Timeline for Can age ever be confounded if it is the independent variable in an observational study?
Current License: CC BY-SA 4.0
16 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 26, 2022 at 12:13 | comment | added | sean.mcgrath | thank you. My (contrived, to make my point clear) Q1 was an attempt to show that the example is problematic. Assuming then that age can be confounded due to a survival bias, this leads to my Q2: Moving back to the Wiki example: Health clearly can create a survival bias and it can also impact happiness. My question is whether there is a clear theoretical reason whether to choose to control for health or not. | |
| Dec 26, 2022 at 8:55 | comment | added | Sextus Empiricus | The alternative title sounds better. | |
| Dec 26, 2022 at 8:52 | comment | added | Sextus Empiricus | @sean.mcgrath I don't follow the numbers in your example and why you choose them but "let's say wealth completely determines happiness and has a strong effect on longevity." sounds like the survival bias and relatively more unhappy people die such that at an older age you have more happy people (but caused by wealth instead of age). So yes indeed, the Wikipedia example is a bit problematic. | |
| Dec 25, 2022 at 23:19 | comment | added | sean.mcgrath | Thank you @sextus for the explanations. And given that 2/3 of my cellar is Bordeaux, I can certainly resonate with the example. I understand what you say, and I believe that I can follow it. However, I still struggle with my Q1. Can you point out what is the fallacy in it? Also can you confirm my suggested title. If so, I will change it. | |
| Dec 25, 2022 at 22:33 | comment | added | Sextus Empiricus | With humans, instead of wines, you can also have vintage effects. For instance when some virus infects mostly a particular age group, then it might not be the age that has the effect, and it could be the different history or exposure that the age group had (e.g. Hoskin's effect or different vaccination programs in different era's like end of the smallpox vaccination program). | |
| Dec 25, 2022 at 22:28 | comment | added | Sextus Empiricus | A good example of confounding effects when age is the exposure variable, as described in the last comment, would be when you compare exclusive Bordeaux wines from two different vintages. If the older ones tastes better than the younger ones, then it doesn't need to be that older is neccesarily better or causes the wine to be better. It can be that the particular vintage happens to have good effects. | |
| Dec 25, 2022 at 22:23 | comment | added | Sextus Empiricus | Correction of that last comment. I can imaging a confounding effect. You might argue that being born in a different year sort of causes age, (it doesn't literally cause age, but it does cause people in a study to have different age) and that could be a confounding variable. It can be that certain effects of being born in a particular year cause happiness and not the age of the people that are born in a particular year. ---- Possibly this effect is assumed to be small. The example from Wikipedia is not a strict formal 'never'. It is an informal example to explain confounding. | |
| Dec 25, 2022 at 22:16 | comment | added | Sextus Empiricus | @sean.mcgrath as long as a person remains alive (and as long as there is no other selection effect) I do not see what would cause age in an observational study. | |
| Dec 25, 2022 at 21:58 | comment | added | sean.mcgrath | I see. It looks like I misused the terminology. I struggled a bit on the terminology because the Wiki example only wrote about 'control.' Should my title then read "Can age ever be confounded if it is the independent variable in an observational study?" As far as I can see, they argue that it can never occur. Is this incorrect? | |
| Dec 25, 2022 at 21:42 | comment | added | Sextus Empiricus | @sean.mcgrath the phrase 'control for age' relates to the situation where age is the confounding variable in another relationship. You control for age when you perform an experiment where you keep 'age' constant or when you correct for it in the analysis. The example on the Wikipedia page is the situation when 'age' is a variable that is being confounded (and they argue that that situation doesn't often occur). | |
| Dec 25, 2022 at 21:18 | comment | added | sean.mcgrath | Re. the first image: yes, this I believe that I understand this. If health or wealth is a mediator, it should not be controlled. My Q1 is related to my title. I give an example to challenge that one can never control for age. Assuming that the answer to Q1 is that age can be determined (e.g., by age if you are talking about the general population), my Q2 is whether health can then be seen as a common cause of both age and happiness. | |
| Dec 25, 2022 at 21:09 | comment | added | sean.mcgrath | The reason for the title is because the Wiki example that I cite maintains (if I read it correctly) that you should never control for the ind var age because it cannot be affected. That is my 'top level' question. | |
| Dec 25, 2022 at 21:01 | comment | added | Sextus Empiricus | Your title seems to be a different question compared to the situation/example in your body text "Should you ever control for the independent variable 'age' in an observational study?" Here one studies the effect of some variable $X$ on another variable $Y$ and 'age' can be a confounding variable in that relationship by having causal relationships with $X$ and $Y$. | |
| Dec 25, 2022 at 20:58 | history | edited | Sextus Empiricus | CC BY-SA 4.0 |
added 110 characters in body
|
| Dec 25, 2022 at 20:50 | history | edited | Sextus Empiricus | CC BY-SA 4.0 |
added 123 characters in body
|
| Dec 25, 2022 at 18:06 | history | answered | Sextus Empiricus | CC BY-SA 4.0 |