SPSS is a statistical package that was first developed in 1968 as the Statistical Package for the Social Sciences (SPSS), with SPSS, Inc. being established in 1975. In 2010, IBM acquired SPSS, and IBM SPSS became part of IBM’s Business Analytics Portfolio. According to the IBM Analytics web page, SPSS is a predictive analytics software package that “offers advanced techniques…to help you find new opportunities, improve efficiency and minimize risk.”
IBM SPSS has four main components: statistical analysis and reporting, predictive modeling and data mining, decision management and deployment and big data analytics. Here, we will take a brief look at what each component can do.
This package integrates software that assists with all phases of the statistical analytical process, including the collecting, analyzing and reporting of data. With this software, you can organize and explore data to discover trends and relationships. There are also many options for graphs and charts for displaying the data in an easy-to-read fashion. Software written in other programming languages can be integrated with SPSS Statistics to provide further customization.
The modeler is a “predictive analytics platform” to help you make decisions. Using data that has already been collected, the modeler allows you to make accurate predictive models without advanced programming. Modeler can pull data from text-based documents, such as survey results and social media, to create useful categories of results that can then be further analyzed.
SPSS Collaboration and Deployment Services
This component allows you to automate processes used in other components of IBM SPSS and to share the results securely. This widens the scope of use of the analytics and allows for collaboration across distances, increasing the efficiency of the analytic processes.
SPSS Analytic Server
The Analytic Server aids in using big data to make predictions. This component pulls data from Hadoop distributions so that it can be used by the SPSS Modeler to improve the decision-making process.
Each of these components is available at different levels, for different costs, allowing you to choose your level of need. All components can be integrated with each other and are compatible with other software.
Uses of SPSS
Here are some examples from the IBM SPSS website that show how IBM SPSS has been used to aid a business and government agency.
Cisco’s human resources department used IBM SPSS to analyze feedback on its employee surveys in an effort to determine what factors could lead to employees leaving. One project included using the SPSS Modeler to do “text sentiment analysis” on employees’ written comments to assess how positive or negative employees felt about cultural issues within the company. Another project focused on engineers, who make up 40% of the company’s employees, and found that the company needed to focus on retention packages based on employees’ individual motivations to help them feel valued and productive.
Tennessee Highway Patrol
The Tennessee Highway Patrol (THP) turned to IBM SPSS to determine how best to use its troopers to improve highway safety without increasing staff. Using historical data on traffic, accidents, weather and events (e.g., sporting events and concerts) in the local area, THP used SPSS to build a model that could predict future incidents. The model used geo-tagged data from three years of trooper reports, weather data and other information to find correlations between incidents (e.g., crashes and DUI arrests) and other factors, including time of day. Within one year, the predictions helped THP reduce the number of highway deaths and injuries by 6% and increase the number of DUI arrests by 34%. With continued data analysis, the THP is aiming to improve even more on keeping its highways safe.
If you are interested in developing a deeper understanding of software packages such as SPSS, you may want to consider pursuing a degree in data analytics.