Ask a Silly Question
Do you remember as a child, when you asked about something you didn't understand? The usual response for me was:
"Ask a silly question, get a silly answer"
That seemed to make sense as a child, and when I received this response I usually thought that I must be asking something that other people understood. The response indicated that my question made me look silly and I quickly learned to ask less questions.
And as I grew up, I threw the saying in the same group as other 'clever sayings', useful for making yourself look smart and others feel a bit more stupid.
"If you don't have anything worth saying, don't say anything at all"
"If you don't know where you're going, you'll end up somewhere else"
"Who is more foolish? The fool or the fool that follows the fool?"
But, the thing is that the saying "Ask a silly question, get a silly answer" causes major problems when it comes to adult life. Because we learn not to ask for clarity, we learn not to query what we have been told, we learn to accept blindly what the 'experts' put in front of us, whether its the media, the government...and yes, even the statistician.
So I'm going to suggest something controversial.
Silly questions are good and should be encouraged.
When I train others I often start by saying that there are no silly questions, that if one person doesn't understand then it is likely that others don't as well but are afraid to say so. How many times have you spoken up in a meeting and started with "This may seem like a silly question but...", only for others to also say that they haven't understood?
I remember being in an important meeting once. There were many senior people around the table, most of whom wouldn't even have known who I was. The meeting had started clearly enough, but very quickly it had developed into a jargon filled discussion that was too difficult to follow. I remember plucking up the courage, and speaking up.
"Excuse me, this may be a silly question, but does anybody understand what is being said here?"
Hushed silence, annoyed glares...and then somebody else piped up
"Yes, I'm not really sure what is being discussed - could someone explain in simple terms?"
There's always a breath of relief when someone else supports you, even more so when they are more senior (and more respected) than you. It was no different here.
But you know the really funny thing, the thing that sticks in my mind years later? Nobody could explain anything. The discussion had gotten so out of hand that nobody knew, or understood, what anyone else was saying. And because nobody had wanted to admit their lack of understanding, the discussion had progressed unintentionally into parody. Even more to my amazement, the person that I had interrupted, who was chairing the meeting, also had to admit that he really didn't know what he was talking about either.
Sheepish smiles, muttered apologies...but it wasn't the last meeting I attended with them where I suspected the same thing was happening.
So what's my point, and how does this relate to data and analysis?
We are in a current culture where we are competitive by nature, where we want to be seen as the "expert". I have blogged previously that everybody seems to think they are an "expert" in data analysis, and how this is a dangerous thing. But within our own professions it is also dangerous to put someone on a pedestal as an expert, if that means that we are no longer able to ask the silly questions...if we are unable to query whether what we are being told is right. Often we decide that it is the more experienced analyst, our manager, our director, the media, the government etc. that know better than us, or shouldn't be challenged. We assume, just as I did with my parents, that being in authority automatically made someone 'the expert' .....and by implication knowing all the answers. While I firmly believe in showing respect to others, this doesn't mean that we can't think for ourselves. And being able to ask silly questions about data is vitally important, because once you do it becomes clear that others, sometimes "the experts", don't have the answers after all.
Let me give you an example.
We have heard a lot recently about unemployment rates, that there are more people in employment than ever before, that fewer people are becoming unemployed. The government uses this to suggest that their policies are working, that austerity has been a success, that changes in benefits have encouraged people back into work. I am not going to say whether I agree with this or not, but I do want to show that there are some 'silly questions' about the data analysis that need asking. Here I am just going to concentrate on one (although there are more).
Is a decreasing unemployment rate a good thing?
This is a silly questions, right? Of course a decreasing unemployment rate is a good thing, isn't it? Fewer people being unemployed surely means more people are remaining in work?
Let's look in a little more detail. Here we're referring to unemployment prevalence, in its simplest terms defined as:.