By Dr. Mike Clark, Research and Statistical Support Services Consultant

Well, it has been about a year so it’s time for SPSS’ (pre-patched) new version, version 16.  For those with academic research needs, there is little need to upgrade from one version to the next every time one is released, and often you can go a couple versions without noticing much difference. It looks however that this may be one of the most significant upgrades for SPSS in a long time for one simple reason: It can access the code of a statistics package that, for us at RSS and, in our opinion, for many academic researchers, has a lot more to offer-R. Version 16 is not in for another month, so we have no details regarding this implementation, but it could be a great thing for SPSS users as it has been sorely lacking in the utilization of more modern approaches in its base package such as resampling, robust methods etc. Academic users have had to look elsewhere for to execute efficiently, as well as basic functionality (e.g. sample size estimation, testing of various assumptions, missing values analysis) without costly add-ons. Of course, R is still free and if one is willing to use R there is no need for SPSS. However, while R has several GUI projects, those are nowhere near SPSS’ interface yet in terms of user-friendliness (though R-commander and others actually have more functionality for a few things), and if this release allows people who otherwise would not have exposure to a package that can deal with real data issues without the requirement of several thousand dollars’ worth of add-ons, we’re for it.

I also want to mention another thing I’m looking forward to: SPSS has finally provided resizable dialog boxes.  That may not seem like much to some, but that’s always been a source of frustration to me and probably anyone that deals with data on a regular basis. One might also notice the new graphics engine, and while it wouldn’t take much to improve SPSS’graphics capabilities, I would be surprised if it is offering anything that R hasn’t been able to perform for some time now, but which will still be clunky if one has to click their way to customize it. I found it very amusing that on the install pdf of the R plug-in feature SPSS states “SPSS is not making any statement about the quality of the R program”. Actually, this R implementation is making a very loud statement that SPSS can’t or simply refuses to implement the techniques and methods that R can do, and rather than develop the tools themselves or make them affordable it has to rely on another statistical package that can offer many things advanced researchers have been requesting for years now.

Here are the notes from SPSS:

What’s New in SPSS 16.0: SPSS Base

Highlights of new features and capabilities in SPSS Base are described below. Separately, read what’s new in SPSS’ add-on modules and SPSS’ enterprise-level offerings.

In addition to a new interface, SPSS 16.0 offers expanded analytical capabilities and programmability enhancements, enhanced data management and reporting capabilities, and greater performance through multithreaded algorithms.

A new, more flexible user interface

       The new, Java™-based interface makes SPSS even easier to use. You can instantly resize dialog boxes to accommodate long variable names and lists, and quickly drag and drop variables from one pane to another to set up your analysis.  

                                    

You can easily spell-check variable names and value labels in SPSS 16.0.This interface also supports Unicode, so that you can work with data in multiple languages from a single application. You can treat text data according to Unicode properties for tasks like sorting and case conversion.

Expanded analytical capabilities

Also, through the SPSS Programmability Extension, an SPSS Integration Plug-In for R is available. This enables users to access the wealth of statistical routines created in R and use them within SPSS as part of SPSS syntax.

Additional programmability enhancements

                              

An SPSS Programmability Integration Plug-in provides the crucial link and configuration instructions that enable an SPSS syntax job to take advantage of a specific external programming language or dynamic link library (DLL). Through the SPSS Programmability Extension, SPSS currently offers the following plug-ins:

·   SPSS-Python Integration Plug-In

·   SPSS-.NET Integration Plug-In

·   SPSS-R Integration Plug-In

New plug-ins are being developed by SPSS Inc. and will be available for download at SPSS Developer Central.

Also available for download at SPSS Developer Central is the new SPSS Programmability Extension SDK. This provides software developers with the information needed to develop an SPSS Programmability Integration Plug-In for a programming language’s use with the SPSS Programmability Extension.

Enhanced data management and reporting capabilities

SPSS 16.0 includes many enhancements to data management that users have specifically requested. With SPSS 16.0, you can:

·   Change the string length or the data type of an existing variable, using syntax

·   Define missing values and labels for data strings of any length

·   Choose either to round off or add decimal places to calculated dates when using the Date/Time Wizard

·   Benefit from new capabilities in the Data Editor, including the ability to:

o        Find and replace information,

o        Spell check value and variable labels,

o        Configure the Variable View, such as the

o       Ability to sort by variable name, type, or format, etc.

o       The ability to show/hide dictionary attributes

·   Find and replace text in syntax, using the Output Viewer, enabling you to detect warnings to identify problems in your output.

·   Transfer data and SPSS output to and from Excel 2007

·   Suppress the number of active datasets open on the desktop

·   Set a permanent default working directory

·   Reporting enhancements include a new, more powerful visualization engine, which replaces the Interactive Graph Properties (IGRAPH) feature. This makes graph editing faster and easier. (Existing IGRAPH syntax will continue to work.) The enhanced Chart Editor delivers a similar level of functionality as the previous IGRAPH editor.

·   In addition, SPSS 16.0 introduces Python as the default front-end scripting language. Python supersedes SAX Basic as the scripting language for tasks such as automation of repetitive tasks and customization of output. (Existing SAX Basic scripts will continue to work in SPSS 16.0.)

Greater performance and scalability

In SPSS 16.0, several algorithms are multithreaded, which improves performance on machines containing multiple processors and multi-core processors.

The following algorithms in SPSS Base are multithreaded:

·   Linear regression

·   Correlation

·   Partial correlation

·    Factor analysis

 (注：本文转载自http://www.unt.edu/benchmarks/archives/2007/october07/rss.htm）