Timeline for In linear mixed model, what if the dependent variable is repetitive for a random factor?
Current License: CC BY-SA 4.0
36 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 22, 2023 at 13:22 | comment | added | BenP | Would be OK for me, already went there | |
| Dec 22, 2023 at 7:21 | comment | added | Eve | Let us continue this discussion in chat. | |
| Dec 22, 2023 at 7:17 | comment | added | Eve | Hi, BenP. Thanks so much for your suggestions! Firstly, I checked the attempt results, so sorry that I put the same picture for subject and channels by mistake. I have modified that. And about the anova across channels for each subject, I put the results in the post. There is exactly no variation across channels except for one subject. Yes... I guess that's the reason. But I'm not sure how the similar papers can make it. And my mentor also suggested me keeping the nested model... | |
| Dec 22, 2023 at 7:10 | history | edited | Eve | CC BY-SA 4.0 |
added 414 characters in body
|
| Dec 21, 2023 at 18:30 | comment | added | BenP | You could also make the anova table for each subject, so 13 such tables in total. The proportion variance explained (SSQ channels / SSQ total) is very low in your table. If this would also be the case for each subject, this would confirm the idea of no or hardly any variance between channels within subjects. And that would be an argument to drop the variance in your model3. | |
| Dec 21, 2023 at 13:25 | history | edited | Eve | CC BY-SA 4.0 |
added 46 characters in body
|
| Dec 21, 2023 at 13:11 | comment | added | BenP | You showed an anova like table for the 7 channels, but that was not exactly what I meant. You have 13 subject, and I was wondering how the 7 channel means of subject 1 vary. And those of subject 2 etc. So a 13 by 7 table with mean values of zrating. My guess is that in each row there is hardly any variation. | |
| Dec 21, 2023 at 13:07 | comment | added | BenP | Also, the total variance of all random effects is the same for you model3 and the other models shown in (4). | |
| Dec 21, 2023 at 13:01 | comment | added | BenP | Looking again at (4) Attempts made, you say that you also used subject and channels separately; in the summaries shown I see both time 13 groups (subjects). It's unclear for me which models you precisely ran here and converged. Your Model3 had a "boundary fit", but the optimizer said "convergence code 0 (OK)". Just wondering what would happen if you would use ML instead of REML, or another optimizer. I have the impression that the results for your model3 are actually OK, meaning that there really is NO variance across channels within subjects. | |
| Dec 21, 2023 at 10:57 | history | edited | Eve | CC BY-SA 4.0 |
added 3 characters in body
|
| Dec 21, 2023 at 10:52 | comment | added | Eve | You're right, EEG data is complicated and exhibits complex temporal variations. But here, my independent variables are averaged across timepoints during happyrating, so it doesn't matter. What I want to explore is the overall effect of EEG data. | |
| Dec 21, 2023 at 10:32 | history | edited | Eve | CC BY-SA 4.0 |
added 5 characters in body
|
| Dec 21, 2023 at 7:44 | history | edited | Eve | CC BY-SA 4.0 |
added 359 characters in body
|
| Dec 21, 2023 at 7:29 | comment | added | Eve | Thanks for your suggestion! I just tried that, which provided some information. I added the result in the post! | |
| Dec 20, 2023 at 3:16 | comment | added | BenP | Your guess could be right. Did you check that, for each subject, the channel means of zrating show some variation? Maybe that the two means for sub-a are very very close and this could be the case for all subjects. In that case you may not even need different channels, one could be enough. | |
| Dec 19, 2023 at 17:04 | comment | added | Eve | Thanks for your comment. After answering the question as an comment, I actually also edited the post. | |
| Dec 19, 2023 at 16:26 | history | edited | kjetil b halvorsen♦ | CC BY-SA 4.0 |
deleted 14 characters in body
|
| Dec 19, 2023 at 16:26 | comment | added | kjetil b halvorsen♦ | Please add new information as an edit to the post, and not only as an comment. On this site we want posts to be self-contained, and comments are ephemeral and often not seen by many. | |
| Dec 19, 2023 at 0:40 | history | edited | Eve | CC BY-SA 4.0 |
deleted 5 characters in body
|
| Dec 18, 2023 at 12:44 | history | edited | Eve | CC BY-SA 4.0 |
added 58 characters in body
|
| Dec 18, 2023 at 12:39 | comment | added | Eve | Thanks for your comment. If I understand you right, It's not different type, happyrating is a continuous variable, and a subject should do happyrating for many times(trials). The reason why "each channel corresponds to the same set of dependent variable data " is that these channels simutaneously recorded the EEG data while subjects are doing happyratings. | |
| Dec 18, 2023 at 12:06 | comment | added | CaroZ | Do I understand it right : you have different types of happiness ratings ? Because you write "for the same participant, each channel corresponds to the same set of dependent variable data (happiness ratings for many trials, although not identical for all trials)" | |
| Dec 18, 2023 at 10:23 | history | edited | Eve | CC BY-SA 4.0 |
added 130 characters in body
|
| Dec 18, 2023 at 10:11 | history | edited | Eve | CC BY-SA 4.0 |
added 484 characters in body
|
| Dec 18, 2023 at 8:52 | history | edited | Eve | CC BY-SA 4.0 |
deleted 2 characters in body
|
| Dec 18, 2023 at 2:47 | history | edited | Eve | CC BY-SA 4.0 |
edited body
|
| Dec 18, 2023 at 2:42 | comment | added | Eve | Sorry for confusing description and thanks again! I added more information. | |
| Dec 18, 2023 at 2:41 | history | edited | Eve | CC BY-SA 4.0 |
added 590 characters in body
|
| Dec 17, 2023 at 15:33 | comment | added | Robert Long | Thanks for adding that, but I'm sorry, I still don't understand the problem. You said "Take a partial data of sub-a as an example, every channel records many same trials.". What does this mean ? Also, please can you describe the study design and tell us your research question(s). | |
| Dec 17, 2023 at 12:12 | history | edited | Eve | CC BY-SA 4.0 |
added 18 characters in body
|
| Dec 17, 2023 at 12:11 | comment | added | Eve | Thanks for your suggestion! I have put some data in the description. | |
| Dec 17, 2023 at 12:10 | history | edited | Eve | CC BY-SA 4.0 |
added 153 characters in body
|
| Dec 17, 2023 at 11:38 | comment | added | Robert Long | Please can you explain what this means ", within each subject, the dependent variables are the same dataset for different channels. (that's because in every trial of a subject, EEG data in different channels correspond to the same behavior index, which is the dependent variable)." ?? Perhaps it will help if you show us an extract of your data, for instance the data for 1 subject. | |
| Dec 17, 2023 at 9:55 | history | edited | Eve | CC BY-SA 4.0 |
deleted 13 characters in body
|
| S Dec 17, 2023 at 9:54 | review | First questions | |||
| Dec 17, 2023 at 12:57 | |||||
| S Dec 17, 2023 at 9:54 | history | asked | Eve | CC BY-SA 4.0 |