Timeline for What exactly is the critique in this regression?
Current License: CC BY-SA 3.0
18 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 18, 2015 at 13:29 | comment | added | EdM | This was a linear regression, not a logistic regression, so a strict rule of thumb isn't appropriate. See this CV question and answers for an introduction to a more thorough discussion. | |
| Sep 18, 2015 at 13:17 | comment | added | BCLC | EdM, thanks. So if we are as liberal as possible, we should have 10*11 = 110 and if we are as conservative as possible, we should have 20*11 = 220 and the paper is problematic as 142 is closer to 110 than 220? | |
| Sep 18, 2015 at 13:13 | comment | added | EdM | In analyzing binary outcomes the rule of thumb is to have no more than 1 predictor for each 10 to 20 cases of the least-frequent outcome. It's hard to give a similar rule of thumb for linear regression as in this paper, since so much depends on the structure of the particular data set. I tend to start worrying as the ratio of cases to predictors approaches a lower limit of 10, which is why I would have appreciated some effort by the authors of this paper to validate their model. | |
| Sep 18, 2015 at 12:57 | comment | added | BCLC | EdM, what makes you say 'few' in 'Trying to examine that many variables (11) with so few cases (142) can lead to problems' ? What is the rule of thumb? | |
| S Sep 18, 2015 at 12:54 | history | suggested | BCLC | CC BY-SA 3.0 |
ote regression coefficient is -28.56 not -28.26
|
| Sep 18, 2015 at 12:53 | review | Suggested edits | |||
| S Sep 18, 2015 at 12:54 | |||||
| Aug 31, 2015 at 22:37 | history | bounty ended | BCLC | ||
| Aug 31, 2015 at 22:37 | vote | accept | BCLC | ||
| Aug 31, 2015 at 15:32 | comment | added | EdM | I edited the answer to include some of this discussion. I may have over-explained from your perspective, but on this site we try to post answers that will be helpful to others who might come across this question later and might not be so sophisticated. | |
| Aug 31, 2015 at 15:25 | history | edited | EdM | CC BY-SA 3.0 |
reorganized answer to address question more directly and incorporate comments
|
| Aug 30, 2015 at 2:24 | comment | added | EdM | Standard errors are in table 2 of the original paper, to which I have access, but not shown in the slideshow. Extraversion (E) and openness (OtE) are variables 4 and 6 in table 1 (under "critique" in the slideshow); they have a significant correlation with each other of 0.27, and the sign of the correlation coefficient for each is negative. In this case trying to "control for" OtE in testing the relation of E to FICO by linear regression might have masked a true relation of these variables to FICO. For beta the ratio to the standard error is what matters for significance, not the value itself. | |
| Aug 29, 2015 at 21:51 | comment | added | BCLC | How do you know those are the standard errors? What's overcontrolled? (Don't remember; can't find it online too) | |
| Aug 29, 2015 at 21:51 | comment | added | BCLC | Read again. I don't think I understood earlier, but I think understand better. To clarify, the beta of -28.56 is not statistically significant but the correlation of -0.17 is? What are the two correlated predictors? OtE and E? Are they correlated with each other? Or do you just mean at least two of the predictors are correlated? What do you mean by 'When predictor variables are correlated this can even lead to changes in the signs of coefficients between single comparisons and multiple regression.' ? | |
| Aug 29, 2015 at 1:58 | comment | added | EdM | In this case, the authors may have overcontrolled by including two correlated predictors, so there was no room for evidence of a link of either to FICO in the multiple regression that they ran. I would not be surprised if they performed a different type of analysis like LASSO they might have found one of these 2 to be related to FICO. | |
| Aug 29, 2015 at 1:49 | comment | added | BCLC | EdM, I think I get it now (with your edit). You mean there was no expectation of a link with all other things controlled, but there may be otherwise? | |
| Aug 29, 2015 at 1:46 | history | edited | EdM | CC BY-SA 3.0 |
Added link to relevant CV page
|
| Aug 29, 2015 at 1:46 | comment | added | BCLC | Not sure I understand, but thanks. I'll read again later. For now, I didn't read the paper. The critique is in slides 26-27 in the slideshow... | |
| Aug 29, 2015 at 1:42 | history | answered | EdM | CC BY-SA 3.0 |