• General Linear Mixed Models (part of the Advanced Statistics Add-On) allow you to “create more accurate models for predicting non-linear outcomes based on hierarchical/nested data or data that includes repeated measures.”
• Faster Performance: Supposed to offer faster compiled transformations in SPSS Statistics server, as well as faster pivot table generation in SPSS Statistics Base (which is a good thing — tables in SPSS 16+ seemed to take a lot longer than they did in SPSS 15).
• Statistics Web Portal: Another feature for those of you using the SPSS Statistics Server lets you allow clients to do their own ad-hoc queries and analysis via a web interface. Also requires SPSS Collaboration and Deployment Services.
• Automated Linear Models: This should be fairly interesting. It allows you to build “powerful linear models in an easy, automated way in SPSS Statistics Base and in the SPSS Statistics Server.” It promises “automated data preparation to get data ready for analysis in a single step; techniques to improve model stability and accuracy; support for categorical predictors; more variable selection methods; interactive visualizations.”
• Syntax Editor Improvements: Apparently more than a dozen performance and ease-of-use enhancements in the syntax editor.
• Default Measurement Level: Allows you to automatically set your measurement level. The software automatically looks at the data when you open the file and decides what type of measurement level is best. Not that this took a lot of time, but every little time saver helps I suppose.
• New Direct Marketing Capabilities include an improved interface; a scoring wizard; better, smarter output; integration with Salesforce.com.
• IBM SPSS Collaboration and Deployment Services: New software to let you share and re-use data sets, secure them, and make them easier to use across the company.

There are lots of information and documentation available for SPSS 19. Read “IBM SPSS Statistics: What’s New” for detailed descriptions of the new features or you can download a free 14-day trial of IBM SPSS Statistics 19 from the SPSS web site. All of the user manuals for IBM SPSS Statistics 19 are also available online, including plug-ins, tools, command syntax reference, and guides to each of the modules.

[Note: New features of IBM/SPSS Statistics 19 have been announced]

• Improved research and reporting tools
• New bootstrapping module for more stable models
• New non-parametrics tests procedure
• Automated data preparation for easier model building
• New Direct Marketing add-on module
• More customizability in the base PASW Statistics
• Easier enterprise integration, deployment and management

PASW Bootstrapping Module

As I understand it, this new module lets you automatically run statistical tests on thousands of randomly drawn sub-samples of your dataset, which lets you minimize impact of outliers and anomalies on your results. If this is easy to use it could be exceptionally useful and time saving. It can be applied to the following SPSS/PASW procedures:

• Descriptive Procedures
  • Frequencies (PASW Statistics Base)
  • Descriptives (PASW Statistics Base)
  • Examine (PASW Statistics Base)
  • Means (PASW Statistics Base)
  • Crosstabs (PASW Statistics Base)
  • T-tests (PASW Statistics Base)
  • Correlations (PASW Statistics Base)
  • Partial Correlations(PASW Statistics Base)
• Modeling Procedures
  • One-Way (PASW Statistics Base)
  • UniAnova (PASW Statistics Base)
  • GLM (PASW Advanced Statistics)
  • Regression (PASW Regression)
  • Nominal Regression (PASW Regression)
  • Discriminant (PASW Statistics Base)
  • Logisitic Regression (PASW Regression)
  • Binary Multi-nominal Logistic Ordinal Regression (PASW Statistics Base)
  • GENLIN (PASW Advanced Statistics)
  • Linear Mixed Models (PASW Advanced Statistics)
  • Cox Regressions (PASW Advanced Statistics)

PASW Direct Marketing

SPSS PASW Direct Marketing is a new model specifically designed to make statistical procedures that can be used to better understand customers more accessible and easier to use.  This is accomplished by cleanly packaging six different procedures/techniques that might be accomplished through other means but that are unlikely to ever be done simply because they are so complex. It appears here that SPSS has made it fairly easy. These procedures are:

• RFM Analysis allows users to score and rank customers by analyzing a combination of how recently they made their last purchase, how frequently they purchase, and how much they spend. The results can then be analyzed and output in a variety of ways.
• Cluster Analysis allows you to group your customers/data into different market segments which can be used to improved your targeting. It includes an interactive model viewer that allows the user to visualize and play with the results in real time.
• Control Package Test allows you to utilize the results of marketing campaigns to figure out which of your customers are most likely to respond to similar campaigns.
• Propensity to Purchase Analysis lets you determine which of your customers are the “lowest hanging fruit” so you can minimize your marketing costs.
• Postal Code Analysis helps you determine best locations for brick-and-mortar stores and other geographic sensitive decisions.

PASW Statistics Developer

PASW Statistical Developer is a new module that gives PASW Statistics users access to thousands of free R algorithms and packages using an interact that is “indistinguishable” from PASW Statistic’s built-in dialogs.

Non-Parametric Testing

Offers more non-parametric testing, which is the ability to make multiple comparisons with non-normal (that is, non-bell shaped) data.

A few links with more information:

• Press Release for SPSS PASW Statistics 18.0
• PDF Detailing PASW Bootstrapping Module
• PDF Detailing PASW Direct Marketing Module
• Details of PASW Statistics Developer (R & Python)

PSAW Statistics 18 will be released on August 17, 2009. SPSS will be hosting an online event in which they intend to go through all of the new features. Sign up here.

• List of the new features in SPSS Statistics 17
• Home page of SPSS Statistics 17.0 (on the SPSS web site)

SPSS, I really did try. I wanted to like it. I’m one of those people who can’t stand to be using old versions of software — I always want to have the latest features (admittedly, I’m still using Office 2003, but that’s mostly a work thing), and so I always upgrade as fast as I possibly can. And when I received SPSS 16 in the mail last fall I did rush to install it — but after only a few hours of trying to use it, I gave up and for the rest of 2007 and most of 2008 went back to using SPSS 15 (as far as I could tell there weren’t that many differences between the two versions — at least for the features I used).

But I couldn’t very well not upgrade after two “major” version releases of SPSS. I couldn’t suffer the embarrassment of having to tell people that I was still using a version of the software that was now almost 3 years old. So I sat down, installed the software, and even went so far as to adjust the options so that the settings would be just the way I liked them.

I should admit right off the bat that I’m not much of a statistician. There are a lot of features in SPSS that I don’t use all that regularly. I mean, I use just about everything on the “Data” menu and the “Transform” menu on a regular basis, but I don’t use a lot of the fancy options on the “Analyze” menu (incidentally, did you notice I keep saying “Menu” — I’m not much of a syntax geek — VB.NET, SQL, and Actionscript are enough for me on the computer language front). So if all of the magic and wonder of SSPS Statistics 17 is in the SPSS syntax, or in the Python, or in whatever fancy programming options they’ve introduced, I won’t notice them.

What I do notice is the fact that the Java interface feels funny. When I try to select SPSS 17 menu items on my admittedly high resolution screen, I miss. When I select cells, they don’t react in the same way as all of my regular Windows applications. When I click drop down menus, they look weird. Frankly, nothing is really in the right place, or the right size, or organized in a way I like it.

And yes, I did customize the SPSS 17 in the “Options” menu to look as Windowsy as I could. I was hopeful for a little bit when I hit “Accept” and the window changed colors to look less Java-like and more like the regular SPSS windows I’m used to seeing.

But then I tried to run Custom Tables 17. Which, besides looking weird, and besides not offering ANY new features from what I’m used to seeing in SPSS Custom Tables 15, has funny looking output.

Actually, I should take that back. SPSS Custom Tables 17 does have one new feature that I am absolutely not crazy about. After you make your table and try to copy the results into Excel (which is where I make all of my charts — call me old-school) the results now retain their SPSS Output formatting. That is, in the past, when I would copy data from SPSS to Excel, only the data itself would be copied. But now when I copy the data, it brings along all of the formatting, including the percentage signs (which tends to mess with how Excel deals with the data).

So to use SPSS Statistics 17 I either need to change the way I work with data in Excel, or I need convert it all back to what I’m used to.

The thing is, I can’t really find any new features in this new version of SPSS that I actually need. For example, for my simple marketing research purposes I don’t care about the new algorithms, or the speed improvements. I don’t have the Missing Values module, so I don’t care about the new multiple imputation. I don’t use syntax, so I don’t care about the updates to the syntax editor, and I don’t care all too much about the new EZ RFM module. I’m also never going to integrate third party applications.

Improved SPSS 17 integration with Microsoft Office ought to be a plus, but if that simply means that the software is going to format my data against my will (as described above) I’m not impressed.

So for the meantime I am sticking with my beat up old copy of SPSS 15. SPSS Statistics 17 can stay in its box, right next to the box for SPSS 16, and maybe I can look forward to either (a) exciting new features in SPSS Statistics 18.0 that will make me feel compelled to shift to the new version (give me folders to help me organize my variables and I’ll shift for sure!) or (b) maybe SPSS will give up on this Java interface and go back to the Windows interface me and my colleagues at work have come to love.

• Sources of Data (in-person surveys, online survey, etc)
• Import Data Formats (SPSS data files, Excel, CSV, database)
• Average number of variables in the file
• Average number of rows in the file
• Other SPSS products used in conjunction with SPSS Statistics 17
• Menu/Dialog use vs. Syntax
• Use of advance programming features (Python, R, .Net)

Last question offers a list of possible new features for SPSS Statistics 18.0:

• Additional enhancements to programmability
• Improvements to charts and graphs
• Better tools for automatic data preparation (auto transformations)
• Continued development of the customer dialog builder
• Additional enhancements to the Syntax Editor
• Campaign/Direct Marketing Tools (e.g., customer lifetime value)
• Enhanced integration with other SPSS applications
• Updates to existing procedures (e.g., add updated analytics & options)
• Continued enhancements to multiple imputation procedure
• Interactive tools for viewing, exploring, and manipulating output from predictive models

You are also given the opportunity to sign up to become a beta tester of SPSS Statistics 18, as well as the opportunity to discuss your answers with SPSS.

Personally, as someone who deals with literally hundreds of variables at a time, I would really like it if SPSS 18 allows you to (a) color code your variables and (b) put variables into folders (preferably with the ability to put folders inside folders).

I would also love SPSS Statistics 18 to have the ability to easily make a copy of a variable complete with all of the data. You know — just copy a variable on the “Variable View” page and paste it to the bottom as a new variable with everything exactly the same (except maybe that the variable name now has the word “Copy” appended to the end.

I would also like SPSS 18 to get out of Java and go back to the traditional Windows interface. I think it is fair to say that while everyone in my department has SPSS 16 installed, we all prefer to use SPSS 15 because the interface isn’t as weird feeling (we’re still waiting for SPSS 17 to arrive).

If you’d like to participate in the development of SPSS 18, take the survey.

• http://spss-voc.com/mrIWeb/mrIWeb.dll?I.Project=STATISTICS18CUST

SPSS 17.0 new features will include:

• Various new algorithms, plus improved speed and performance with additional multithreaded procedures.
• New multiple imputation procedure in SPSS Missing Values that helps you more easily compelte datasets for more reliable analysis
• Updated syntax editor that makes it easier to create, test and correct syntax
• Improved integration with Microsoft Office
• New SPSS EZ RFM (Recency, Frequency, Monetary value) module to help analysts identify top customers and clients
• Ability to integrate third-party applications, procedures, and graphics packages created in R (through plug-ins)
• Enhanced administrative tools to improve IT configuration of software.

Also to be released is a new SPSS EZ RFM module, designed to help marketers use recency, frequency, and monetary value analysis to find their most valuable customers.

Key reasons offered by “Discover Secrets Your Spreadsheet Can’t Tell You”:

• Easy access to descriptive statistics and frequencies: True. while you can do descriptives in Excel using some of the built-in functions and the data analysis add-in, it is a lot easier and faster in SPSS.
• Wider variety of charts & graphs: True, although I tend to find Excel easier to manage
• Better, more flexible pivot tables Sortof true. That is, true if you have SPSS Tables. If you don’t, then in my opinion Excel pivot tables are easier to work with. SPSS Tables, on the other hand, is extremely easy to use and lets you do a lot more things that you can’t do with Excel Pivot Tables.
• Full set of statistical tests: True. While it is definitely possible to run statistical tests in Microsoft Excel, they’re much harder to find and work with compared to SPSS, where they pretty much come “free” with every function you run.
• Easy to run similar reports and graphics for subsets: True. Using the “Split” function in SPSS, it is relatively easy to create tables and charts for subsets without doing any extra work. Or you can create syntax (SPSS’s macro language) that lets you reuse your tables and codes over and over again.
• Labels instead of codes in your reports: I love this feature. Just because your survey software makes Male=1 and Female=2 doesn’t mean you want to see lots of 1s and 2s in your reports. And while it isn’t difficult to use search/replace in Excel to change all of your 1s to Male and your 2s to Female, SPSS lets you keep your values intact.
• Accurate results when some data is missing: Sortof true. For this item, they point to the benefits you get from using the SPSS Missing Value Analysis add-on module (an extra $800 or so). This tells you whether the questions that were skipped by your respondents will impact your analysis, and will even estimate what these values should have been. Obviously Excel can’t do anything like that, but keep in mind you need to buy the extra module to get it to work.
• Helps you spot data-entry errors or unusual data points: Certainly SPSS can help with this one, but I think you can get these types of results pretty easily in Excel.
• Easy import functions: I’m not sure that I completely agree with this one. It is true that it is easy to bring in text files. And they do provide functionality to bring in ODBC databases, including Excel spreadhsheets, Access tables and SPSS databases. But the interface for doing so is a little funky and the experience isn’t as clean or smooth as it is with Excel.
• Unlimited rows: This point describes how SPSS can handle an unlimited number of rows while Excel can only handle 65,000. Microsoft Excel 2007 can handle unlimited rows too, but SPSS’s assertion may have been true when the article was published.
• Using SPSS saves time and increases productivity: I suppose that really depends on what it is that you’re trying to do. There are a lot of analysis that I find easier to do in Excel. But certainly if you’re doing statistical analysis it is easier and faster in SPSS.
• SPSS makes it easy to understand statistical results. SPSS has added a lot of extra help files and tutorials that explain how you can/should interpret a lot of the statistical jargon that the software spits out. Excel obviously does not.

A few reasons why I still do a lot of stuff in Excel:

• For most people, the learning curve is much less steep with Excel: Learning SPSS was initially an unpleasant experience. It has a lot of options that don’t make sense until after you’ve spent a lot of time with the program. Once you’ve learned the software you’ll be amazed that you ever lived without it (or some other data analysis package) but until then you’ll spend a lot of time cursing it.
• It’s expensive. Especially if you already have Excel. Expect to spend over $1,700 for a copy.
• Charts are easier to manage/control in Excel: In my opinion. While SPSS has a lot of neat charting features, they aren’t as dynamic as Excel’s chart functionality — that is, when creating a presentation, I often need to go back in and tweak the charts or rearrange the data or rearrange the bars. In Excel, this is as easy as editing the underlying spreadsheet, which would automatically update the Powerpoint. In SPSS, you have to recreate the chart and recopy it into the presentation.
• More flexible use of functions: Excel has a lot more functions than SPSS and gives you more flexibility in how you use them.

Read “Discover Secrets your Spreadsheet Can’t Tell You”

I haven’t personally seen SPSS 16 for the Mac, but looking at the pictures Bertolt provided, I can certainly see why one might be concerned. It really does look like something that belongs on Windows 3.1.

SPSS for Mac 16.0′s main toolbar

If you are thinking about getting SPSS 16 for Mac, you may wish to read Bertolt’s open letter to SPSS first. He basically says that the only thing SPSS 16 has going for it over R is the graphic user interface.

For my own part, I’m still using SPSS 15 even though I own SPSS 16. While SPSS has done a good job of incorporating every feature into the new Java interface, it doesn’t feel like the Windows applications I’m used to. It’s kind of wierd. While I can appreciate that it will lower the development costs for SPSS because they can use the same basic code for three different platforms, it generally means that nobody is going to feel 100% at home.

• Bertolt Meyer’s open letter to SPSS about SPSS 16 for Mac

I’ll be the first to admit — I didn’t start using SPSS until version 11.5 so I’ve always had the menu-based, visual interface to work with. I can do some reptitive tasks in SPSS syntax, but most of what I do isn’t repetitive and the menus are much faster.

My objective for SPSS: clean up the data, weight it, and spit out attractive tables broken down by a variety of different market segments.

SPSS Tables makes it easy to do just that. Provided you have your variables set up properly, it is very easy to drag-and-drop all sorts of different crosstab scenarios, and just as easy to create a “banner” with multiple variables (some of the banners I’ve created have 12-15 different variables going across the top with over 50 columns).

The SPSS Tables Editor

The layout of the SPSS Tables editor is fairly straightforward. In the upper left corner is a list of all of your variables with little icons indicating whether they are nominal, ordinal, or scale. This is, in fact, one of the special places in SPSS where it actually makes a difference whether or not you take the trouble to identify each variable type. No worries though — if you want to produce results that are really only appropriate for nominal variables but your variable is set as “scale” you can right click on the variable and temporarily change it.

The lower left panel of the SPSS Tables dialog box (blank in the image shown above) shows you a preview of all of the different values of your categorical variables. You can’t do very much with them other than see them — which is why I tend to hide the “categories” area and use the entire space on the left side to list my variables (all you have to do is click on the space between the “variables” list and the “categories” list and drag down).

On the right is the main work panel, where you can see I’ve already dragged a couple of variables (level of education on top and geographic region on the left). Adding variables is as easy as grabbing them from the list over on the left and dropping them into the “Rows” bar or the “Columns” bar in the work area.

Now while this might sound very Microsoft Excel Pivot Table-like, SPSS Tables has a few extra tricks up its sleeve. For example, let’s say I want to add another variable. In Excel — and in SPSS Tables — I could do something like this:

Notice how I added “Gender” over on the left, and I now will be able to see all of the data for each level of education broken out by gender. I can also do this:

This time the variable is above the geographic indicator. Admittedly, these are both things that you could do in a pivot table. What makes SPSS Tables somewhat special is the ability to do the following:

In the example above, notice how I put the “Gender” variable next to the geographic indicator — in effect creating a banner. This can be very useful when you want, for example, want to quickly see a variable broken down by a variety of different market segments. And because SPSS Tables allows you to select and drag all of your column headers from the left side at once, building such tables go very quickly.

You aren’t limited to one variable going one way or another. You can also create relatively complicated tables to meet your needs. Consider the following:

 In the example  above, I have two variables on the left. I have the geographic indicator on top, but I now have the gender columns broken down by whether they are retired or not retired. Let’s see you do that in Microsoft Excel!

SPSS Tables also has something called a “compact” view to keep your work space from getting to confusing (it is also useful if you have a lot of boxes displayed, which can really slow down your computer).

Incidentally, if you want you can filter in a way equivalent to the “Page” capabilities in the Excel Pivot table by pushing the “Layers” button in the upper right corner of the screen. This will allow you to show (for example) on the results of female respondents in your table. Tables created this way are dynamic — you can switch between the different “layers” in the output screen.

Once you’ve set up your rows and columns, you have several options regarding the types of statistics you want shown. Using the “Summary Statistics” button you can show counts, unweighted counts, row percentages, column percentages, whole table-based percentages, etc, etc. You can display only one statistic per row/column combo or you can display multiple statistics (you might want to show both the count and the column percentage). You can also do the same thing for the totals and subtotals.

This display of categories and subcategories — as well as variables values that you would like to exclude — can be controlled from the “Categories and Totals” page. Here you can also tell SPSS Tables what order to put the variables.

SPSS Tables can also do statistical testing: it can test for independence (Chi-square); compare column means (t-tests); aond compare column proportions (z-test). Results are displayed in relevant tables below your main table. SPSS Tables also has options which allow you to income what should be done with cells that contain missing values.

You have a lot of control over the look and feel of the output using TableLooks. SPSS Tables uses the TableLook you select in the SPSS Options dialog box. Shown above is the TableLook called “CompactAcademic Times Roman” which is my personal favorite. But you do have more than a dozen different formats to choose from.

Tables created by SPSS Tables can be copied and pasted right into Powerpoint or Excel. Tables pasted into Powerpoint appear as graphics, while tables pasted into Excel enter as data fields (without formatting, which can be very helpful).

Incidentally, if you’re one of those people who do like to have syntax for everything you do, you can use the “Paste” button to paste the code needed to generate your table into the SPSS Syntax Editor.

I use SPSS Tables a lot — it is, as I mentioned above, my favorite SPSS add-on module and definitely the one I use the most. But alas — such great functionality comes with a price. SPSS Tables is available for purchase (Windows, Mac or Linux) for $949 on top of the $1,699 that you’ll already have to pay for SPSS Base 16.0 (because SPSS Tables is an add-on you will only be able to use it if you have SPSS Base). Maintenance will cost you an additional $237 a year. It appears that some web sites offer it for as little as $510 for students.

But it is extremely useful, and it will save you a lot of time. If you don’t have SPSS Base, you might want to consider SPSS Desktop Reporter, which contains many of the same features (although I have to admit, Desktop Reporter can be extremely difficult and frustrating to use if you don’t have SPSS Base to edit your data).

• SPSS Tables home page
• SPSS Tables 16.0 at the SPSS Online Store
• SPSS Tables online demo
• $510 student edition of SPSS Tables at JourneyEd
• SPSS Tables instruction manual

Dear SPSS for Mac customer:

Your investment in SPSS for Mac demonstrates your commitment to maximizing
the value your organization gains from statistical analysis. SPSS Inc. is
equally committed to developing technologies that help you derive valuable
insights from your data.

The purpose of this e-mail is to inform you that SPSS will not support the
PowerPC chip in our next product, SPSS 17.0 for Mac, which is scheduled for
release in the third quarter of 2008.

This decision, while not an easy one, was based on two key factors:

--Most Apple computers now rely on Intel chips. Since August 2006, Apple
Inc. has released computers that run on Intel chips instead of PowerPC
chips; as a result, PowerPC market share has declined substantially

--The strategic allocation of development resources. Focusing development on
Intel-based hardware enables SPSS to devote more resources to developing
other features and advancements that SPSS customers have requested

We regret any inconvenience this change may cause you. We appreciate your
loyalty to SPSS for Mac, and we hope that you will select SPSS 17.0 for Mac
to meet your statistical analysis challenges.

Please feel free to contact me with any questions you might have.

Sincerely,

Arik Pelkey
Senior Product Manager
SPSS Inc.

At the time Intel-based macs were released, the MacOS’s overall market share (as reported by NetApplications) was about 4.29% of all computers on the Internet. In February, 2008 it represented 3.07%, while the Intel-based Macs represented 4.39% (grand total of 7.46% — huge increase for Macs overall in 2 years).

It still means that a little over 40% of all Mac users are still using PowerPC based Macs, which means that there are still going to be a lot of people who don’t get to upgrade to SPSS 17.

To be perfectly honest, even though I have SPSS 16.0 I still really only use SPSS 15.0. I’m still having trouble getting used to the new java-based interface. Unless SPSS 17.0 gets some exciting new features, I may continue to stick with 15 anyways.

So…no big loss for the MacOS users?

• Intel/MacOS market share breakdown at NetApplications
• Methodology of how market share was determined by Net Applications
• Link to the original message from SPSS

Raynald’s SPSS Tools: Raynald has been using SPSS for over 15 years and has answered over 2,100 questios in various SPSS mailing lists. His site contains over 675 different syntax, macro and scripting examples, FAQs, a Newbie’s corner, and a support section for the SPSS book he published.

East Carolina Dept. of Psychology SPSS Guide: This page initially doesn’t look like much, but if you’re looking for information about univariate and multivariate analysis, there is a lot here to work with ranging from introductory lessons to much more advanced applications. Most of the SPSS tutorials are in the form of word documents that are quick and easy to print out as a reference.

comp.soft-sys.stat.spss: Not really a web site per se, but it is a great discussion board that contains thousands of messages on every conceivable SPSS related topic. If you have an SPSS question, odds are the answer is already here or someone can give you an answer in just a day or two. The link above takes you to google groups — you can also use a newsfeed to keep up with the latest in your favorite rss news reader.

SPSS Guide by Gil Einstein and Ken Aberethy: Alright, it is a little dated, but if you’re looking for a quick and easy to read step-by-step guide to computing a t-test, running an ANOVA, or setting up a data file, this is a good source.

Internet Guide to SPSS for Windows: This guide is well written and covers all of the basics and not so basic procedures you need to manipulate your data ranging from the basics (syntax, data files and output) to more advanced topics related to data transformation and analysis. It is very syntax oriented and reasonably old (1998), but most of what it describes still works and is helpful for understanding the software better.

If you have any others you’d like to share send me a note!