Reporting Regression Analysis Results

[E. Bernardo]

Are there online articles/papers concerning guidelines in reporting regression analysis results.

 

Thank you.

---

Open emails faster.
Yahoo! recommends that you upgrade your browser to the new Internet Explorer 8 optimized for Yahoo!.Get it here! (It's free)

[Marta Garcia-Granero]

Oooops... this should have gone to the whole list, not to Eins Bernardo
alone (sorry for the duplicate posting, Eins).

MGG


Eins Bernardo wrote:

> Are there online articles/papers concerning guidelines in reporting
> regression analysis results.

Hi Eins:

Not exactly a free online resource (you need subscription to get the
article), but this paragraph from Royston et al. (Prognosis and
prognostic research: Developing a prognostic model. BMJ 2009; 338:b604.
doi:10.1136/bmj.b604) might give you some clues:

"Developing a model is a complex process, so readers of a report of a
new prognostic model need to know sufficient details of the data
handling and modelling methods. All candidate predictors and those
included in the final model and their explicit coding should be
carefully reported. All regression coefficients should be reported
(including the intercept) to allow readers to calculate risk predictions
for their own patients."

Also, Lang and Secic, in their highly recommended book "How to Report
Statistics in Medicine. Annotated Guidelines for Authors, Editors, and
Reviewers" ACP Series ISBN: 0-943126-44-4) devote a chapter (#7) to the
reporting of regression analyses. They suggest to present the
coefficients (for logistic and Cox regression, also EXP(b) should be
reported) with their SE, 95%CI and p-values. You can take a look at the
chapter (at least, part of it) in Google books (check if the link gets
split in more than one line, it's rather long)

http://books.google.es/books?id=kBUBRh1AWG4C&pg=PA85&lpg=PA85&dq=Lang+Secic+How+to+report+statistics+in+medicine+Reporting+regression&source=bl&ots=Mgr-MlhLtl&sig=jDqSpXfEzfqg3JpNqtAb7ozORXo&hl=es&ei=oDy7S7fxGcfW-Qa7gLHACA&sa=X&oi=book_result&ct=result&resnum=3&ved=0CBQQ6AEwAg#v=onepage&q=&f=false

HTH,
Marta GG

--
For miscellaneous SPSS related statistical stuff, visit:
http://gjyp.nl/marta/

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Hui K.]

Hi Eins Bernardo,

Further to Marta's email, you may also find the following book useful for reporting regression analysis results.

Title: Using multivariate statistics (5th ed.)
Authors: Tabachnick, Barbara G.; Fidell, Linda S.
Publication date: 2006
http://www.amazon.co.uk/Using-Multivariate-Statistics-Barbara-Tabachnick/dp/0205459382/ref=sr_1_4?ie=UTF8&s=books&qid=1270714912&sr=1-4

All the best,
Victoria


Handle them carefully, for words have more power than atom bombs. - Pearl Strachan Hurd
===============================
Miss Victoria Ka-Ying Hui

Research Fellow
School of Health Sciences
Room 4005, Nightingale Building (Bldg 67)
University of Southampton
Highfield, Southampton
SO17 1BJ
United Kingdom

Tel: [phone] (ext. 27888)
Email: [hidden email]
===============================
-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Marta García-Granero
Sent: 06 April 2010 15:00
To: [hidden email]
Subject: Re: Reporting Regression Analysis Results

Oooops... this should have gone to the whole list, not to Eins Bernardo
alone (sorry for the duplicate posting, Eins).

MGG


Eins Bernardo wrote:

> Are there online articles/papers concerning guidelines in reporting
> regression analysis results.

Hi Eins:

Not exactly a free online resource (you need subscription to get the
article), but this paragraph from Royston et al. (Prognosis and
prognostic research: Developing a prognostic model. BMJ 2009; 338:b604.
doi:10.1136/bmj.b604) might give you some clues:

"Developing a model is a complex process, so readers of a report of a
new prognostic model need to know sufficient details of the data
handling and modelling methods. All candidate predictors and those
included in the final model and their explicit coding should be
carefully reported. All regression coefficients should be reported
(including the intercept) to allow readers to calculate risk predictions
for their own patients."

Also, Lang and Secic, in their highly recommended book "How to Report
Statistics in Medicine. Annotated Guidelines for Authors, Editors, and
Reviewers" ACP Series ISBN: 0-943126-44-4) devote a chapter (#7) to the
reporting of regression analyses. They suggest to present the
coefficients (for logistic and Cox regression, also EXP(b) should be
reported) with their SE, 95%CI and p-values. You can take a look at the
chapter (at least, part of it) in Google books (check if the link gets
split in more than one line, it's rather long)

http://books.google.es/books?id=kBUBRh1AWG4C&pg=PA85&lpg=PA85&dq=Lang+Secic+How+to+report+statistics+in+medicine+Reporting+regression&source=bl&ots=Mgr-MlhLtl&sig=jDqSpXfEzfqg3JpNqtAb7ozORXo&hl=es&ei=oDy7S7fxGcfW-Qa7gLHACA&sa=X&oi=book_result&ct=result&resnum=3&ved=0CBQQ6AEwAg#v=onepage&q=&f=false

HTH,
Marta GG

--
For miscellaneous SPSS related statistical stuff, visit:
http://gjyp.nl/marta/

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD