Linear Regression - Robust standard error

Discuss the jamovi platform, possible improvements, etc.
Jorge
Posts: 22
Joined: Thu Apr 01, 2021 8:44 pm

Re: Linear Regression - Robust standard error

Post by Jorge »

Hi,

That's OK, I understand that you are busy.

Do you have any idea on when will you be able to work on this? I ask mainly to know if it will be ready for the beginning of the next spring semester.

Thanks.
User avatar
MAgojam
Posts: 421
Joined: Thu Jun 08, 2017 2:33 pm
Location: Parma (Italy)

Re: Linear Regression - Robust standard error

Post by MAgojam »

Hi, Jorge.
Pay attention to the possible times of the year, for Jonathon the spring period begins today... :')

Cheers,
Maurizio


----------------
Sorry if Off-Topic :sunglasses:
Jorge
Posts: 22
Joined: Thu Apr 01, 2021 8:44 pm

Re: Linear Regression - Robust standard error

Post by Jorge »

MAgojam wrote:Hi, Jorge.
Pay attention to the possible times of the year, for Jonathon the spring period begins today... :')

Cheers,
Maurizio


----------------
Sorry if Off-Topic :sunglasses:
Fair enough. :D :D

For me, the beginning of the spring semester means february. So, shall we expect this feature in jamovi by february 2022? ;)
User avatar
jonathon
Posts: 2609
Joined: Fri Jan 27, 2017 10:04 am

Re: Linear Regression - Robust standard error

Post by jonathon »

hard to say. scrape together some sponsorship, and i'm sure we can have it by then.

cheers

jonathon
rconroy
Posts: 5
Joined: Thu Jan 30, 2020 12:13 pm

Re: Linear Regression - Robust standard error

Post by rconroy »

I would strongly support the inclusion of robust standard errors. They allow researchers to work with data that are heterogeneous, even when the underlying factors that cause the clustering cannot be measured or even specified. That is, they allow us to work with the messy reality of real world data.

Robust standard errors are now part of safe statistical practice, and even on teaching courses we teach students to switch them on at all times.
Jorge
Posts: 22
Joined: Thu Apr 01, 2021 8:44 pm

Re: Linear Regression - Robust standard error

Post by Jorge »

rconroy wrote:I would strongly support the inclusion of robust standard errors. They allow researchers to work with data that are heterogeneous, even when the underlying factors that cause the clustering cannot be measured or even specified. That is, they allow us to work with the messy reality of real world data.

Robust standard errors are now part of safe statistical practice, and even on teaching courses we teach students to switch them on at all times.
Thanks, that's exactly my point. Regression analysis without robust standard errors is pointless and we always tell our students to don't forget to include it in their analysis.
User avatar
jonathon
Posts: 2609
Joined: Fri Jan 27, 2017 10:04 am

Re: Linear Regression - Robust standard error

Post by jonathon »

sure. scrape together some sponsorship, and we'll take it from there.

cheers

jonathon
mickeyi
Posts: 2
Joined: Mon Oct 18, 2021 6:59 pm

Re: Linear Regression - Robust standard error

Post by mickeyi »

Hi Jonathon:

Can I chime in and say that I too would really appreciate the addition of robust standard errors in JAMOVI. Since reading the now classic paper titled, On the Unnecessary Ubiquity of Hierarchical Linear Modeling" (https://pubmed.ncbi.nlm.nih.gov/27149401/), I have to come to appreciate the use of robust standard errors. The article makes a strong case that in psychology we overuse MLM, when all we need with clustered data (for most questions) is robust standard errors. I would be happy to throw some money at this if that will expedite this. Seems like something relatively easy to program...

Please help!

Michael Inzlicht
Jorge
Posts: 22
Joined: Thu Apr 01, 2021 8:44 pm

Re: Linear Regression - Robust standard error

Post by Jorge »

mickeyi wrote:Hi Jonathon:

Can I chime in and say that I too would really appreciate the addition of robust standard errors in JAMOVI. Since reading the now classic paper titled, On the Unnecessary Ubiquity of Hierarchical Linear Modeling" (https://pubmed.ncbi.nlm.nih.gov/27149401/), I have to come to appreciate the use of robust standard errors. The article makes a strong case that in psychology we overuse MLM, when all we need with clustered data (for most questions) is robust standard errors. I would be happy to throw some money at this if that will expedite this. Seems like something relatively easy to program...

Please help!

Michael Inzlicht
Hi Jonathon.

I hope you are well.

Given the interesting message from Michael Inzlicht, do you think it would be possible to move on with the robust standard error feature in jamovi?

Thank you,
Jorge
Post Reply