I just wrote a LinkedIn recommendation for one of my team who’s leaving Cambridge in the new year. It included this phrase
unafraid of the difficult (to ask and often answer!) questions
And he’s not the only one. Questions are a tester’s stock-in-trade, but what kinds of factors can make them difficult to ask? Here’s some starters:
- the questions are hard to frame because the subject matter is hard to understand
- the questions have known answers, but none are attractive
- the questions don’t have any known answers
- the questions are unlikely to have any answers
- the questions put the credibility of the questionee at risk
- the questions put the credibility of the questioner at risk
- the questions put the credibility of shared beliefs, plans or assumptions at risk
- the questions challenge someone further up the company hierarchy
- the questions are in a sensitive area – socially, personally, morally or otherwise
- the questions are outside the questioner’s perceived area of concern or responsibility
- the questioner fears the answer
- the questioner fears that the question would reveal some information they would prefer hidden
- the questioner isn’t sure who to ask the question of
- the questioner can see that others who could are not asking the question
- the questioner has found that questions of this type are not answered
- the questioner lacks credibility in the area of the question
- the questioner lacks confidence in their ability to question this area
- the questionee is expected not to want to answer the question
- the questionee is expected not to know the answer
- the questionee never answers questions
- the questionee responds negatively to questions (and the questioner)
- the questionee is likely interpret the question as implied criticism or lack of knowledge
Some of these – or their analogues – are also reasons for a question being difficult to answer but here’s a few more in that direction*:
- the answer will not satisfy the questioner, or someone they care about
- the answer is known but cannot be given
- the answer is known to be incorrect or deliberately misleading
- the answer is unknown
- the answer is unknown but some answer is required
- the answer is clearly insufficient
- the answer would expose something that the questionee would prefer hidden
- the answer to a related question could expose something the questionee would prefer hidden
- the questioner is difficult to satisfy
- the questionee doesn’t understand the question
- the questionee doesn’t understand the relevance of the question
- the questionee doesn’t recognise that there is a question to answer
Because they’ll make me think, suggest that I might reconsider, force me to understand what my point of view on something actually is. Because they expose contradictions and vagueness, throw light onto dark corners, open up new possibilities by suggesting that there may be answers other than those already thought of, or those that have been arrived at by not thinking.
Because they can start a dialog in an important place, one which is the crux of a problem or a symptom or a ramification of it.
Because the difficult questions are often the improving questions: maybe the thing being asked about is changed for the better as a result of the question, or our view of the thing becomes more nuanced or increased in resolution, or broader, or our knowledge about our knowledge of the thing becomes clearer.
And even though the answers are often difficult, I do my best to give them in as full, honest and timely a fashion as I can because I think that an environment where those questions can be asked safely and will be answered respectfully is one that is conducive to good work.
* And we haven’t taken into account the questions that aren’t asked because they are hard to know or the answers that are hard purely because of the effort that’s required to discover them or how differences in context can change how questions are asked or answered, how the same questions can be asked in different ways, willful blindness, plausible deniability, behavioural models such as the Satir Interaction Model and so on.
Thanks to Josh Raine for his comments on an earlier draft of this post.