parameter is the difference in prediction values between the current (focal) group and the reference group, while intercept is the prediction value in the reference group. In your model you have 2 factors and no interactions. Then, naturally, each term of the said differences is the average across the levels of the second, opposite factor. And the intercept is the prediction on the reference x reference subgroup. | |
| Sep 26, 2021 at 2:42 | history | edited | Simon Harmel | CC BY-SA 4.0 |
edited title
|
| Sep 26, 2021 at 1:33 | history | edited | Simon Harmel | CC BY-SA 4.0 |
added 2 characters in body
|
| Sep 26, 2021 at 1:28 | history | edited | Simon Harmel | CC BY-SA 4.0 |
added 2 characters in body
|
| Sep 25, 2021 at 22:06 | comment | added | Simon Harmel | @ttnphns, dummy-coding. | |
| Sep 25, 2021 at 22:05 | history | edited | Simon Harmel | CC BY-SA 4.0 |
edited body
|
| Sep 25, 2021 at 21:53 | comment | added | ttnphns | What was the categorical encoding type aka contrast type? | |
| Sep 25, 2021 at 21:38 | history | edited | rnorouzian | CC BY-SA 4.0 |
edited body
|
| Sep 25, 2021 at 21:25 | history | edited | Simon Harmel | CC BY-SA 4.0 |
added 226 characters in body
|
| Sep 25, 2021 at 21:11 | history | edited | Simon Harmel | CC BY-SA 4.0 |
deleted 5 characters in body
|
| Sep 25, 2021 at 20:52 | history | edited | Simon Harmel | CC BY-SA 4.0 |
edited body
|
| Sep 25, 2021 at 20:23 | history | asked | Simon Harmel | CC BY-SA 4.0 |