Friday, 11 September 2015

Why Bother With Data Collection?

Why is data gathering so important?
Throughout all of our improvement projects we bore people rigid with our desire to gather data, we do this for two main reasons. Firstly it helps us and the project team to understand the process, specifically whether the process is stable and meeting customer expectations, if we find that it isn’t, it helps us to focus any improvements to address this. Secondly it is important to gather data on the process pre and post improvement in order to measure how effective any improvements have or have not been.

What are the difficulties around data collection?
What, when and how much?
It can be difficult to collect the right data at the beginning before you know what improvements will be implemented, for this reason we try to collect standard types of data for all our projects, measures that we hope to improve following any improvement project and I shall discuss these more later. The nature of the University year can make it difficult to obtain accurate data as volumes and staff workload can be very different according to the time of year e.g September. In contrast to the manufacturing environment many of the University’s processes are non-standard and ill-defined. This makes much of the data e.g. process time, staff time etc extremely variable and by not collecting enough you can get a false impression of the process.
In our experience much of the system data available is partial, either because part of the process is not carried out or recorded using the systems or because the information out of a system is only as good as the information that has gone in.
People can feel threatened by data collection because they fear it is being used to record their performance etc. It is therefore extremely important to involve everyone in data collection and make it clear that it is needed to improve a process for everyone involved and not as a comment on individuals.
Time consuming
Data collection can be extremely time consuming for staff involved in the process and those that need to analyse it after. It is therefore important to make it as easy and intuitive as possible.  

What sort of data do we collect and how?
As discussed earlier PIU have standard measures that we use for our PI projects. These measures allow us to analyse whether the process is stable. They also allow us to compare between our own projects.
  • Process time
This is the total time the process has taken from start to finish (both active and inactive). For example, a customer completed their printing estimate request form on Monday at 3pm, they received their estimate on Wednesday at 3pm. The total process time was 48hrs. This measure is particularly useful in helping you to see how stable the process is (were most done within 48hrs) and whether the customer expectation is consistently met (do they expect it within 24hrs?). You can use a range of techniques to measure the process time, sometimes this information will be available in the system, if not we often use chitties.
  • Staff time
Similar to process time but this is the time it takes for staff to actually work on the form or product. If we take the print estimating example, the estimator had to complete 5 tasks in order to produce the estimate and each step took 5mins so the staff time is 25mins. You want to get the process time as close to the staff time as possible as not only will this improve the service for the customer, it will also potentially help you to eliminate waste for the process. Chitties are also a useful technique to measure this along with process stapling (following the process from start to finish and recording how long it takes staff at each stage.
  • Volumes
When designing a new process it is important that you design one that is fit for purpose and can cope with the volume of work going through it. Volumes can be simple to record and we often ask people to keep 5 bar gates to do this.
  • Error types and rates
This includes measuring how many times the process went wrong and things had to be corrected. Errors want should be reduced in the new process and it is important to understand when and why they occurred to do this. Customer complaints can be a really useful way of identifying these.
Qualitative data can be really useful in understanding the current process and this can include talking to staff within the process about what works well and what doesn’t as well as customers of the process to understand their experiences. We use a number of different ways to gather this sort of data including; interviews, surveys and focus groups.

Collecting data is really important as it helps you to ensure you are addressing the right problems and allows you to measure success. As Sir Arthur Conan Doyle once said ‘It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts’.

No comments:

Post a Comment