First some comments on terminology. Multilevel models, hierarchical (linear) models and mixed models are the same thing. The term mixed models is used (particularly in biosciences when modeling over time) because of the two aspects of the model - fixed effects (that is averages) and random terms (that is variance- covariances). Personally I use multilevel and not HLM because it  is possible to analyze data  that is not strictly hierarchical such as when pupils are nested in neighborhoods and pupils are nested in schools but schools and neighborhoods are crossed. I believe that the SPSS algorithm  is able to handle such complex structures.

SPSS does have a mixed model facility but it depends on what package you or your institution  have bought and what has been installed on your machine. SPSS was renamed PASW when taken over by IBM and  it has now reverted to its old name.

I have two problems with Mixed models in SPSS

1 I do not like the interface as it presumes that you are doing a repeated measures longitudinal analysis and everything is set within that framework but I am sure you could get used to it

2 (And related to 1) while SPSS with estimate the higher-level variances (eg between school variance) it does not estimate the specific random effects eg the school effects - in the work that I do, this is of substantive interest. I suspect that this is due to the repeated measures notion as people would not be interested the level2 effects which are in this case would be individual people.

Chris Charlton has converted of some of our online Lemma training materials into SPSS. Two Modules have been completed.

• Module 3 on using Multiple regression this can be used as a pre-cursor to the multilevel  part of the course

• Module 5: Introduction to Multilevel Modelling.

Importantly the latter show you how to use SPSS syntax to calculate higher- level residuals. The material separates out concepts from implementation

The course already has material for MLwiN, Stata and R.

http://www.bristol.ac.uk/cmm/news/57.htmlhttp://www.bristol.ac.uk/cmm/news/57.html

http://www.bristol.ac.uk/cmm/learning/online-course/index.html

If you want to understand  about different forms of structures;

see Module 4

http://www.bristol.ac.uk/cmm/learning/online-course/course-topics.html#m04

I am not sure if this is true in the most recent version, but I believe it is under Analyze -> Mixed Models -> Linear. Here is a tutorial for doing so (although it may not be Version 22, I would imagine the steps are fairly similar).

http://methodsconsultants.com/tutorial/24/Estimating-HLM-Models-Using-SPSS-Menus-Part-2

My understanding is the HLR is not the same as HLM.  I found a recent paper by Woltman et al. (2012) that indicates I need an SPSS add-on to run HLM.  Disappointing.

I assume James means HLR = hierarchical linear regression, and HLM = hierarchical linear model.  In the statistical dialect I speak, HLR is a linear regression model (usually OLS) in which variables are entered in steps, but in which the data analyst controls which variables are entered in which step.  Thus, it is not what I would call a stepwise method--stepwise implies that some mindless algorithm decides (e.g., on the basis of the largest change in R2) which variable gets entered next.

HLM goes by at least a couple other names:  multilevel model and mixed model.  I prefer multilevel model, mainly because it helps one avoid the confusion with HLR on the one hand, and with mixed design and mixed model ANOVA on the other hand. 

Anyway, assuming HLM = what I call a multilevel model, you can indeed do it via the MIXED procedure (Analyze > Mixed Models > Linear), as Daniel said--see the link below for examples.  You could also use the GENLINMIXED procedure (Analyze > Mixed Models > Generalized Linear).

http://www.ats.ucla.edu/stat/spss/topics/MLM.htm

Hi James, I struggled A LOT trying to run a mixed model on SPSS and have seen one writer report that it can provide incorrect results, which have concerned me in the past. Let me know if you want that reference. I ultimately decided to run the model in SPSS. Here's a very helpful article I found for longitudinal data that should still be helpful even if you're not using longitudinal data. It's a very practical article and helped me a lot. It conflicts a little with the SPSS instructions, though. I'm currently trying to learn R just so I can run mixed models (painful though). Good luck!

Thanks, Bruce and Andrew.  Yes, HLR is a linear regression model, and HLM is a hierarchical linear model.  I've not seen HLM identified as a mixed model, but I'll check out the articles.  The one I looked at indicated that SPSS requirers an add-on called "Predictive Analytics Software," or PASW for HLM.

I didn't have to download anything additional to SPSS to run the HLM. I saw something about requirements for add-ons, too, but the article I saw was for a prior version of SPSS. I'm running v 19. Please let us know if you find something other than the definitions Bruce had.

Will do.  Thanks.

First some comments on terminology. Multilevel models, hierarchical (linear) models and mixed models are the same thing. The term mixed models is used (particularly in biosciences when modeling over time) because of the two aspects of the model - fixed effects (that is averages) and random terms (that is variance- covariances). Personally I use multilevel and not HLM because it  is possible to analyze data  that is not strictly hierarchical such as when pupils are nested in neighborhoods and pupils are nested in schools but schools and neighborhoods are crossed. I believe that the SPSS algorithm  is able to handle such complex structures.

SPSS does have a mixed model facility but it depends on what package you or your institution  have bought and what has been installed on your machine. SPSS was renamed PASW when taken over by IBM and  it has now reverted to its old name.

I have two problems with Mixed models in SPSS

1 I do not like the interface as it presumes that you are doing a repeated measures longitudinal analysis and everything is set within that framework but I am sure you could get used to it

2 (And related to 1) while SPSS with estimate the higher-level variances (eg between school variance) it does not estimate the specific random effects eg the school effects - in the work that I do, this is of substantive interest. I suspect that this is due to the repeated measures notion as people would not be interested the level2 effects which are in this case would be individual people.

Chris Charlton has converted of some of our online Lemma training materials into SPSS. Two Modules have been completed.

• Module 3 on using Multiple regression this can be used as a pre-cursor to the multilevel  part of the course

• Module 5: Introduction to Multilevel Modelling.

Importantly the latter show you how to use SPSS syntax to calculate higher- level residuals. The material separates out concepts from implementation

The course already has material for MLwiN, Stata and R.

http://www.bristol.ac.uk/cmm/news/57.htmlhttp://www.bristol.ac.uk/cmm/news/57.html

http://www.bristol.ac.uk/cmm/learning/online-course/index.html

If you want to understand  about different forms of structures;

see Module 4

http://www.bristol.ac.uk/cmm/learning/online-course/course-topics.html#m04

Thanks, Kelvyn for these resources.  Given the complexity of these issues and my desire to make my life less complicated, I've handed my data over to a statistician for analysis.