THE ASIA-PACIFIC CENTRE FOR RESEARCH, INC.

home > articles >

New SPSS Survey Research Product Transforms Open-Ended Responses into Attitudinal and Behavioral Insight
Source: www.spss.com
Copyright SPSS, Inc. 2004
07/05/04

This is a dated announcement. The material in this announcement could be superceded by more current announcements.

SPSS Text Analysis for Surveys™ software dramatically reduces the time, effort and expense associated with analyzing open-ended responses

CHICAGO, 06/17/04 — Open-ended survey questions have traditionally provided both a wealth of information and a host of challenges to survey researchers due to the many difficulties associated with analyzing responses. To address this opportunity, SPSS Inc. (NASDAQ: SPSSE) today announced the availability of SPSS Text Analysis for Surveys, a new software product that enables commercial, academic and government researchers to more efficiently obtain valuable information from survey text responses. Built on SPSS Inc.’s proven text mining technology, SPSS Text Analysis for Surveys brings a powerful new set of analytic capabilities to a broad class of users.

"SPSS Text Analysis for Surveys' ability to distinguish between positive and negative open-ended responses is the 'Holy Grail' of survey text analysis," said Bob Muenchen, manager of the Statistical Consulting Center at The University of Tennessee Office of Information Technology. "The product’s linguistic capabilities in this regard are excellent. I have used four others that attempted to automate text categorization and only SPSS Text Analysis for Surveys could consistently detect satisfied respondents from dissatisfied ones."

SPSS Text Analysis for Surveys assists survey researchers in transforming unstructured, qualitative data into structured, quantitative data by reliably extracting and classifying key concepts from open-ended survey response. Previously, researchers experienced difficulties in analyzing these responses as they are time-consuming, tedious and expensive to categorize. Using SPSS Text Analysis for Surveys’ linguistic algorithms, researchers can now unlock open-ended data more efficiently, significantly improving analytic quality and decision-making ability as the categorized text responses can be quantitatively evaluated alongside other data. Categorical responses can also be further refined through manual techniques.

"SPSS Text Analysis for Surveys is exactly what the university has been searching for to extract critical nuggets of information from long text survey responses," said John Lemon, senior computing advisor, University of Aberdeen, Scotland. "Overall, the enhanced insights that SPSS Text Analysis for Surveys delivers will significantly improve the quality of our survey analysis and the decision-making ability of our organization."

By using this new product, survey researchers will glean greater analytical value from text responses as well as save time by simplifying in-house creation of code lists and eliminating the need to read verbatim responses word for word. The categories or "codes" produced can be re-used to provide consistent results across the same or similar studies. By reducing the need for outside coding services, survey researchers can deploy scarce funding elsewhere. Additionally, the product’s automated categorization capabilities minimize inconsistencies that can occur due to human error from manually coding.

SPSS Text Analysis for Surveys can import data from multiple sources, including ODBC-compliant databases, Microsoft® Excel files (.xls), and SPSS files (.sav). The product can work in conjunction with the SPSS statistical and data management family of products and the Dimensions™ market research technology platform. Once created, categories can be readily exported for further analysis in applications such as SPSS® for Windows® and Excel.

System Requirements

Operating system: Microsoft® Windows XP, Service Pack 1a, 2000, Service Pack 3 or Me

Hardware:
* Processor: Intel® Pentium®-class
* Memory: 256MB RAM minimum; 512MB recommended
* Disk space: 150MB minimum; more recommended for larger datasets
* Monitor: 1024 x 768 (SVGA) resolution
* CD-ROM drive for installation
* Web browser: Internet Explorer 6.0 or later for online help