Portland, OR—With the explosive growth of corporate databases, companies are becoming information pack rats. From product specifications to employee absences to customer complaints, we collect massive amounts of data. But what are we doing with it?

"Not enough," says Marte Sheeran, director of marketing for Northwest Analytical, Inc., maker of NWA Quality Analyst software. "Databases offer huge potential for bottom-line gains, but much of that potential goes to waste because, without specialized software, it's too difficult to glean meaningful information from the massive amounts of data they store."

"Individual data points by themselves are not all that valuable, but they speak volumes when you look at them as part of a process," Sheeran says. "The key is to identify and analyze variation in that process."

Every process has a certain amount of natural (random) variation that is statistically predictable. When analyzing data, the key is to identify which data is within that predictable range and which data is outside it. This is where statistical process control (SPC) software linked to corporate databases can be a powerful tool. It can analyze the data statistically to determine what is predictable variation and what is unusual.

Statistical analysis offers you all kinds of opportunities for learning about a process. If the analysis of your database shows that your process is generating unusual (non-predictable) variation, your first step is to investigate and deal with the cause(s), which typically come from outside the process itself (e.g., a bad batch of raw materials).

Once you have a process that is statistically predictable ("in-control"), you can perform a capability study to see whether the natural variation in your process is acceptable. For example, the natural variation in a manufacturing process might predictably generate 4-5 percent defects, but this level of defects might be too high. The analysis tells you that process changes must be made in order to reduce defects. Further statistical analysis can help evaluate the ability of various process changes to reduce defects.

The beauty of SPC analysis is that it not only evaluates the process to see what can be expected statistically with the current process, it can then display the analysis graphically so that even a non-technical staff person can easily see if the situation is unusual or not.

In choosing SPC software for analyzing data, Sheeran says, companies should consider three issues:

• Ease of use. Many software packages will do the job, but make it so hard that only experts can use them. Software that is easy to learn and use will encourage employees at all levels to work with it.
• Connectivity with a wide-range of databases. Some analytical software packages that are labeled "database compatible" will only connect with one or two types of databases and then only when IT can provide significant support. Your software should be able to easily connect with all standard databases.
• High-quality graphics. Even people with "statistics phobia" can clearly see when a process is going out of control. High-quality graphics allow employees and customers to quickly and easily see when problems develop and are resolved.

With easy-to-use analytical software available, there's no need for companies to be information pack rats. "Don't let your data languish," says Sheeran. "Use it for process improvements and watch how the results enhance your bottom line."

Northwest Analytical, Inc. (NWA) provides analytical software for understanding processes and improving quality. The NWA software suite includes NWA Quality Analyst (comprehensive off-line SQC charting and analysis), NWA Quality Monitor (workstation software for plant floor SPC) and NWA Quality Analyst Web Server (a company-wide solution for SQC reporting and analysis). NWA's software is used in more than 4,500 companies worldwide.

back to top