SPSS Statistics

 View Only

SPSS: 50 Years of Innovation

By Douglas Stauber posted Thu April 05, 2018 09:34 PM

  


This year, SPSS is celebrating its Golden Anniversary. Decades before the “data science” term was coined, SPSS provided the world with a user friendly, reliable, and trustworthy means to yield actionable insights from data. To commemorate this important milestone, we've highlighted the most important moments from the past 50 years and discuss our exciting future.

[caption id="attachment_7544" align="alignnone" width="300"] Example of a computer punch card[/caption]

SPSS, or Statistical Package for the Social Sciences, started development in 1965 from a group of researchers who were frustrated in serving the teaching and research needs of the political science department at Stanford University. They had to use different programs to perform their data analysis and write time consuming code with minimal documentation. To streamline data analysis tasks, they began working on SPSS - an integrated system with statistical programs built in. The first release was in 1968 as a batch processor that ran on punch card systems. By the end of the '60s, SPSS grew to over 60 universities.

[caption id="attachment_7546" align="alignnone" width="235"] The popular "redbook" from 1975[/caption]

In the '70s, SPSS grew quickly as more users saw the value of a statistical software package. By 1975, nearly 600 organizations - everything from universities to NASA - used SPSS. By this time, SPSS versions were released which expanded the analysis techniques. In the '70s, SPSS contained descriptive statistics, simple frequency distributions, correlation, ANOVA, multiple regression, and even scatter diagrams. SPSS also no longer required punch cards, but was used on mainframes with a client terminal.

[caption id="attachment_7549" align="alignnone" width="300"] An example of SPSS bar chart from the '80s[/caption]

The growth of SPSS continued into the '80s. SPSS version 10 was dubbed SPSS-X for the mainframe and introduced a light version that, for the first time, ran on IBM-compatible PCs. It required MS-DOS 2.0 and at 3MB took nine floppy disks to install. The PC version set a new standard for statistical software on personal computers and became immediately popular. The age of analytics for everyone had begun.

[caption id="attachment_7550" align="alignnone" width="300"] SPSS Data Editor on Windows 3.1 [/caption]

In the '90s, the success of SPSS on the PC naturally led to the first version of SPSS for Windows. Microsoft Windows and Macintosh versions were introduced. These releases really took advantage of the point and click interface and in doing so made the software approachable by more users. The release over the next few years introduced pivot tables, right click menus, and the general linear model procedure. In addition in 1998, Clementine joined the SPSS family and was later named SPSS Modeler. This brought data mining techniques into the SPSS company. Along with other acquisitions, this evolved the SPSS brand into a full featured descriptive and predictive solutions company.

[caption id="attachment_7552" align="alignnone" width="300"] SPSS version 15 Data Editor[/caption]

In the 2000s, SPSS made some huge innovations:


  • V10 introduced near unlimited file sizes, added Excel import, and a client-server version.


  • V14 introduced Integration with Python. This led to a more open environment including .NET, Java, and compatibility with other apps for easy importing and exporting of data.


  • V16 introduced a new UI based on Java, and with this technology came drag and drop, resizable windows, and the ability to simultaneously release across operating systems and user languages.


  • V16 also introduced integration with the R programming language.


  • v18 saw a temporary name change to PASW (Predictive Analytics Software) to reflect the growing product differentiation across the brand. SPSS also joined IBM around this time.


  • V19 restored the SPSS brand, much to user delight.



[caption id="attachment_7562" align="alignnone" width="300"] SPSS Statistics Subscription running the new Bayesian procedures[/caption]

The 2010s focused on addressing the needs of our users. Throughout this period, product betas became a requirement to ensure our end users had the ability to provide feedback based on each version. This user feedback led to the investments in Custom Tables, one click importing and exporting, and stronger ties to open source found in version 24. Then, version 25 introduced Bayesian Statistics, attractive charts, and popular enhancements to our MIXED procedures. User feedback also led to the creation of SPSS Statistics Subscription, the easiest SPSS software ever to buy, setup, and manage.

[caption id="attachment_7567" align="alignnone" width="300"] Screenshot of the SPSS Statistics Preview[/caption]

The innovation of SPSS continues to this day. Both SPSS Statistics and SPSS Modeler have announced major investments to make our software easier to use and more efficient. In addition, our commitment to user feedback continues with the largest open beta of SPSS software ever: all users can try the new user interface of SPSS Statistics today.

SPSS has come a long way. From punch cards to today's advanced software, SPSS has made countless innovations in the past 50 years, and will continue to innovate for our users as we start our next 50.

Special thanks to our sponsored and beta users, our entire SPSS development team, and the countless users around the world that have entrusted us with their data analysis.

[video width="1920" height="1028" mp4="https://developer.ibm.com/predictiveanalytics/wp-content/uploads/sites/48/2018/04/7708_020918_subtitled_YT1.mp4"][/video]





Sources:


  • Nie, N. H. (1975). SPSS: Statistical package for the social sciences. New York: McGraw-Hill.

  • SPSS-X users guide. (1988). Chicago, Ill: SPSS.

  • Norušis, M. J. (1993). SPSS: SPSS for windows base system users guide. Chicago, IL: SPSS.

  • Fridlund, A. J. (1985, February 11). Taking the Numbers Bull By the Horns. InfoWorld, 7(6), 42-50.

  • IBM Release History


  • http://www.unige.ch/ses/sococ/cl/bib/soft/spss.history.html










#50
#newui
#SPSS
#SPSSModeler
#SPSSStatistics
7 comments
7 views

Permalink

Comments

Thu November 22, 2018 11:33 PM

I thought the whole PASW thing was a response to Norman Nie's lawsuit about use of the name "SPSS", which I thought was resolved prior to (or as a result of) the IBM purchase. I thought 19 was first true IBM release.

Mon June 04, 2018 02:24 PM

Hi Peter - Good catch. I've made that change.

Sat June 02, 2018 10:22 AM

Congrats, but wasn't it IBM that tried to change the name of SPSS to PASW statistics? If so, the comment at V19 'SPSS also joined IBM around this time', should go to V18.

Fri April 06, 2018 01:54 PM

I used to need SPSS in my master's degree and M.D. degree till i decided to study it professionally. It is really amazing and i am sure full of more and more treasures.

Fri April 06, 2018 07:49 AM

I congratulate SPSS with its 50 years anniversary. I started to work for SPSS in 1983, and would become technical support specialist at SPSS Benelux BV, that would be opened in January 1984. In 1985 SPSS-PC was introduced and that became a success! We, at SPSS Benelux BV, started to open offices in the UK, Sweden, Germany, which became successful. I am pleased to have taken part in the success of SPSS, and still we are, as now Smit Consult are Business Partner of IBM for data analytics, and we support IBM SPSS Statistics and IBM SPSS Modeler. I expect that IBM SPSS Statistics will be used by many users for many years to come. It is really versatile, comprehensive and easy to learn and easy to use.

Fri April 06, 2018 06:31 AM

It invokes nostalgia whenever there is mention of SPSS rich legacy. I started using SPSS in the year 2001 as a Data Miner. Before using Clementine I was using Intelligent Miner for Data from IBM for predicting the quality of rolled coils at the Hot Strip Mill in a steel manufacturing company and also predicting amount of mono-injection and co-injection for desulphurisation and I also used SQL Server Analysis Services some data mining exercises. Then I joined SPSS South Asia in 2001 where I got exposed to SPSS. About a year a later when Lexiquest for augmenting the capability of Clementine as a Data Mining Tool by bringing in Text Mining. In 2004 June I joined Software Labs at IBM in India. I then again switched back to Intelligent Miner and then in the year 2009 SPSS got acquired by IBM. I then again switched back tp SPSS PASW. And ever since I have used SPSS Modeler for many Client Projects, PoCs, PoTs and Solutions (Analytical Framework for Business Intelligence). It drives me nostalgic as I recall many years of association with SPSS and I still feel I do not know anything in SPSS. Fortunately I have access to many experts who guide me on my client engagements.

Thu April 05, 2018 09:48 PM

I started using SPSS in 1976 while in college. We had the punch cards and it was helpful to know FORTRAN. Compared to do all the stats by hand calculator, this was screaming fast even if you had to submit your card deck and wait overnight for the output (& hope there were no syntax errors).
Fast forward to today and I'm celebrating my 42nd year with my desktop Mac install. Still serves me well...