As scientists, we value truth and accuracy. As a manager, the tool you're using to measure is you yourself - and we all have inherent biases that will throw off your readings. The good news is that learning about these biases will help you correct for them and understand your surroundings a little better.

Experience Bias

You have the background of an engineer, or a scientist. That gives you certain biases. For example, let's say that someone ask you to solve a general problem like "let's cut costs". If you're awesome at technology, you might first look for a technology solution first, whereas someone with a finance or HR background would approach the problem differently. To quote:

"In times of trouble, go with what you know."

Availability Bias

You have a set of experiences that you can recall. So, you might make a decision based on something that's happened to you - a remark someone made, a funny situation you witnessed, and so on. However, you're just one person, and your sample data is seriously non-representative, so there's bias towards your own recollection. Big caveat - don't let this stop you from going and talking to people to get some data samples to understand what's going on. Some data is better than none at all, and you're never going to have time to get all the data you need.

Anchoring Bias

Go and find a friend of yours, and ask them to write down the last three digits of their phone number on a piece of paper. Then, a little later, ask them to estimate when Genghis Khan was born. The chances are pretty good that they'll guess a year that has three digits. If you do this experiment with enough people, you'll notice that a statistically significant number of people make the same mistake.

This is just one crappy example of the anchoring bias. There are many more important ones. For example - the first time you meet a new minion and they show up late, it's incredibly easy to assume they're generally not punctual - even though you only have one data point. Another - if you don't like the people you've met from one particular college, you might assume that the next person you meet from that college is also an ass. Again - you might well be right, but be aware that your opinion carries bias.

Calibration Bias

Go and find a friend of yours, and ask them to make an honest assessment of where they stand in ability compared to everyone else in the office - the top 10%, the top 25%, middle, bottom 25% or bottom 10%. Find someone else, and repeat until you have a fair few responses. Collate these, and the chances are pretty good that your data can't represent reality, because a hugely disproportionate number of people have put themselves in the top 10% or 25%.

It turns out that we suck at assessing ourselves. There's a reason for it - we may need to believe that we're pretty good in order to have pride in our work, and that's a good thing. But trying to translate belief into objective assessment is a minefield.

Correction

The moral of these stories - make sure that you're not reporting bias as data. The more aware you are of your own biases, the better you will become at measuring the world around you. And finally, as with any complex bias, understand that even with correction, you're never going to be completely accurate.