Today I got negative feedback. Few days ago, during a meeting, after hearing lots of percentages and how they are different, I asked “did you run any significance test?“. The answer was worse than I expected: “what’s that?”. I said a plain “thank you” and tried to listen further. Someone asked “what’s a significance test?” and I said shortly “it’s a test that shows you if those numbers are significantly different and not just by chance”. A hostile voice said “what, you studied statistics?!”
So after this meeting, I was told that the presenter felt offended, like I contested her work. Well, it’s not me the one who offended but herself: if she had answered “No” nobody would realize what happened but “what’s that?” showed that she had no idea about what I was talking about. And yes, I did contest her work! I tried to open some eyes, to make people more reserved while analyzing the figures.
I’m sick and tired of being presented all sorts of numbers without ever being told how they came up with them: sample size, is the sample representative, significance tests run in order to be able to say X is significantly more likely to do something than Y, level of confidence, margin of error and you name it. I see this in media but hear it from people too. For god’s sake, at least remember this: if the difference between X and Y is less than the margin of error then X is not significantly different from Y.
Media is bombarding us with numbers, making a big fuss out of some obscure stuff. The worst is that we grow accustomed to expressing about everything in percentages, in statistics. But nobody teaches us how to use this numbers, how to look at them and what is meaningful or just rubbish meant to increase audience. Except when you are studying statistics in university.
When someone do try to understand if the conclusions are reliable or not, the others are like awakened from their sweet sleep: grumpy and hostile. The truth is that they prefer the ignorance, they prefer to be lied with “statistics”. And it hurts. On a larger scale it hurts to see the government telling lies backed up with “statistics”. It’s more convenient for the government to say “the average salary is xxx” using the arithmetic mean for this while the median is much much lower. This is just an example but there are books just about this subject.
A good example is one of my former employers who justified my poor salary increase, of only 10%, saying that my colleagues from Canada had an increase of only 6%. My prompt reply was "yes but their 6% values more than my whole romanian salary including the raise!" He changed the subject as he couldn't contradict that.
As for negative feedback, I'm just wondering what is worse: to make someone feel embarased by questioning his work or to mislead a room full of people with numbers that you cannot really understand?
<<.Previous | Back to summary | Next.>>
