New additions to our web site

The ATS Statistical Consulting Group is proud to announce two new additions to our web site. First, we have completed most of the textbook examples for Applied Survival Analysis, Second Edition by Hosmer, Lemeshow and May. We have completed most of the examples in most of the chapters in SAS, Stata and SPSS. We will start doing the examples in R as time permits. If you are familiar with the first edition of this text, or if you are interested in survival analysis, we think that you will find this book very useful. To see the examples, please visit http://www.ats.ucla.edu/stat/examples/asa2/default.htm .

The second big addition to our website is a series of movies featuring Professor Muthen describing, explaining and illustrating new methods for latent variable modeling in Mplus 5.1. Cool new methods include exploratory structural equation modeling and a new method of handling missing data. There are several movies that you can watch by visiting http://www.ats.ucla.edu/stat/mplus/seminars/whatsnew_in_mplus5_1/default.htm .

Besides all of this, we have also added several new FAQs, some dealing with the analysis of spatial data and others dealing with the explanation and interpretation of two-way and three-way interaction terms in regression.

We hope that you find these new materials useful! Enjoy!

-crw

Patch work

We are sorry that the blog has been quiet for so long! It isn't because nothing has been happening, but rather because so much as been happening. There will be some nice additions to our website soon, so stay tuned for that.

In the meantime, let's talk about patches for statistical software. In general, we recommend applying the patches that vendors provide for their products, as they usually contain fixes for known bugs and may contain new features or improve the performance of the software.

Three programs for which there are recently released patches include SPSS, HLM and SUDAAN. The latest SPSS patch, which upgrades SPSS version 16 to version 16.0.2, contains many important bug fixes. If you are running SPSS version 16, please be sure to visit the SPSS website, download and install this patch. A patch for HLM upgrades version 6 to version 6.06, and the patch for SUDAAN upgrades version 9 to version 9.0.3. All of these patches can be obtained free of charge from the respective vendors' websites. You can also visit http://www.ats.ucla.edu/stat/stat_pkg.htm for more information and relevant links.

We would also like to mention that an upgrade for Mplus will be available soon. That will make the current version of Mplus version 5.1. From what we hear, this upgrade contains some very nice enhancements to version 5, so you will want to watch for it.

- crw

Reminisces of the first WCSUG

Well, the first West Coast Stata User Group meeting is now history, and by all reports, it was a success. There were lots of interesting talks, some about Stata, and some just about statistical issues. In addition to the ATS Stat Consulting group, the meeting was attended by many graduate students, researchers, several Stata employees, and some authors of Stata Press books. Compared to the user group meetings for other stat packages, this meeting may seem small with only about 60 attendees per day. However, many people found the small size to be a big benefit, because there were many opportunities to chat with everyone there. At the end of the first day (it was a two day event), many people went to dinner and enjoyed good food and conversations. The second day ended with the popular "Wishes and Grumbles" session, during which users could tell members of the Stata team about their ideas for improvements or additions to Stata.

If you would like to see the slides of this year's presentations, please see http://www.stata.com/meeting/wcsug07/abstracts.html . If you didn't make it to this year's meeting, you may want to think about attending the next one. The meeting was a lot of fun, and there were plenty of interesting things to learn!

-crw

A new tool for the toolbelt

Well, the Fall Quarter has just begun so I guess I better get writing again.

We are waiting with great anticipation the release of the program SuperMix from Scientific Software International. This looks to be a very useful tool for data analysts.

Basically, SuperMix performs mixed-effects models for continuous, binary, ordinal, nominal and count response variables. It can handle two- and three-level models with random intercepts and random slopes. Although SuperMix is a new product, it is really just the combination of four programs written by Donald Hedeker and Robert Gibbons (both of University of Illinois, Chicage) that were previously available online. SuperMix combines MIXREG, MIXOR, MIXNO, and MIXPREG into one package with a consistent user interface.

The webpage for SuperMix indicates that its planned release is September 2007. Since it is a bit past that time, we are looking the the official release anytime now.

pbe

Free research resources?!

This is just a short little note to let people know that we have compiled a page of free, research-related resources for UCLA researchers. This page can be found at http://www.ats.ucla.edu/stat/research_resources.htm . This page is actually a modification of a flier that we put together to let graduate students know about the free, research-related resources that are available to them. The flier (in .pdf form) is linked from the web page. If you know of any other research-related resources that are free and available to all UCLA researchers or to all UCLA graduate students, please email us and let us know. Research in general, and graduate school in particular, are tough enough, and a little free help is always welcome!

-crw

What's the meaning of this?

If you have ever wondered "what's the meaning of this?" when looking at your output, help may be on the way. We have added lots of new annotated outputs to our web site. We have also reorganized the listing of these pages to make it easier for you to find what you're looking for. We have even added an Annotated Outputs link to our home page. The analyses annotated on these pages are the same analyses that are discussed on our Data Analysis Examples pages, so you can learn about an analysis on the Data Analysis Examples page and then learn more about each part of the output on the corresponding annotated output page. If you don't see an annotated output page for a particular analysis, please check back periodically. We are continuing to add new annotated outputs (and Data Analysis Examples) pages. We hope that you find these pages useful!

-crw

Where did all the consultants go?

Consulting was closed for most of last week, and some of you may have wondered why. This wasn’t our typical end of the qaurter “dark week.” Instead, three of us (Xiao, Christine and Rose) went to the major North American gathering for our tribe, the Joint Statistical Meetings in Salt Lake City, Utah. Now, Salt Lake City in early August may not sound like paradise to everyone, but hidden away from the heat in a convention center, we enjoyed three days of learning, discussing and networking. We even got to see other statisticians dance, and we were surprised at the number who actually do so quite well.

So what goodies did we bring back to share with all of you and each other? I (Rose) took a full day course on adjusting for multiple hypothesis tests, which introduced me to a lot of new (to me) techniques for dealing with multiplicity. I still have some homework to do, but this should be really useful when working with clients who are running a large number of significance tests. Christine enjoyed talks on the use of the ACS (the American Community Survey, which is replacing the long form of the Census), methods for handling missing data in survey non-response, imputation methods in survey research, and finding out about the new features in SUDAAN 9.2 (which should be out any day now).

When asked about the conference, Xiao said, "I really enjoyed the sessions, especially some of the invited sessions, such as the session on causality, with Paul Holland, Judea Peal and Donald Rubin as the three speakers. The round table lunch with Professor James Robins from the School of Public Health at Harvard University was extremely educational and entertaining. I also enjoyed the opportunities to talk to some really awesome statisticians from SAS and Mplus. I learned a lot from them. I (Xiao) also enjoyed my little afternoon trips, such as visiting the University of Utah by tram, having a glimpse of the gleaming Salt Lake and riding the tram to Sandy, watching the great Wasatch mountains go by from afar."

So that’s where we were and what we were doing.

r.a.m.