yevster commented 6 years ago

(Please assign to me - I don't seem to have permissions)

yevster commented 6 years ago

Here is the first draft. I'd love to get some comments! (Will probably need to replace hyperlinks with footnotes for the printed version)

Exercise: The Bias Checklist

Estimated time to complete: 15 minutes first time, 2-5 minutes subsequently.

Materials necessary: Editable copy of this text, or print-out and a writing implement

Exercise type: Checklist

Bias is the single greatest threat to the open organization. This is no exaggeration. In traditional organizations, responsibilities for evaluating ideas, strategies, contributions—even people—typically fall on (presumably) trained managers. In open organizations, that responsibility rests with contributors of all sorts.

"In organizations that are fit for the future," writes Jim Whitehurst in The Open Organization, "Leaders will be chosen by the led. Contribution will matter more than credentials [...] Compensation will be set by peers, not bosses." According to Whitehurst, an open organization is a meritocracy: "Those people who have earned their peers' respect over time drive decisions." But the way humans allocate their respect is itself prone to bias. And imagine what can happen when biased decision-making results in the wrong leaders being chosen, certain contributions being over- or undervalued, or compensation being allocated on something other than merit.

The following checklist covers several documented phenomena that, sometimes unconsciously, skew the decision-making practices.

The bias checklist

☐ Has this person done a favor for you? If so, you are more likely to impulsively do something that feels like a favor to them. This principle of reciprocity is very powerful. This is why when charities send you solicitations by mail, they will often include address labels, calendars, other cheap goodies. When you feel like you've been given something, you are more likely to find yourself giving something back. Reciprocity is especially dangerous in an open organization, where the odds of two contributors having to evaluate each other's contributions are drastically higher than they may be elsewhere. (Positive Reciprocity)

☐ Has this person done something unfavorable to you? Reciprocity works both ways. Even if you can swear up and down that you're not the type who bears a grudge, if someone has previously rejected contributions from you, be especially alert for the desire to reject a contribution from them. (Negative Reciprocity).

☐ Has someone else done something particularly favorable or unfavorable to you? While we're on the subject, be on the lookout for Generalized Reciprocity - a tendency to "pay it forward", or to "do onto others" what someone else has done to you. There has been evidence of a higher tendency to pass on negative or unfair behaviors than positive ones, so if you feel "wronged" at the moment you're evaluating a contribution, this feeling can influence your evaluation. One study found writing a message to the original perpetrator of an unfair or unjust behavior explaining your feelings can reduce your impulse to pass that unfairness onto others.

☐ Have you already done something nice (however small) for this person? This one's a bit more subtle than reciprocity, but no less powerful. Subscribe to the mailing list of a major political party, and you will be inundated with requests to fill out a survey or make a small donation (e.g. $1). These are effective solicitation techniques, because once you have done something positive for someone, you become more likely to do something even bigger for that person. This phenomenon is sometimes called the "Foot In The Door" technique. If you have spoken favorably of an idea, a contribution or a contributor, or approved past work of the contributor, you will be likely biased in favor of that contributor.

☐ Have you previously spoken ill of an idea, a contribution, or a person? The same need for consistency with oneself that powers the Foot in the Door technique also works in the negative. If you have previously treated an idea, a contribution, or a person with disdain or have otherwise acted unfavorably toward a person, you will be predisposed to continue to act in that way.

☐ Is the person good looking? It has long-since been documented that the attractiveness of someone performing a task influences our evaluation of how someone performs that task. This is an example of the Halo Effect - a phenomenon whereby one attribute of a person leads us to assume or overestimate other attributes of that person.

☐ Is the person successful or renowned in unrelated areas? This is another example of the halo effect. For example, Stephen Hawking's anti-AI pronouncements received wide press coverage across the world, even though all of Stephen Hawking's education and accomplishments are in physics and cosmology. Another example comes direct from the original The Open Organization. Mr. Whitehurst writes about how Gavin King (creator of the Hibernate Framework) was allowed to embark on a new project to create a new language for the JVM. Today, that language, Ceylon, languishes in relative obscurity, compared to competing JVM languages Scala (which has existed for years prior to Ceylon) and Kotlin (released at about the same time). Data frameworks and programming languages are examples of substantially different areas, so it seems fallacious, in retrospect, to assume resounding success in one area would portend any success in the other.

☐ Does a person or a contribution have similarities to someone or something you respect or admire, or dislike or oppose? When making evaluations, we have a tendency to compare what we see to similar things from our own experience and pass judgment based on similarity to those things. This is why, for example, political activists often like to compare the people they oppose to Hitler. None of those being compared have objectively committed wrongs even remotely resembling Hitler's mass murders, but the comparisons are driven by the intuition that any similarity to Hitler presents the target in a negative light. (Similarity Heuristic)

☐ Does this person meet or violate your norms for how a successful person in a looks? Whether or not we admit it to ourselves, we all have biases and stereotypes for what a successful person in a field looks like. For example, when asked to picture a successful software engineer, many of us will think of a white or Asian male. While gender and racial biases have been widely discussed, awareness of them alone is not to compensate for their effect. We have to make these unconscious biases conscious and force ourselves to second-guess our judgments in their light.

Avoiding biases

Knowing about a bias may not be enough to prevent its effects. Many of us just aren't that good at correcting ourselves. This is why it's crucial to take steps wherever possible to avoid biases:

☐ Evaluate contributions anonymously. Orchestras, for example, will often audition musicians without seeing them. This practice has reduced the gender gap in major orchestras long before it became adopted by televised talent shows.

☐ Define structured and rigorous evaluation criteria for people and contributions.

Opening up on biases

Knowing about, and even anticipating, our biases is the first step in the journey to overcome them. But it is not the end.

Working through the checklist above, for instance, one may be attempted to acknowledge the possibility of the biases—but ultimately deny experiencing them. Bad news: This may itself be a bias talking. Humans have been found to be overconfident in their abilities in virtually all areas, from driving to investing. So there's every reason to assume we may be overconfident in our own ability to screen ourselves for biases.

Fortunately, the structure of the open organization can help us overcome overconfidence. In The Open Organization, Jim Whitehurst writes about the principle of 360-degree of accountability, meaning "you are accountable to everybody." 360-degree accountability empowers participants to recognize biases not only in themselves (which is difficult to do) but also in others (which is easier). "My job is [. . .] to take questions and feedback and engage associates in a conversation about the decisions we're making as a team," Whiterhurst writes. Thus, in an open organization, it should be appropriate—indeed, encouraged—to ask, for example, "have you thought you might be biased in favor of X due to your prior positive statements about it?" The answer to the question, rather than a "yes" or "no," should be the concrete steps one has taken (or will take) to avoid bias. A reasonable response may be: "No, I haven't thought of it, but now that you mention it, let me ask Y, who has not dealt with X previously, for a neutral perspective."

When bias is acknowledged, discussed, and counteracted proactively and systematically, an organization becomes more deeply empowered to scale trust, responsibility, and accountability across the organization. An open organization relies on the soundness of human judgment across its ranks. Taking every possible step to make this judgment as objective and trustworthy as it can be is therefore absolutely imperative.