Showing posts with label bayesian statistics. Show all posts
Showing posts with label bayesian statistics. Show all posts

Wednesday, October 7, 2009

Using Data to Refine Perception

If you find yourself saying the following:

I don’t need to spend time putting a measurement system in place to tell me what I already know. I’m living the issues every day. I know where the problems are and what to do to fix them.

Shark photo credit to Stock Exchange user: lumix2004You might be in a situation where perhaps you should reconsider your position. Sometimes it is necessary to collect data for other reasons - like convincing others.

Three indicators that investing time collecting data might be worth the investment:

  1. People say they agree with you about what the problems are and what needs to be done to fix them, but their actions are to the contrary. [persuasion via demonstration]
  2. When the same problems keep recurring despite having a process or countermeasures in place to prevent them. [verification]
  3. Other problems keep getting all the attention and resources from the leaders and your peers, despite your recommendations or insistence otherwise. [proportional prioritization – magnitude/scale]

Perception is reality. Data is an effective tool to influence perception in order to provide an opportunity to improve the reality.

* Note: Statistics can be [mis]used or misunderstood to support almost any theory like the relationship between shark bites and ice cream sales. I don’t mean for you to use data in a deceptive way just to influence others to support your position. Make sure the data and your understanding of the situation match - checks and balances.

Friday, May 15, 2009

Questions Lead – Answers Follow

One of the fundamental mistakes I have seen repeatedly that leads to the loss of expected impact of improvement to a business process is making changes based on individual observations, education, and experience alone. What is missing is asking questions of all the people impacted by a proposed change. The people that operate, support, audit, and reap the benefits of a process, collectively referred to as stakeholders for the purposes of this post, often have valuable knowledge about a process.
Processes are used to complete a task, or multiple tasks, repeatedly in a consistent manner. The nature of the repetition gives the local stakeholders many observations with which to build a mental repository of data about a process. There is one tool (often under-utilized) at everyone’s disposal that allows the harvesting of that valuable data. Asking questions will yield valuable insight about exceptional conditions, frequencies and probabilities of occurrences, pareto of key information, historical perspective, and trends. The information may not appear as reliable as data from a sophisticated electronic data collection device, but undervaluing it will certainly reduce the effectiveness of an “improvement”.