3.11 I just want a number!
Main page |
When leaders and project managers with no scientific skills get pressed for time, this is often what one will experience:
- It can't be that difficult.
- Just make the experiments and give them a result in the form of a number.
- They don't care about the details on why it can't be done right.
- For some reason a false result is better than no result.
- They just want a number. How hard can it be?
You think they sound like a bunch of whiners? They are!
Usually I say 42. That's the meaning of life, the universe and everything.
What? You don't like the number? How about π or e? Those are perfectly good numbers.
No, I'm not trying to be funny or provoking. I'm telling you that they are trying to make you break the law, and take the blame if they get caught using falsified results.
I can understand why some people want to please the bosses and co-workers with a can-do attitude. I once had a colleague who wanted to test twice on the same subject. It was well established that the test was destructive for both product and test subject, so the first test would change the test subject so a second test would be affected to the point where the result wouldn't be reliable in any way. He was told that it was the wrong way to do the test by me and another colleague, but he just wanted a number. It was all he focused on. The number. He couldn't relate to the fact that he was in reality falsifying his results. This is actually a serious offense. He might as well have used random numbers by throwing a dice. It would have saved time and money and the results were just as reliable.
I can't stress this enough: Do NOT ever falsify results. No matter how much they want a result and you want to accommodate them, you NEVER falsify your results. There is no "just a number".
Then what do you do? You have to do something. There are a couple of options here:
Some will argue that you can do a "quick and dirty" test and you can give them some valid results later.
Don't ever do that!
You give them some results and tell them that the results may be off or even wrong. They only hear the number. It may not be the ones you gave the results, but somewhere along the line the things you told them about why the results may be faulty are left out and the numbers are what's left. It's human nature.
- It can't be that difficult.
- Just make the experiments and give them a result in the form of a number.
- They don't care about the details on why it can't be done right.
- For some reason a false result is better than no result.
- They just want a number. How hard can it be?
You think they sound like a bunch of whiners? They are!
Usually I say 42. That's the meaning of life, the universe and everything.
What? You don't like the number? How about π or e? Those are perfectly good numbers.
No, I'm not trying to be funny or provoking. I'm telling you that they are trying to make you break the law, and take the blame if they get caught using falsified results.
I can understand why some people want to please the bosses and co-workers with a can-do attitude. I once had a colleague who wanted to test twice on the same subject. It was well established that the test was destructive for both product and test subject, so the first test would change the test subject so a second test would be affected to the point where the result wouldn't be reliable in any way. He was told that it was the wrong way to do the test by me and another colleague, but he just wanted a number. It was all he focused on. The number. He couldn't relate to the fact that he was in reality falsifying his results. This is actually a serious offense. He might as well have used random numbers by throwing a dice. It would have saved time and money and the results were just as reliable.
I can't stress this enough: Do NOT ever falsify results. No matter how much they want a result and you want to accommodate them, you NEVER falsify your results. There is no "just a number".
Then what do you do? You have to do something. There are a couple of options here:
- If you don't have a reliable method at all, you tell them to wait for whatever it is that's missing to be developed. They may bitch and moan, but you NEVER give anyone any results based on an unreliable method.
- You have some material but not enough for a full test or the test takes longer than the time available. Make the tests you can from the philosophy, a few tests done the right way - not the other way around. Why? Because the first type of results is reliable, the other is not.
Some will argue that you can do a "quick and dirty" test and you can give them some valid results later.
Don't ever do that!
You give them some results and tell them that the results may be off or even wrong. They only hear the number. It may not be the ones you gave the results, but somewhere along the line the things you told them about why the results may be faulty are left out and the numbers are what's left. It's human nature.