![]() |
||||||||||||
![]() |
Observing Requirements
|
|||||||||||
|
If you really want to know how something is done, start by observing requirements from someone performing the task. In some circles, this is called as job shadowing. There may be no better teacher than doing, but watching runs a close second. The advantage to Observation over other forms of Gathering Requirements is that you get to see how the job is actually done as opposed to what users will tell you. That's not to say users aren't truthful. They may, however, not be consciously aware of certain aspects of Business Requirements in their day-to-day activities. For example, when asked to describe data entry procedures, a user might give what he believes to be a detailed answer. Once observed, however, he realizes that he left out "simple" steps that he performs without thinking. For instance, would you remember to include "hit the Enter key" in your instructions? Many would not. We often know how to do our jobs without knowing how to describe that work to another. Observing Requirements prevents little details like that from slipping through the cracks. Observers should be prepared to remain neutral and let the users do their work without commentary. However, observers are permitted to ask questions of the users when clarification of a task is needed. Observers should also be aware of the "Hawthorne Effect." A 1920 study revealed that observed workers performed their tasks more thoroughly and efficiently than unobserved workers. Observing Requirements will not be effective if users exceed normal performance levels and hide any deficiencies in the process.
ClearlyPUT's Requirementing process uses
observing requirements
in it's approach. Summary
If you really want to know how something is done, start by observing someone performing the task. Observing Requirements lets you see how the job is actually done as opposed to what users will tell you. Observers should not involve themselves in the work but may ask questions when necessary. Be aware that users may perform more efficiently if they know they are being observed, thereby invalidating the results.
|