Search This Blog

Tuesday, July 2, 2013

28 B Example: Improving the library process

In most cases the largest cause in a Pareto chart is the first target for improving a process. However, in this example the students aren’t complaining about the time they spend in the stacks. The students’ primary complaint is about the second shortest step in the overall process. Based on these observations the process improvement team decides to divide the problem into two pieces. First they will determine if they can reduce the amount of data required for the checkout process, thereby addressing the main complain of the students, and second they will analyze the process in more detail to see if the overall time can be shortened without major investments in new equipment or facilities.
Examining the rational for requiring the student’s local and home addresses for checking out books determined that this policy traces to the era before the college established computer databases of student information and the use of student ID numbers. It was incorporated into the checkout software just because it was part of the process at the time the software was introduced. Therefore the team checked to ensure they could get access to a student’s addresses if they had a student’s name and ID number. Confirming that they could they dropped this requirement from the checkout process and modified the library’s policy documentation to reflect this change. All librarians and assistant librarians were notified that student address data is no longer to be collected.
The team examined the raw data collected on the individuals using the library to see if any additional insight might be gained on the process. This led to discussions with the assistant librarian that collected the data.  The discussions revealed that if a student returned to the catalog after being in the stacks then each of the times the student spent waiting for the catalog computer, using the computer or searching the stacks were added together. For example, if a student returned to the catalog computer three times then the total time using the catalog computer is the sum of the three independent times. Examining the raw data indicates that about 50% of the students return to the catalog more than once and about 25% return three or more times. Discussions with students returning to the catalog computer more than once showed that these students were noting only one book at a time when doing a catalog search. They then go to the stacks, examine the book and if it doesn’t have the information they need they return to the catalog computer, repeat the search and examine the next most likely book.
Based on this new information the team concluded that the instructions for using the catalog are inadequate, possibly explaining why 81% of those surveyed found the instructions of little or no use. The team modified the instructions to include recommending that the students note several books from each search that might contain what they need before going to the stacks and examining any of the books. The modified instructions were posted beside the catalog computers and notices were posted reminding the library users that the instructions had changed.
Next the team prepared a work flow diagram and a deployment chart to see if any additional insight into improving the process is provided by either of these tools. Work flow diagrams and deployment charts are items that I have found useful that are not in my 1987 copy of the Memory Jogger. These are based on process flow charts and add information not usually associated with flow charts that helps identify process improvements. For our purposes the definitions of these items are:
        Flowchart- A schematic step by step description of a process. It can be top level or very detailed.
        Work flow diagram- A floor plan of the workspace that includes the movements of people and items involved with the process.
        Deployment Chart- A matrix of process steps and workers showing who is responsible for each step.
Simple examples better explain why these diagrams and charts are useful. Important information for process improvement efforts not included on flow charts is the spatial relationships of steps in the process and who the primary and secondary workers are that are responsible for each step.  This information is provided in work flow diagrams and deployment charts. Examples of these are shown in figures 26 and 27 for the book checkout process.
Figure 26 Work flow diagram for checking out a book.
The work flow in figure 26 shows that this work space can be improved to make the student’s job easier by moving the card catalog closer to the stacks so that students that have to make a second and third trip to card catalog need to take fewer steps. The work space can be improved to make the librarian’s job easier by moving the librarian’s computer closer to where the students bring their books at the checkout station so the librarian doesn’t have to move back and forth to access the computer and work with the student. These improvements are not apparent from just the process flow chart illustrated in figure 21.

Figure 27 Deployment chart for finding and checking out a book. (The rectangles indicate a primary responsibility and the ovals a secondary responsibility for a given task)
The deployment chart shown in figure 27 adds new information to the process description, although in this simple case it doesn’t suggest any improvements to the process. An alternate version of the deployment chart lists process steps under columns labeled with the titles of individual workers involved in the process. In general it is good practice for process improvement teams to start by developing the process flow chart, the work flow diagram and the deployment chart as a first step and to try various forms of these charts and diagrams rather than just preparing a single flow chart before beginning analysis.
Often just putting charts and diagrams on paper suggests simple improvements, as the work flow diagram did for the book check out process. The observant student will have noticed that the description of the library processes is much simpler than what actually happens. For example, the flow chart, work flow diagram and deployment charts lack the loop back to the catalog if the student doesn’t find the information needed in the books examined from the first search of the catalog. If all three charts are developed at the beginning of a process improvement activity it is more likely that missing steps are identified.
Checking the effectiveness of candidate improvements
The process improvement team moved the catalog computer and the checkout computer as suggested by the work flow diagram. They then collected new data for a week to see if these changes and the changes they had made to the checkout process had any impact on the number of complaints received or on the time to find and check out books. The post improvement data is shown in figures 28 and 29.

Figure 28 Post improvement complaint check sheet.

Figure 29 Post improvement times for finding and checking out books.
The post improvement data shown in figures 28 and 29 indicate that the objective of reducing the number of complaints was met and some reduction in the average time to find and checkout books was achieved. Although only a small reduction in time the students spend in the stacks was achieved there is a significant reduction in all other times. The students spend less time in queues and less time providing data to the librarians. Note that the time the librarians spend in processing books is also cut in half. This is attributed to the reduction in time listening to students complaining. The time listening to complaints was also increasing the time students had to wait in the queue for checking out books. The librarians found that they had more time for other work and they were enjoying their jobs more without having to listen to so many complaints.
This example of process improvement contains several lessons learned that are worth listing. These include:
·       Processes are often much more complex in actuality than workers first perceive. This is usually because they think of how the process ought to work rather than how it actually does.
·       First attempts at collecting data often miss key data because of incomplete descriptions of processes.
·       Fixing a problem in one step can often have unexpected benefits in other steps in a process. In the library example eliminating the need for providing unnecessary data to the librarians not only reduced complaints it reduced the times librarians spent in listening to the complaints and the time students spent in queues waiting for free librarians.
·       Processes often contain unnecessary steps due to some past rational that no longer applies.
·       Improving processes almost always makes workers jobs easier and if continued over time can reduce the number of workers required to carry out processes.
Process improvement should be an ongoing effort and if it is the improvement teams build on past work so that their flow charts and data collection become refined with time. The fact that first efforts are not as complete as they might be isn’t important as long as improvement activities continue.
I hope that this example shows how continuous process improvement with trained and empowered workers contributes to higher quality work, more motivated and happier workers and a more effective organization. This is the type of organization that survives and thrives in today’s competitive global environment.
I also hope this example has convinced you to learn statistical process control in more depth. This example is only an introduction to statistical methods. There are different types of control charts that are used for different analyses. These include p charts, np charts, c charts and u charts in addition to the x-bar – R chart used in the library example. In addition to control charts you need to know about run charts and scatter diagrams. The Memory Jogger book recommend in an earlier lecture defines all of these methods.
You may try statistical methods on simple work processes based on what you have learned in this lecture. I must caution you that although that may work well you could cause more problems by trying to apply these simple methods to a process that demands more complex treatment or different analysis than described here. If you have access to someone versed in statistical techniques then go ahead and begin using these techniques but be sure to ask the expert to review your work to ensure the methods apply to the processes you intend to improve.
Since some books and training courses on statistical process control do not discuss work flow diagrams and deployment charts you can learn to prepare these tools by practicing on processes around your home. Develop a flow chart, a work flow diagram and a deployment chart for the activities you engage in from the time you get up in the morning to the time you leave for work. If you are married I don’t recommend trying to get your spouse to change the morning routine based on any process improvement you identify from your charts. You may be judged as having gone a bit overboard.

If you find that the pace of blog posts isn’t compatible with the pace you  would like to maintain in studying this material you can buy the book “The Manager’s Guide for Effective Leadership” in hard copy or for Kindle at:
or hard copy or for nook at:
or hard copy or E-book at:

No comments:

Post a Comment