I Just Made a Mistake

If a worker knows how to do a job, wants to do a good job, and is capable of doing a good job, why do they screw up sometimes? When somebody tells us "It's not my fault, I just made a mistake!", do we let them off the hook? Do we take them to task and hold them accountable? Do we throw our hands up in despair? Is there anything we can do to prevent these kinds of screw ups, or do we just have to powerlessly resign ourselves to the "fact" that there will always be human error?

John Evans (UK) of http://www.human-error.com/ talks about Risk Influencing Factors (RIF's), all kinds of little things that contribute to the risk of an error occuring. So, if a person is entering data from a paper form into a computer, there's a risk that they will enter some data wrong. Even if the person can do the job, wants to do the job well, and knows how to do the job (aptitude, attitude, training) a bunch of little RIF's added together still might conspire to produce errors. And if you ask someone why they made those mistakes, they will honestly be baffled. Holding them accountable would be damaging and ineffective.

For this example, some contributing factors might be ambient noise, a confusing form layout, a screen form that doesn't match the paper form, distraction and interruptions, poor ergonomics, fatigue or illness, stress or worry, poor handwriting, poor lighting, screen glare. Imagine that if four or five of these factors are active, the person can still do the job completely error free. But then, if the situation changes and a few more factors suddenly apply, we cross some threshhold and an error occurs.

Deciding whether and when to hold people accountable seems to be the same as deciding whether to search for the cause of variation in statistical process control. If you search for the "cause" of every random-dance common-cause variation, you are wasting your time and money. Instead, you need to study and modify the system, the tasks, the tools, the procedures to reduce the build in variation and risks. On the other hand, if you don't search for the "cause" of special-cause something-important-just-happened variation, your system is out of control and you're missing a golden opportunity for process improvement.

In the same way, if you hold someone accountable for a problem that is actually produced by the random dance of lots of little risk influencing factors, it will do nothing to reduce the risk of error the next time a bunch of contributing factors all occur at once.

In the absence of strong evidence of incompetence, maliciousness, ignorance or maybe even intoxication, we may still be tempted to hold the person accountable. But we're more powerful if we take this as a cue to look for factors that we can control, rather than fall back on the ineffective blame game.