Wednesday, December 2, 2015

Potentially Dangerous Biases in Policy Planning Meetings

Administrative meetings are are a common part of many people's professional lives.  Teachers, academics, health care workers, engineers...we all at times are required to attend meetings, some of which have a theme of planning for the future, and of changing the way things are done.

Quadrant 4
The importance of reflective, proactive planning is very high.  I often refer--with my patients-- to the four "quadrants" described by Stephen Covey (1989) in his very popular book, The 7 Habits of Highly Effective People.  He wisely points out that many people are devoted to what I call "quadrant 1", which is taking care of short-term, important tasks.  For many workers this could be answering calls, or dealing with emergencies.   While quadrant 1 is important, it often consumes so much time and energy that there is nothing left for long-term planning.

 It is an obvious problem when people are spending too much time in quadrants 2 and 3:  short-term or long-term unimportant tasks (such as engaging in unhealthy habits).

 But Covey's simple but great point of wisdom is that the highest priority ought to be setting aside time for "quadrant 4":  long-term, important tasks.  Many of us are too busy with the urgent matters of the day to make time for these things.   If more attention was given to long-term, creative, imaginative planning, the short-term important tasks could often be taken care of much more efficiently, leaving more time for deeper, higher quality, more meaningful work. 

So, reserving time for meetings, in order to plan for change, can be vitally important, even if they take us away from our important daily work.  

Meetings have Costs
However, meetings have costs!   Many meetings are unproductive and boring.    The cost of the meeting is not only financial (i.e. to pay for the salaries of those in attendance, plus any costs of using the space, or getting food, etc.), but more importantly there is a cost in terms of the work that could have been done if the meeting did not take place.   If there are 40 health care workers in a 3-hour meeting (which is really 4 hours, if we consider the time to get to the meeting place and back), this represents 160 patient-hours of health care which was not provided.   For some types of brief psychotherapy, which consist of 8 sessions over a 2-month period, the meeting has a cost of 20 suffering people not receiving help.    An added cost could be a decrement in morale caused by wasted, unrewarding, unproductive time.  

Cognitive Biases & Heuristics in Meetings
One of the biggest problems with meetings, in my opinion, has to do with the manner in which they occur.  Daniel Kahneman, the Nobel-laureate psychologist and behavioural economist, has written about the strong cognitive biases that occur in group interactions such as meetings.

If a meeting begins with a number of presentations which the rest of the group watches, then these presentations are likely to have more influence than is rationally expected.  This is particularly true if the presenters are seen as "experts."  This is in keeping with other social psychology work, such as that of Cialdini, looking at factors heightening pursuasion.  These six factors have strong, often unrecognized power to manipulate decision-making.  The factors are natural phenomena, and not necessarily "bad"--in fact, as Kahneman points out, these phenomena (heuristics) can help us make important decisions more quickly, because they are "cognitive short-cuts."  For example, if we have a trusted expert educating us about the use of a new technique, it may spare us from having to take a much longer time researching and learning about the technique on our own.  But unfortunately, these persuasive factors are also used manipulatively by marketers, and may also enchant us so much that we do not do the necessary critical thinking about ideas that are presented persuasively to us.  Furthermore, even if those involved with the heuristics do not have any negative or manipulative motive, the biases exist nevertheless.  They cause a potential harmful impairment in rational decision-making, even without anyone being aware of it! 

Cialdini's Elements of Persuasion
Here are five of Cialdini's six elements of persuasion, as manifest in a typical professional meeting:

1) Reciprocity (we are more likely to be persuaded if someone has just given us something of value, since we are likely to feel grateful and indebted):  meetings may begin with tasty snacks & coffee.  We may get our own personalized name tag, which makes us feel valued and important.  We may be entertained by a friendly, humorous, engaging presenter.  It makes us primed to cooperate!  Because we have invested time and effort to come to the meeting, we are probably more apt to be "receptive" to the ideas presented, so as to compensate ourselves for the effort.   This mechanism has been well-researched before in other studies of how people form allegiances to groups via strenuous tasks necessary to join, such as in fraternities or the military.

2) Consistency & Commitment (if we have already agreed to a certain component of an idea, we are more likely to be persuaded to go farther with it in the future, otherwise we could feel as though we are contradicting ourselves or being inconsistent).   This can happen when there are repeated meetings, perhaps spaced out weeks or months apart.  If people have agreed to certain aspects of a plan, they are much more likely to continue agreeing with the same plan, and to go farther with it, even though the plan may at this future point be inappropriate or worthy of questioning.  In sales, if you can convince someone to sign a petition, and to wear a sticker touting some kind of apparently wholesome idea, you are much more likely to be able to convince that person to contribute their time and money to the cause later on, even if that person would otherwise have had a lot of doubts about the plan.  In intimate relationships, if you have said "yes" to someone a number of times previously, it can be harder to say "no" to a similar future request; this often leads to people staying in an undesired relationship, or tolerating negative relationship behaviours, which they regret afterwards. In political movements, initial commitment to a group's policies can have an "inertia" which makes members go along with the group's subsequent negative actions (extreme examples of this phenomenon can be found in various political movements of the past century). 

3) Social Pressure (if other peers are conforming to an idea, we are more likely to as well).   Fellow members of the meeting may be smiling and nodding at the presenter's ideas.

In one meeting, in which electronic gadgets were discussed as a therapeutic modality (not only therapy apps, but also things such as covert monitoring of online activity to monitor health), one could notice that many people in the room had their heads buried in their laptop screens (an irony, speaking to the interpersonal detachment caused by electronic devices!)    Murmurings could be heard from nodding heads, of "yes, yes, therapy apps, yes, what a good idea!"   An objection to the idea of using electronic gadgetry this way would be perceived by the group as a jarring and unwelcome dissonance or even a disrespectful action. 

Also, another tactic in policy presentations can be to show that other large groups across the world are adopting the idea in question.  We may be shown that professional peers, perhaps in Australia, Germany, or Japan, have adopted some new policy, and are thriving with it.   It makes us feel like joining in, or even catching up to our Australian friends!  In an environment with a lot of social pressure, one would feel deterred from objecting to an idea, because of a concern that it might offend our friends, or that it might cause us to look reactionary or old-fashioned!  Clearly, it is important to learn about what our peers are doing, but the problem is when the social pressure becomes a biased cognitive short-cut, leading us to not question or think critically about the matter in question. 

4) Liking.  If the presenter is friendly, likable, humorous, and kind, we are more likely to be persuaded by his or her ideas.  Many salespeople and presenters are indeed friendly, likable people, which gives them a big persuasive advantage!  The trouble is, supposing we had some very shy or less social people in the group, who might have a more abrupt or chilly manner.  Yet, suppose these people had very, very good ideas to contribute to the discussion.  These voices would carry much less persuasive power because of the influence of this factor.    Sometimes quite harmful ideas have been "sold" to groups, because of the likable, engaging demeanour of the person presenting them.  Once again, one does not have to look far in political history to find unsettling or even frightening examples.  

5) Authority.  At the beginning of most presentations, we are usually shown a lengthy list of qualifications that the presenter has.  This might include degrees from major universities, academic awards, publications, etc.  Such a list conveys that the person we are listening to is an "expert," and that we should perhaps trust more strongly what is said.  Often, this is a useful cognitive short-cut.  But in many other situations, the qualifications may have nothing at all to do with the message being conveyed in the meeting.  I can think of some highly-paid professional experts who have much to say about clinical care of patients--but their area of expertise relates to extensive publication of research articles, and extensive attendance of international meetings.  The "expert" may actually have much less experience with clinical care than a humble colleague-- who lacks the impressive credentials-- in the next office, who spends his or her day seeing patients, while not publishing or attending meetings at all.   Qualifications should be respected, but we need to watch out for the impressiveness and "authority" of a presenter causing irrational biases and influence in our decision-making.   Another manifestation of the authority bias has to do with the sincere well-meaning nature of many presenters.  They may be very excited about their new idea.  They may be asked to give speeches and presentations about it across the country, and may even be getting famous as a result.  Their work may be grounded in a passion and commitment for their work, and also the enjoyment of getting acknowledgment and respect professionally.   The trouble is, this very excitement, ambition,  and the existence of professional credit and fame, could lead to a bias in which the presenter becomes overly attached to their ideas, and too eager to push the ideas without critique.  The audience may also really respect the presenter, and the integrity of the presenter's motives, but as a result may be too willing to agree respectfully with the ideas, instead of questioning them rationally. 

Behavioural Economic Biases

Other biases in meetings:  some of these biases are described in behavioural economics.

6) Joint vs Single comparison bias:  in many policy discussions, there may be a presenter excitedly showing results based on a new system or technique.  The results may be very positive.  But very often, there is no fair control group.  When we look at just one thing without comparing, it causes our assessment to be distorted.  Kahneman's classic example is of how people would value  sets of dishes:  in one set, there could be 10 plates in perfect condition, plus an eleventh plate which is defective.  In another set there would be just 10 plates in perfect condition, without the eleventh plate.  If people assess these two sets at the same time, they would value them identically, since they both contain 10 good plates.  But if people assess the sets individually, without comparison, they would give the first set a much higher value than the second.  The existence of the defective plate would make it seem like the whole set has something wrong with it.   In presentations, we often see only "one set of plates."

7) misuse of statistics.  There are many ways to misuse and misrepresent statistics (either deliberately or inadvertently) which ends up unfairly favouring a policy agenda.  A couple of simple, common examples occur to me, pertinent to mental health care:

a) no control group for self-limiting phenomena.    This type of data gathering often occurs in emergency services.  A person in crisis may present to an emergency room.  A symptom questionnaire is administered, showing very high symptom scores.  In the emergency room, the person may be lying on an uncomfortable stretcher in a noisy hallway for hours, be interviewed five or six times by nurses, ER physicians, residents, and then a psychiatrist.   Perhaps a different psychiatrist the next day.  The whole ordeal could be frightening, very uncomfortable, or even frankly traumatizing.    Yet, after three days, the same symptom questionnaire could be administered again, showing a dramatic reduction in symptom scores.  The person would be discharged home, and the data would appear to show that the emergency service was functioning very well.

 In this case,  the person improved mainly because of the passage of time.   Possibly the emergency room offered some measure of safety, and perhaps even some empathic support.  But mainly, it would have been the passage of time for an acute crisis to settle down.   Many of the features of the emergency room stay could have been frankly harmful, yet the data would not show this.   In order to show whether the emergency service was indeed helpful, one would need to have a comparison group, offering a different type of experience (for example, a comfortable, cozy, supportive, community-based environment which had rooms for people to stay overnight).    Probably, the alternative service would show better results, and also a much lower risk of severe adverse effects (such as a patient feeling traumatized).

b) short-sightedness.   Many treatment trials are short term, leading to rapid improvement in symptoms.  If we use potent sedatives to treat insomnia, for example, we are likely to see almost instantaneous improvement in symptom scores.  But if we follow this medicated group for a year, we would most likely see that the medication group was doing no better, or often worse, than a control group.   We see this type of bias all over our culture.  Evidence about the dangers of greenhouse gasses or other pollutants in the atmosphere require data spanning decades.  Short-term data may lull us into believing that there is no risk, or that a harmful policy is benign or beneficial.      In psychotherapy, there is indeed evidence that psychotherapy (of all types)  can have long-term beneficial effects for many people.  But a short-term view would unfairly discourage the use of longer-term psychotherapy.  It is technically easier to gather short-term data, and such data have less technical error (due, for example, to fewer patients lost to follow-up, etc.).   But just because some types of data are easier to gather, and appear to show a positive result, we should not be biased into thinking that long-term approaches are invalid. 

 Zimbardo's Work on the Social Causation of Negative Behaviour
 There are dramatic analogies to be made, looking at historical events.   Philip Zimbardo has written extensively about the manner in which disasterous, or even "evil" events can occur, due to the dynamics of a group context.    In groups, if there is a pre-existing majority position, the meeting can become a time in which the majority position becomes "celebrated" and pushed further, due to contagious group enthusiasm, particularly if the group features a type of hierarchy or a power differential.    A dissenter in such a meeting could be viewed as "negative" or "difficult," and there would be social pressures which discourage contrary points of view.  If one of themes in the meeting is "collaboration," then any comment which seems negative or "non-collaborating" may be seen to be violating the group ethos, even though the comment may be of high ethical import.     This phenomenon occurs in all human groups, including those in which the members are highly intelligent, sophisticated, and "civilized."   The decision making in groups leading to some of the world's great wars of the past centuries have often involved contagious enthusiasm from a strong majority view, in which dissenters were initially suppressed due to social pressure alone, then later due to threat of punishment.

The Brainstorming Myth

Another common event in policy planning meetings is "brainstorming."  This may be a process in which the attendees are split into groups of 5-7 people, and asked to freely, imaginatively, constructively express positive ideas for change, perhaps building on the ideas of others, with notes being taken, and with no criticisms allowed.  But more recent evidence has shown that brainstorming techniques could be inefficient in a variety of ways.  The same social dynamics outlined above could subvert the process, discourage dissent, and discourage a challenge to authority.   It could be much more fruitful, in a group dynamic, not to have "brainstorming," but rather to make strenuous efforts to permit a truly free debate, in which members have fully adequate time to prepare in advance.  In some meetings, there may be opportunity for debate, but the members have had no time to prepare.  This would be like having a court hearing for an important case, in which there would be ample opportunity for debate, but in which one side would be allowed weeks of preparation, while the other side would have to improvise on the spur of the moment with no preparation at all! 

Making Meetings Better

What can be done to improve meetings and to reduce unwelcome cognitive biases?

1) participants in meetings can be asked to contribute ideas before the meeting begins, perhaps anonymously.

2) free questioning, with uncensored feedback, must not only be encouraged, but the right to offer such feedback must be fiercely protected.   This is the foundation of a free society, and of a modern educational process as well. 

2) I do not believe that a decision-making meeting should feature educational presentations at all.  It is simply not possible for a group to critically appraise educational material in an unbiased fashion in the space of only a few hours.   If important group decisions are to be made in a meeting, the presence of a presenter causes very strong influence on this.  We assume that the presenter has good intentions, may be very wise, etc. -- but if there is no opportunity for detailed critical appraisal or debate, then the presenter is given automatic power to influence the group, through various mechanisms of psychological bias.

 Imagine attending a family meeting to decide on a vacation spot, or on a new school for the children, in which there is a presenter advocating some strong point of view on the matter who introduces and leads the meeting, complete with video display and free pastries! There is already a conflict of interest if a group facilitator already has a specific change agenda in mind.  The facilitator is then not really a facilitator anymore, but is a lobbyist.  It is especially harmful if the people attending the meeting assume that the facilitator is neutral, without knowing that there is a strong lobbying effect going on. 

The presenter in our hypothetical family meeting might argue a very convincing case for boarding schools, or home schooling, or on the wonderful possibilities of vacations in Mexico!   But would we want family decisions to be influenced by such a presentation?  My point here is not that presentations are bad -- in fact, they could be very useful and informative -- but the problem is when single presenters have the power to influence a group's decision-making agenda, without fair opportunity for critical analysis, and for other points of view to be fairly and thoroughly explored.   It is not enough to allow an hour during the meeting itself for critical debate!  Many of the arguments presented by a sophisticated lobbyist could require weeks to study.

One of the strengths of the modern legal system is of allowing a fair opportunity for free debate before a decision is made about a legal or ethical matter.  We would never have a situation in a legal proceeding in which the person running the meeting (e.g. the judge!) would give a sales presentation encouraging a particular decision, with no opportunity for the different groups to study the proposal in detail (usually, with weeks to prepare), before decisions are made. 

3) In many cases, meetings do not need to occur at all, as the cost of them (as outlined above) may greatly exceed the benefits.

4) Great care should be taken to minimize bias of all types, and to encourage any marginalized or "hidden" voices to be heard.  In mental health care debates, there are many with strong opinions on the matter.   In many public health analyses, there could be broad population surveys to decide upon future policy.  This is reasonable.  But in many cases, those who have actually accessed and benefited from long-term therapy are not heard from at all in the surveys.   This is analogous to making major decisions about cardiovascular health care by interviewing thousands of random members of the public, while not interviewing those who have actually had heart surgery.   People could conclude from such a survey that a rapid-access, short-term exercise program is what people really want for cardiac health, and that a heart surgery program is far too expensive and inefficient.  But this conclusion would be the result of numerous layers of bias, as well as from an inadequate understanding of the experiences of everyone involved intimately in the issue.  

5) "Pre-mortem"  Another of Kahneman's good ideas about decision making, particularly in group settings, is to have a "pre-mortem."  This means imagining that a particular decision has been made, but imagining further that the decision has led to very negative or even disastrous results in the future.  The exercise is to describe ways in which this negative result could have come to be.  The advantage of a pre-mortem could be to use the group's energy not simply to drift enthusiastically into contagious cognitive biases favouring a particular decision, but to work at exploring hidden risks which the group's enthusiasm would previously have been blind to. 

6) "Non-meeting" meetings:  I think it is a great idea to set aside and pay for times in which employees and colleagues can engage in structured healthy activities, such as an exercise class, a walk in the forest, a fine arts class, or a concert.  While this could seem wasteful and expensive, I think that the benefits for morale, group cohesiveness, enjoyment of work, reduced absenteeism, and work efficiency, could be substantial.  Also this would be an example of self-care on a group level, which I think is an important model for our patients to follow.  If we are doing healthy and enjoyable self-care activities as professionals, it is more likely that our patients will be willing to do the same.

5) Heroism.  Zimbardo concludes his thesis about negative group dynamics by calling for each of us to be a "hero."  The type of heroism he means is to be boldly willing to challenge authority, to speak up freely, even when there is a risk to doing so.  One of the best and strongest aspects of American culture is a respect for free speech (and certainly a respect for heroism).    But this wonderful cultural foundation needs to be constantly nurtured, exercised, and practiced, in order to prevent it from being eroded by other cultural forces.    Be heroic :  speak up!

Cialdini, R. B. (1984). The psychology of persuasion. New York: Quill William Morrow.

Covey Stephen, R. (1989). The 7 Habits of Highly Effective People. Simon & Shuster, USA.

Furnham, A. (2000). The brainstorming myth. Business strategy review, 11(4), 21-28.
Kahneman, D. (2011). Thinking, fast and slow. Macmillan.

Zimbardo, P. (2007). The Lucifer effect: Understanding how good people turn evil. New York.

No comments: