Data Collection And Quality Management

Data Collection and Quality Management Paper

Data is collected to learn the effectiveness of a particular tool in preventing defects or to look into the cause of a particular defect (Burrill, Ledolter, p.381 Ð'¶1). Data removes the trepidation and uncertainty of an unknown element. For example, opinions vary from person to person and what one person thinks as good, there is another that thinks otherwise and yet another that may have a neutral opinion. One reason for collecting data is to gain an understanding of the data by organizing and graphing the individual values (Albert, Rossman, p.1 Ð'¶2). Secondly, and most importantly, collecting data helps draw conclusions about a larger group of information (Albert, Rossman, p.1 Ð'¶3).

There are other reasons to collect data. Data also helps people to understand the processes with which they work. Data can also be collected for the purpose of process control. Numerous manufacturing processes use feedback data to adjust an input, such as temperature, in order to keep the output at the desired level. Feedback data from quality control activities can be used to make adjustments that result in more products meeting their requirements (Burrill, Ledolter, p. 381, Ð'¶ 1). Data can also be used to improve processes with the intent of increasing productivity.

For example, a delivery service makes a guarantee that all domestic packages will be delivered within 24 hours. What happens when a package is delivered in 25 hours or later? Will the delivery service make amends by waiving the delivery fee, or did the delivery company make a reasonable guarantee? The data gathered for this delivery company will determine the variables associated with each delivery and paint a clearer picture on whether the time interval for delivery is enough to continue with the 48-hour delivery guarantee policy. What the delivery company can do is to measure the distribution of all times it will take to deliver a package within a year. These times cannot be measured, however and there is no way to predict how long it will take to deliver a package in 3 months. What the delivery company can do is measure the time it takes to deliver a package for 10-15 deliveries during the past week, graphing and summarizing the time measurements, or data, collected. The information can be extrapolated with known variables, such as delivery times within the past day or week, and help understand the larger group of deliveries within the next year. Data is nothing, unless interpreted and turned into useful information.

What Data is needed?

The type of data needed falls into two categories, quantitative, and qualitative. Quantitative data is expressed as numbers, or is data that is collected as numbers. An example of quantitative data is the number of hours worked per person, or the number of High-Definition Televisions sold per day. Quantitative data arises through using some tool to make measurements like a clock, a scale, a ruler or some similar instrument (Burrill, Ledolter, 1999, p. 378, Ð'¶ 1). Time to failure, time to repair, length, and weight are all quantifiable measures, which are dependent on the use of some measuring device.

Qualitative data is analyzed through appropriate data classification and arranged into one of several classes according to characteristics the elements have in common (Burrill, Ledolter, p. 388, Ð'¶ 3), and is usually described as verbal or nonnumeric. Qualitative data can be gathered through surveys, or the output gathered through a brainstorming session. Questions such as “How can we improve our products?”, or “What obstacles hold us back from delivering total quality?” It is better to work directly with qualitative statements and attempt to classify, group, and arrange the data to gain a better understanding of the underlying message given (Burrill, Ledolter, 1999, p.379, Ð'¶ 1).

How is it collected?

There are numerous ways to go about data collection. The first step in data collection is to determine the relevant metrics that need to be measured so data can be collected accordingly and knowing what the relevant metrics are, allows data to be collected in a way that provides the best possible information for analysis (University of Phoenix, 2007). Sometimes data needed to solve a problem is gathered as part of a special project, such as looking to find a way to reduce or prevent defects. In this context, data can be gathered through a library search, surveys, or through a brainstorming session. Collecting data thorough these means is difficult, as well-defined data collection is near impossible, as gathering and procuring the data presents a problem in itself.

Another way to collect data is through quality control efforts, where there is a more uniform approach to collect data. In quality control, processes are examined closely as it correlates with quality output. Simply put, a process may be hindering production, and since a process can be decomposed into a collection of interconnected sub-processes, studying the production process is a good starting point for a gathering data (Burrill, Ledolter, 1999, p.382, Ð'¶ 2).

As Burrill and Ledolter state, the actual collection of quality control data is part of a detection procedure (p. 382, Ð'¶ 4), and a significant step in a quality control effort is to design a collection mechanism to detect problematic processes that is inefficient or unnecessary to the overall outcome. When building this data-collection mechanism, one must understand clearly all uses that will be made of the data before the design of the system and organize the data collection to ease the task of recording data and preparing it for use. Also, collecting the data in a way that simplifies the task of analysis must be planned in conjunction with how the data is organized, analyzed, and presented; and by whom will do the collecting (Burrill, Ledolter, 1999, p. 382, Ð'¶ 4),

Before classifying data