SF-36 scorer

[michael roughton]

Does anyone where I could get hold of a cheap (or even better free)
package to calculate scores for SF-36 surveys?

Thanks alot,

Michael

[Sanjoy Paul]

There is no free package for the analysis of SF-36 Scores.

I remember, we bought one good software from a US company while working at
the Leicester University. I d't remember the name though.

If you remind me day after tomorrow I can give you the details.

Sanjoy

Dr. Sanjoy K. Paul
Senior Medical Statistician
University of Oxford
Diabetes Trial Unit
OCDEM, Churchill Hospital
Old Road
Headington
Oxford, OX3 7LJ
Tel: +44 (0)1865 857283
Fax: +44 (0)1865 857260
Email: sanjo...@dtu.ox.ac.uk
sambh...@hotmail.com

[Ted Harding]

If you are referring to the MOS SF-36 as developed by RAND:

http://www.rand.org/health/surveys_tools/mos/mos_core_36item_survey.html

then provided you have clean data it should be pretty straightfoward
using any decent statistical/database package. Just enter the results,
encode the scoring rules:

http://www.rand.org/health/surveys_tools/mos/mos_core_36item_scoring.html

and let it run. You may need to build in checks that (for instance)
every item is answered, and that each (multiple choice) item gets
only one answer, though in principle the form is so simple that
these checks could be done on-the-fly during data entry (setting
aside cases of violation to be resolved later).

As to software, there are free relational database packages such
as mysql:

http://www.mysql.com

or postgresql:

http://www.postgresql.org

though frankly the power of a relational database is probably
overkill for so simple an application (and they take some learning
if you're not already familiar with SQL).

Also, provided you have the data in a format which can be readily
imported (e.g. comma-separated varables or CSV, such as you can get
by exporting from Excel -- not that I'm encouraging anyone to use
Excel), then the free R statistical package:

http://www.r-project.org

is an excellent engine for statistical analysis (and will eat
the scoring aspect before even starting on its breakfast).

Indeed, if merely computing the scores is the issue, then it
can be done very simply using any program that can carry out
the very straightforward rules on a line-by-line basis. And
for that I personally would use the fast and simple (and free)
tool called 'awk' (though "Real Men Use Perl", but I don't):

http://www.gnu.org/software/gawk/gawk.html

I'm taking your question at face value -- that your asking about
how to compute scores. It would get more interesting if you need
to relate scores to other variables recorded case-by-case for
the respondents. In that context, perhaps a database solution
used in conjunction with R would be a good way to go.

Hoping this helps,
Ted.

--------------------------------------------------------------------
E-Mail: (Ted Harding) <Ted.H...@nessie.mcc.ac.uk>
Fax-to-email: +44 (0)870 094 0861
Date: 31-Jul-06 Time: 16:22:08
------------------------------ XFMail ------------------------------

[John Uebersax]

michael roughton wrote:

> Does anyone where I could get hold of a cheap (or even better free)
> package to calculate scores for SF-36 surveys?

Ron Hays, who worked with John Ware at RAND on developing the SF-36,
has some free scoring software on his website:

http://gim.med.ucla.edu/FacultyPages/Hays/util.htm

Some of these are for the SAS programming language. You can also try
posting a question in the SAS and/or SPSS newsgroups to see if anyone
there will share their code with you.

You might also check the website or discussion forums here:

http://www.sf-36.org/

Hope this helps.
--
John Uebersax, PhD

[Rhian Daniel]

I'm using xtmixed in Stata for the first time (but am familiar with PROC MIXED in SAS), and am wondering if anybody could please tell me whether or not xtmixed allows for a different covariance structure to be fitted in two different treatment groups? For example, in SAS I would type (in my "random" or "repeated" statement) "type=un group=treat". Is there an equivalent in Stata's xtmixed?

 

Many thanks,

Rhian Daniel

[John Uebersax]

I'm replying to my own message to reset the thread title in Google
Groups to the correct one.

It was changed due to Rhian's new question.

With some newsreader software, you can reply to a message in any
thread, give it a new title, and a new thread is started.
However, in Google Groups, this has the effect of renaming the original
thread.

I'll alert Google Groups to the problem. Meanwhile, I am going to try
to place Rhian's question in a new thread.

--
John Uebersax PhD

[John Uebersax]

Again--this resets the thread name. (If people receive MEDSTAT
messages by email, please notify me and I won't do this again, and
sorry for the possible annoyance).

John Uebersax

[Ted Harding]

Well, some do -- e.g. me! But the main inconvenience is simply
deleting these things!

Whether your re-threading has any effect on people who store
or arrange their messages by thread, I can't say (since I don't
bother with this).

So, in my view, don't worry about inconvenience.

Best wishes,
Ted.

--------------------------------------------------------------------
E-Mail: (Ted Harding) <Ted.H...@nessie.mcc.ac.uk>
Fax-to-email: +44 (0)870 094 0861

Date: 02-Aug-06 Time: 14:15:33
------------------------------ XFMail ------------------------------