So the time has come to execute your test protocols, you’ve spent days tweaking and refining your test cases to ensure that they work first time round and you’ll end up with as few deviations and discrepancies as humanly possible.
What about the Test Step Instructions?
How much time and effort have you really put into explaining how to actually execute the protocol. What happens if you get drafted in on a new project and you have to get some else to execute the protocol for you? Are you confident in the knowledge that you have explained how to execute the document correctly?
Another great article! Many pieces of good advice in it and well worth a read.
A couple of points from my experience are:
we use the term Non Proving Step (NPS) against things you need to do to get the test to work, but where you don’t need any output or evidence. We find it helps align test instruction steps against results.
remember to list out whether other items are required to perform the test eg calibrated stop watch to time an action
while we do it rarely, we do sometimes use witnesses - particuarly if the result may be transient (ie of short duration) and no evidence is possible (eg can’t take a screen shot as the result happens too quickly).
have a place on the test script where you can refer to your deviations - they do happen!