Sarah Wiseman, Sandy Gould, Dr Dominic Furniss & Dr Anna Cox
This class is taught to University students, both undergraduates and post graduates as part of a Human Computer Interaction (HCI) course. The aim of the lecture is to have students consider the impact, cause and importance of error, and how they might design to avoid error in the future.
The lesson begins with examples of error occurring from computer or device use (we choose examples from the medical domain) and what the consequences of these errors were. This establishes a reason for understanding the causes of human error. We then discuss the different ways that error can be classified: including the difference between slips and mistakes, and an introduction to some of Norman’s classification schemes. At this point the class is given the Errordiary classification exercise.
Students are provided with a worksheet listing each type of error and asked to work in pairs and find examples of the errors from the Errordiary website. The website is accessible from smart phones and so we have found that there is at least one method of access to the site per two students. We give the students a link to a randomised version of Errordiary so that every person gets a different set of tweets to classify. We assign 20-25 minutes to this task.
Once the students get involved there is usually a lot of discussion as the task is not straightforward: not every error can be classified, and there is not always enough information to do so. Students often want to discuss particular examples with the lecturer. After 15-20 minutes the lecturer asks for examples, and particular errors are discussed with the class.
This task is useful for teaching about human error classification, and also for teaching the students about dealing with real world (messy) data. By giving them a task that is not a straightforward matching task (for instance, providing 8 examples of error and asking the class to classify them) the students are required to fully understand the definition of each type of error and think more deeply about what they have just learnt. It also acts as a good break between two halves of the lecture and gives the students a chance to relax, chat and have fun.
After the Errordiary exercise we cover more theories about human error, including post-completion errors and memory for goals.
Tips for teachers:
Have examples of each type of error when explaining them, preferably from your own experience or Errordiary – contemporary errors are more interesting to students than Norman’s ones. Equally, if you can update the news story about error and computers/devices then this will make the lecture more engaging.
During the Errordiary task, if there is an error that is particularly tough, remember it and bring it up with the class to discuss as a whole. Ask the students to say why they think it should be classified as A or B.