I have a question about selection biases

[sadpatato-897]

There have been many studies done to find out if cell phones cause infertility. Most have found that there is a positive correlation between phone use and infertility. However, most studies are done at fertility clinics and their study population is fertility patients. My question is, does the fact that the study population is fertility patients create a selection bias? Or, is the study still valid as even though the population is already experiencing fertility problems, it is still a level playing field as people who use phones the most are 'more infertile' than others? Thanks for reading this!

Edited December 12, 2020 by sadpatato-897

[studiot]

12 minutes ago, sadpatato-897 said:

There have been many studies done to find out if cell phones cause infertility. Most have found that there is a positive connection between phone use and infertility. However, most studies are done at fertility clinics and their study population is fertility patients. My question is, does the fact that the study population is fertility patients create a selection bias? Or, is the study still valid as even though the population is already experiencing fertility problems, it is still a level playing field as people who use phones the most are 'more infertile' than others? Thanks for reading this!

 

Let's see some evidence for your opening assertions please.

It is true that if  most studies are, in fact, conducted only at fertility clinics - something I very much doubt as that would present mamoth difficulties in finding control groups and properly analysing the data and in the final analysis it could only be valid for people at fertility clinics, whatever their fertility status, selection bias would be a big problem and many studies would fall below the bar of statistical validity.

[swansont]

7 minutes ago, sadpatato-897 said:

There have been many studies done to find out if cell phones cause infertility. Most have found that there is a positive connection between phone use and infertility.

Really? Can you cite some of them?

At best I suspect you mean correlation rather than connection, which implies some kind of causality has been established.

7 minutes ago, sadpatato-897 said:

However, most studies are done at fertility clinics and their study population is fertility patients.

They would likely not count as valid scientific studies

7 minutes ago, sadpatato-897 said:

My question is, does the fact that the study population is fertility patients create a selection bias? Or, is the study still valid as even though the population is already experiencing fertility problems, it is still a level playing field as people who use phones the most are 'more infertile' than others? Thanks for reading this!

Yes, I would say so. How is there a control group if you have already limited the population to people with fertility problems? You’d have to show the studies, but I’m not seeing how you can determine cell phone use as a cause if all of your subjects have fertility problems.  And regardless of what you might be able to show, you have to confirm that cell phone ownership or use isn’t representing some other variable.

(a famous example is the correlation between owning a pickup truck and voting republican. The pickup isn’t the cause of one’s political leanings)

[sadpatato-897]

1 minute ago, studiot said:

 

Let's see some evidence for your opening assertions please.

It is true that if  most studies are, in fact, conducted only at fertility clinics - something I very much doubt as that would present mamoth difficulties in finding control groups and properly analysing the data and in the final analysis it could only be valid for people at fertility clinics, whatever their fertility status, selection bias would be a big problem and many studies would fall below the bar of statistical validity.

Thank you for the reply! What about what I said in regards to there being a level playing feild (like, could it still be that phone users are 'more infertile' than non-users)? Is that something to be concerned about?

1 minute ago, swansont said:

Really? Can you cite some of them?

At best I suspect you mean correlation rather than connection, which implies some kind of causality has been established.

They would likely not count as valid scientific studies

Yes, I would say so. How is there a control group if you have already limited the population to people with fertility problems? You’d have to show the studies, but I’m not seeing how you can determine cell phone use as a cause if all of your subjects have fertility problems.  And regardless of what you might be able to show, you have to confirm that cell phone ownership or use isn’t representing some other variable.

(a famous example is the correlation between owning a pickup truck and voting republican. The pickup isn’t the cause of one’s political leanings)

Thank you so much! Yes, you're right it's more connection rather than a actual, proven link. Here is a article going into a bit more detial about the situation: https://www.pbs.org/newshour/science/radiation-eggs-sperm-cellphone

[studiot]

One article ?

You suggested that most studies were at fertility clinics.

[sadpatato-897]

Just now, studiot said:

One article ?

You suggested that most studies were at fertility clinics.

Yeah, I did. I just thought that article summed it up well. There are also:

https://healthcare.utah.edu/healthfeed/postings/2014/06/061914_cellphone-cause-infertility.php

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6240172/

https://modernfertility.com/blog/sperm/

https://www.berkleycenter.com/about-the-berkley-center/blog/2020/january/beware-men-your-smartphone-could-affect-your-rep/

[studiot]

Thank you.

However a quick reading of the government comparison of studies (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6240172/)  shows three things 

1) The study compared many sources not only cell phones.

2) The study looks at many situations not just clinics.

3) The study appeared to be about male infertility only.

 

What was you impression of this study ?

[sadpatato-897]

37 minutes ago, studiot said:

Thank you.

However a quick reading of the government comparison of studies (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6240172/)  shows three things 

1) The study compared many sources not only cell phones.

2) The study looks at many situations not just clinics.

3) The study appeared to be about male infertility only.

 

What was you impression of this study ?

I don't know, I honestly know very little about science. What do you think?

1 hour ago, swansont said:

How is there a control group if you have already limited the population to people with fertility problems?

I think what they did is they asked fertility paitents how much they used a phone, and found that the higher use of a phone the lower the patients fertility was. I don't know if they still need a non-fertility paitent control or if how they did it was legitimate and non--flawed.

[studiot]

2 hours ago, sadpatato-897 said:

I don't know, I honestly know very little about science. What do you think?

Then why are you talking about selection bias, which is quite a technical matter ?

If you are a non scientist seeking explanation of stuff you are reading in some of these reports, there's nothing at all wrong with that.
Plenty of folks here will happily help with that.

 

I think the government study I picked out looks well thought out, not too technical and offers many useful points for considerations and discussion.

 

Would you like to do that ?

[sadpatato-897]

7 minutes ago, studiot said:

Then why are you talking about selection bias, which is quite a technical matter ?

If you are a non scientist seeking explanation of stuff you are reading in some of these reports, there's nothing at all wrong with that.
Plenty of folks here will happily help with that.

 

I think the government study I picked out looks well thought out, not too technical and offers many useful points for considerations and discussion.

 

Would you like to do that ?

Yes, thank you!