If you haven’t seen it already, check out a “leaked” version of Valve Software’s “Employee Handbook“. It reads like an INTJ designed how the company would work – fantastic!
I especially loved the bit about how they view hiring: “More important than breathing”. True^2.
They’re currently hiring lots of positions, including “Economist”. (A compelling reason to get a graduate degree in the dismal science?)
Today I had the opportunity to present to 5 classes at Oklahoma State University. 3 of them were “Intro MIS” classes where we talked about “People factors in IT projects”. The other 2 were to “Strategy and Integration in Organizations”, a senior-level management class, where we talked about “When we fail”. Hopefully at least 1 of the 400 students got something useful out of today, but I had a lot of fun.
It’s (thankfully?) not often I present to five groups *in a row*, but it did give me a chance to observe a couple of things that I thought were interesting – maybe if you find yourself in front of a group of people in the future, you might find some of this helpful:
Okay, Clips are up 87-84, so I’m going to call it a night and cheer for the Thunder.
I’ll be back on campus in Stillwater next week to talk to hundreds of business students in the “Intro to MIS” classes. I’ll be giving my “People factors in IT projects” presentation, which is really a trojan horse for me to bring up the concept of “complex systems”. Some people might see the “People factors” title and be afraid it’s going to be dry, but I think it’s actually a pretty fun one – how bad can it be if it starts and ends with discussing Thunder basketball and Nick Collison’s contract extension from a couple years ago?
In my “The ends are mean”, I tried to explain why I don’t think that the results of something necessarily tell us a lot about the people involved. Sometimes things go wrong for people who did the absolute right thing… and sometimes things go right for the people who did the absolute wrong thing. So for us to use the rule “the result tells us what we should think of a person’s decisions/actions” can be misleading.
Now, I love me some behavioral interviewing. I like asking people of specific examples of times when they had to solve a problem or demonstrate an attribute we’re interested in. The behavioral interview can be another great data point to use in deciding who we should hire. However, I take a little bit of an issue with the “traditional” approach of a good behavioral interview answer.
If you don’t know, most people learn that a good answer is a “STAR” – that is a good answer covers the “Situation or Task”, covers the “Action” taken, and finally the “Result” of the action. Cover all 3 of these pieces, and you have attained the lofty “full STAR” answer!
I don’t like two things about this approach:
If we buy into the idea that the result doesn’t tell us whether the decision/action was good bad, then we can see that we have to dig into Why a person did what they did. Did they do their homework? Did the jump to a logical conclusion? What alternatives were considered and why did they choose what they chose? I want to find out all of this information in order to find out if this is the type of person we want to hire.
Now, many of you interviewers might already be getting this info instinctively, as part of the “Situation or Task” description, but too many times I’ve seen the “why” glossed over since “W” isn’t in the “STAR”. But, if it helps, please think of this as the STAR++ method, or as the “STWAR” (rhymes with “GWAR”) answer.
I’m assuming you’re hiring for the long-run, in which case, it’s better for your hire to be good than lucky.
One of these days, I’m going to make a list of the things I believe. But I want to elaborate on one that’s been bugging me lately. Namely, “the ends are mean”.
When somebody tells me, “well, it turned out okay, so it didn’t matter”, I hear “I’m either lazy, stupid, or unwilling to admit that I screwed up.”
I think most people can agree that even if you “do the right thing”, things won’t always turn out right. There could be something you didn’t know, something you didn’t expect, some unintended consequence – but that doesn’t mean that you screwed up. It’d be silly to punish people who did the best they could in the situation they were in, and if you ever have a manager that doesn’t get that, you have my condolences.
But what about the case where everything turned out okay? Does that mean that the decisions and actions of those involved were correct?
So, I’m going to paraphrase an example from London’s Metaphysical Society, circa 1877, given by William K. Clifford. In “The Ethics of Belief”, he tells the tale of a ship’s owner who sends his ship a-sail with a boatload of passengers. The shipowner believes his ship to be sound before the voyage, but only because the ship had sailed without incident so many times before. However, during the course of the voyage, the ship sinks and all passengers are lost.
Is the shipowner responsible for the death of all the passengers? Surely. He was grossly negligent by not investigating if his ship was seaworthy before the trip. He simply hoped for the best… and the worst happened. Any of us would expect the shipowner to know that his ship was safe before we board. The shipowner is guilty.
But the story gets interesting when we say, that instead of the ship sinking, the trip went off without a hitch. All the passengers had a pleasant voyage, and the ship made it to port, none the worse for wear. In this case, was the shipowner guilty?
Yes. Yes he was.
In neither case did he actually know that his ship would be able to make the trip. Yet, he loaded up a bunch of people and basically rolled the dice with their lives. Just because the ship didn’t sink doesn’t mean he wasn’t guilty – it’s just that he wasn’t found out.
The point of this story? That the ends don’t tell us if someone did a bad job OR a good job. For us to really know how a person did, we have to look at WHY they decided what they did, and if they believed what they believed for good reason.
The ends are just mean that way – they allow us to rationalize what we think about someone’s efforts, without having to investigate what really happened.
In which case, we’re no better than the guilty shipowner.
They’ve just posted the contest rules for the online class I’m taking at Udacity:
It’s been a great CS 101 class where you learn basic CS concepts while building a search engine, so if you have ever had an interest in learning how to program, I highly recommend it.
Now I just have to think of a way to revolutionize search so I can win a trip out to Palo Alto…
Had a great time presenting to BAP tonight at OSU. Really enjoyed meeting some students and getting to visit with the faculty as well. I hope they had at least half as much fun as I did.
I kept this particular presentation pretty high level; not nearly the depth I’ve gone into before with an Industrial Engineering class, for example. We went through several “fun” examples, but I got the impression that they kept expecting me to really drill down into one of the examples. And I made the mistake of not explicitly saying “I’m talking about ‘systems’, not ‘computer systems’, so it took a few of them a bit to pick that up. Next time I give this presentation, I may switch it up to see how it goes. But the cow puns killed. (You probably had to be there…)
Now I think I’ll watch the remainder of the Thunder/Lakers game to “unwind”. (Okay, maybe not my best plan ever.)
“This event was the best MISSA-coordinated event we have held during the past two springs.”
OU “Price Matters”, Vol. 6, Issue 2 (see page 6)
Tomorrow I’ll be in Stillwater to talk to Beta Alpha Psi. Yes, that’s an accounting group, and I’m an “IT guy”. Yes, that’s OSU when I bleed crimson and cream. But it was a really smart group to talk to last semester, and since they were nice enough to ask me back, I’m going to bring out the good stuff.
The name of the talk is “Problems of Purview: Applying Systems Thinking in a Complex World.” If they only leave with one idea at the end, I want it to be that we have to be careful that our picture is “big enough” when we’re making decisions – if we focus too narrowly, we’ll end up solving problems in a way that makes things worse overall.
Purview problems pop up all the time if you look for them – my favorite examples being, of course, sports examples. How many times will a team sign a player with incredible individual stats, but that’s more than offset by the negative impact that player has on the team’s ability to win? (Yes, I’m thinking specifically of Carmelo Anthony… but maybe I’m just still bitter from him ripping through the entire Big 12 in the NCAA basketball tournament…)
And when we make a bad decision, it tends to stay bad. As people, we seem to have an incredible ability to pick out one piece of information and give it entirely too much weight. (Psychologists call this “anchoring”.) And once we’ve locked in onto what we want to think, we’re hesitant to let it go, both because we’re trained to be consistent with our earlier words and actions (so we’re not a “flip flopper”), and because once we have an opinion we tend to only pay attention to data points that reinforce what we already believe. (“Confirmation bias” is the name for this one. Or “Fox News”.) And even without all that, there’s still the problem that admitting a mistake would often be worse for an individual than if they just try to get by without anyone finding out… even if admitting the mistake, correcting it, and learning from it would be better for the big picture.
And that brings us back to purview. It’s how we figure out what we should pay attention to, and how we should judge the outcomes of our decisions. Yes, expanding our purview makes us realize that problems are often more complex that we’d care to admit. It also means that the “simple” solution we have in mind probably won’t work. But it’s the first step towards us making effective decisions that make the “big picture” better. And the future could use our help…
…and a little less Carmelo Anthony.
I updated my blog software, and managed to delete all my prior posts.
This blog obviously needs a change control process. [/kidding]
