The first talk on the Project Management/Business Analysis track was by Robert Handler, VP and Distinguished Analyst at Gartner, Inc. For those of you who aren’t in the IT business, Garner, Inc. is a well known information technology research and advisory firm. This talk was heavy on psychology; it occurred to me, as I listened, that I could have made an entirely different career from my current one by combining my psychology studies at Stanford (I leaned toward social psychology) with my IT knowledge and going in more of a business than technical direction. But on to the notes.
Governance is great when you’re the governor; “It’s good to be the king.” But there are many ways it can go wrong, that can be predicted with social psychology and behavioral economics. IT governance assumes rational actors, but people are irrational, and predictably so. Understanding how people are irrational allows for good IT governance. Handler recommends reading Dan Arielly, Ph.D., “Predictably Irrational” and “The Upside of Irrationality.”
How could a rational process (right out of a Gartner research paper) fail? For instance, in principle, you may want to follow a process in which you evaluate solely on merits, fit alignment, value, and desired future state, and use best practices and very rational processes called gates. In practice, it is likely that you will succumb to some of the crazy stuff we do as individuals and as groups or teams, and fail to make optimal IT decisions.
First, the meshuggeneh things we do as individuals:
Bias: Overconfidence
Explanation: Humans are overconfident of their abilities, even in hindsight. An example all too common in the IT world is overly optimistic time estimates.
Bias: Hindsight
Explanation: Your brain takes the good things, and forgets the bad, so that you judge history as more favorable or predictable than it actually was.
Bias: Sunk-cost effect
Explanation: Attachment increases as investment increases. [LG: Some of us may remember encountering this effect in at least one failed relationship.]
Bias: Recency Effect/Availability Bias
Explanation: We place too much emphasis on recent, available data.
Bias: Confirmation bias
Explanation: We tend to gather, analyze, and rely on information that confirms our beliefs. For example, if an article is written such that it balances information supporting a liberal or conservative point of view, and given to both liberals and conservatives to read, liberals and conservatives will each remember most readily the points that support their point of view.
At this point, Handler asked us to think of our area code, and then asked whether Atilla the Hun was born in the first millenium or the second millenium. Few raised their hands for either millenium, perhaps because few, at a professional event, were willing to commit themselves to what might be the wrong answer. But, for the record, my own thoughts as he asked the question were: “First millenium. No, it has to be the second millenium. Why can’t I remember this correctly right away? This is something I should know.” In fact, Handler had set us the question to illustrate the next bias:
Bias: Anchoring
Explanation: We allow our initial reference (e.g. ballpark estimate) to anchor our frame of reference (even if it’s something like your three digit area code making you more likely to think Atilla the Hun was born in the first millenium). Many ISTJs in IT don’t understand this, and think that if their initial estimate was correct within an order of magnitude, it’s close enough. But the business people to whom they gave the estimate (without a margin of error) were expecting it actually to match reality.
Bias: Illusory correlation
Explanation: We see patterns that don’t really exist. We then can’t be persuaded otherwise, even by experts who have checked the correlation. [LG: I think, for example, of the long and repeatedly debunked story of the association between vaccinations and autism.]
Bias: Egocentrism
Explanation: We give more credit to ourselves, and more blame to others, than is deserved.
These are things that normal human beings that are somewhat functional do.
Group biases (and how to use them to your advantage): Handler recommends Robert, The Science of Persuasion, and Cialdini’s Six Principles.
Contrast Principle: Decisions are based on comparisons. This is used in sales; for example, if a comparison is set up between light, standard, and platinum versions of a product, people will gravitate toward “standard.”
Principle of Reciprocation: People are willing to return favors. Don’t squander that opportunity. For example, people will accept more governance in return for favors.
Principle of Scarcity: People value what is perceived to be in short supply (even if that perception is false).
Principle of Authority/Credibility: People attend to the requests of experts and authority figures. Handler discussed how sometimes the same thing will be more readily believed if an authority figure repeats it.
Principle of Social Proof: People trust their perceptions when they’re consistent with those of others.
Principle of Consistency: People act to support previous decisions they have made. If you can get people to take actions consistent with what you want, under circumstances where they don’t see themselves as coerced, they’ll adjust their beliefs to match their actions.
Principle of Liking: People will more readily comply with a request if they like the person making the request. Admitting an honest, truthful flaw can help with this. [LG: I'm reminded of the debate in which Obama, pressed for a flaw, admitted to a messy desk, and got praise for admitting an actual flaw, rather than one of the "flaw that's really a virtue" variety.]
Handler discussed an experiment in which 80% of the people who had gotten a response of “No” were still happy, if they saw the process as fair and transparent, but that percentage dropped to 37% if the process was not as fair.
Crazy stuff groups or teams do:
Group think: Going along to get along (Irving Janis).
Process losses: “Failing to achieve group synergies” (Juan Steiner) Conformity, in group/out group, domination, lack of individual accountability, information stinginess, filtering, role confusion, etc.
Organizational/cultural barriers:
Yes: False public consent followed by private dissent.
No: Silence followed by …
“Smart Talk,” Jeffrey Pfeffer, MD
Veto power disables all but “no”
Maybe: Analysis paralysis
Combat biases through awareness. Promote fairness through process. Leverage the culture, and create a diverse set of participants to limit biases.
Recommended for further reading: Why People Make Irrational Decisions and How Project Management Leaders Can Deal With It
An example all too common in the IT world is overly optimistic time estimates.
That is so dang true. My policy was to take my first estimate, double the numerical part, then increment the unit type.
So if my first estimate was that I could do something in a week, I’d give an estimate of two months.