Date: Fri, 23 May 2008 16:51:12 -0700
Reply-To: Dale Glaser <glaserconsult@sbcglobal.net>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: Dale Glaser <glaserconsult@sbcglobal.net>
Subject: Re: SPSS Advanced Models for V15.0 vs. HLM 6?
In-Reply-To: <315A1127C9694847B86D0EB16A5235CE04C80B@itexcn05.uchc.net>
Content-Type: text/plain; charset=iso-8859-1
.and if SPSS is to be used Bickel, R. (2007) Multilevel Analysis for applied research is a gentle intro with lots of SPSS examples...
that being said, I have a presentation where I have compared the same model in SPSS, HLM6.0, and Mplus and for the most part I get the same parameter estimates and variance components across softwares except SPSS computes (regardless if I use REML or FIML) a markedly different variance component for the slope (but not the intercept)...................SPSS was going to get back to me about that but I hadn't heard word yet...(this was almost a year ago)....I tested this last with SPSS v 15.0.1 so maybe it has been rectified with v. 16.0, but I am very wary of using SPSS for mixed linear models......
dale
"Burleson,Joseph A." <burleson@up.uchc.edu> wrote:
Justin:
While I am a diehard user of SPSS for 40 years, I would say that the HLM has major advantages, and is bullet-proof in the integrity of its' algorithms. Don Hedekor's program is also excellent. The Singer and Willet text on longitudinal modeling using HLM notes these two, plus anohter popular one from UK (sorry not to have the software names right at hand). Finally, SAS is always on the cutting edge of HLM, but can be difficult to use, easier for more technically inclined folks.
SPSS's MIXED procedure has undergone various improvements, starting with correcting real errors a few versions ago. It's syntax is "SAS-like" and so a SAS user can see that SPSS just stole the syntax style.
One of my troubles with SPSS is that the documentation of what you need and how to get it is very skimpy, the syntax, while simple, is not intuitive, and the documentation on line is slim.
Marisa Nurissis's SPSS book on advanced modules has a nice, but short, chapter on MIXED procedures. She is an excellent writter, I just with she had been contracted to write a book ala Singer and Willett (which is masterfully written), on all the aspects of HLM.
Finally, Raudenbush and Bryk's 2nd edition book of HLM is a pithy and highly technical text that accompanies the HLM program, but their elaborations are nil, and they are extremely succinct, assuming a very technical background, one more familiar to biostaticians and programmers.
So, I would get HLM and suffer the learning curve. SPSS is very expensive, too, if all you want is MIXED option. I use SPSS, HLM, and SAS in efforts to get the best of each.
Joe Burleson
Joe B.
________________________________
From: SPSSX(r) Discussion on behalf of Justin Meyer
Sent: Fri 5/23/2008 11:56 AM
To: SPSSX-L@LISTSERV.UGA.EDU
Subject: SPSS Advanced Models for V15.0 vs. HLM 6?
Hi SPSS Listers,
I am at the very beginning stages of learning to use Hierarchical Linear
Modeling (HLM) techniques to analyze student standardized test data. The
company I work for wants to buy a software package for me to use based
on my recommendation. This would be for occasional use. I am
considering:
-SPSS Advanced Models for V15.0
-Scientific Software International Hierarchical Linear Modeling (HLM) 6
Am I missing any other software that I should be considering? Can anyone
tell me about their experiences with either of these? What are the pros
and cons of each? Consultants that do analysis for the company use HLM 6
(but they would not be helping me) so that is a consideration. I won't
be replacing the work that they do, just doing some additional analysis.
Picking the better software is more important than price.
Thank you for any information you can provide. If this is not an
appropriate venue for this question, I apologize in advance.
____________________________________
Justin Meyer
Researcher
Rowland Reading Foundation
1 South Pinckney Street, Suite 324
Madison, WI 53703
phone: 866-370-7323 fax: 608-204-3846
www.rowlandreading.org
____________________________________
=======
To manage your subscription to SPSSX-L, send a message to
LISTSERV@LISTSERV.UGA.EDU (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
LISTSERV@LISTSERV.UGA.EDU (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
Dale Glaser, Ph.D.
Principal--Glaser Consulting
Lecturer/Adjunct Faculty--SDSU/USD/AIU
President, San Diego Chapter of
American Statistical Association
3115 4th Avenue
San Diego, CA 92103
phone: 619-220-0602
fax: 619-220-0412
email: glaserconsult@sbcglobal.net
website: www.glaserconsult.com
=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@LISTSERV.UGA.EDU (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
