Midpines, CA (PRWEB) September 23, 2004
Soft data is getting harder, according to dxResearch, a firm specializing in survey techniques and data analysis. But even with the myriad of easy-to-use survey software and tools available to help, you'll get useless information if you don't know what you're doing.
For instance, the satisfaction survey is a standard tool many businesses use to gather data and assess perceptions and opinions of customers, employees, patients, etc. Such tools are necessary because data shapes information, information shapes decisions, and decisions commit resources. ÂInadequate data design and improper scaling have a low information yield, and the result is risky business. We are in the business of reducing that risk by helping people to generate solid, reliable information,Â says Dr. George Chynoweth, President.
ÂData contain information - they are not information themselves,Â continues Chynoweth. ÂAnd some data contain more information than others. Scaling is the process of turning perceptions, attitudes, opinions, etc., into quantifiable data. But not all numbers are created equal, and the resulting scales can produce as much error as information.Â
The result is misinformation, or even disinformation. The good news is recent research in scaling techniques has shown it is possible to obtain very accurate information from survey data. While the current industry standard is typically less than 75 percent accuracy, it is now possible to exceed 90 percent.
ÂYou can easily create a scale in which the numbers DO NOT contain all the information you need.Â explains Chynoweth. ÂConsider a simple example using two numbers, 40 and 80. 80 inches of rainfall is twice as much as 40 inches, but 80 degrees Fahrenheit is NOT twice a warm as 40 degrees. The rainfall data contain more useful information than the temperature data because they are on a scale that allows you to compare the two at face value. Knowing how to design data is paramount to maximizing the information contained in those data. Â
However, information accuracy is only one of several factors necessary for decision-making information. ÂIn various manufacturing and business processes we are concerned with things like specifications, tolerances, and variation. We can measure these things and make adjustments in order to reduce scrap and re-work, and to improve product quality and the financial bottom line. In an analogous fashion we can design survey items as if they were manufacturing processesÂprocesses from the customer, employee, or patient perspectives,Â says Chynoweth.
Assuming these process items are well scaled, Statistical Process Controls can be used to extract information such as performance, criticality, and variation. These processes underlie and support Key Performance Indicators, such as employee morale, patient satisfaction or customer loyalty. Processes that perform well, that are critical to success and are in control can easily be identified and leveraged. Key processes that need attention because of poor performance or inconsistency can be targeted, and non-critical processes can be safely ignored. Chynoweth calls these Âstrategic process metrics,Â and they require designed data to function.
The final step is visual and cognitive clarity in reporting/displaying information. Chynoweth says information should be understandable at a glance; critical processes should be immediately visible and their performance and variation quickly evaluated. Struggling with the significance of 5.84 and 5.43 on a 7-point scale does not promote cognitive clarity. However, a clean display of a 5% difference in performance, in context, can be quickly seen, understood and evaluated. Additionally, sorting the data by statistic and showing multiple metrics simultaneously results in enhanced information that reveals comparative details and relationships.
ÂThe real beauty of these conversions, sorts, and comparisons,Â Chynoweth explains, Âis that the information contained in the data has not been lost or altered. Information and data integrity remain intact, but now itÂs understandable at a glance. We can easily see where to allocate resources to improve or optimize the Key Performance Indicators.Â
The survey methodology developed by dxResearch includes data design and scaling, development of strategic process metrics, information extraction via Statistical Process Controls, and the visual display of information. ÂActionable informationÂ is not industry jargon at dxResearch, it is their guarantee.
Founded in 2002 by George Chynoweth, dxResearch is on the leading edge of survey design and the application of Statistical Process Controls to soft data. dxResearch can assist with survey design, development, deployment, analyses, presentation, and/or training in this methodology. Chynoweth is also the architect of the Interactive Customer Evaluation (ICE) system now in use throughout the Department of Defense.
George Chynoweth, PhD
# # #