Conditions for Intuitive Expertise

Very interesting paper by two of my favorite authors on intuition and expertise. Here are my takeways from their paper.

1. Is my new work environment stable? That is, in the words of Klein and Kahneman does it have high-validity? High validity is characterized by sufficient regularity that ensures that same processes will consistently produce similar outcomes, and that they are predictable.

2. Does the new environment allow me to have sufficient opportunities to “learn” these regularities through frequent practice? In other words, are my exposures to these regularities frequent or rare?

3. Does the new environment offer accurate and rapid feedback to ensure that I am staying on track, allowing more rapid learning?

If the new environment is insufficiently stable – i.e. has “low-validity” My outcome success will not be predictable but highly variable. It will be characterized by tasks that are not simple or that produce singular solutions. The clues will be imbedded in “noisy environments”. They may be valid, but weak. In this type of environment, I will need to augment my repeated tasks with tools that have statistical algorithms that are designed to both detect and report the embedded clues. I will also need augmented feedback tools so that my “unaided” human fallibilities will not overly bias my decisions.

Skill achievement in high-validity environments will need to be measured by the ratio of successful outcomes divided by the number of trials. In low-validity environments skill attainment is more difficult and success should be measured more on the consistency of applying valid processes, multiplied by both number of trials times trial validity (i.e. difficulty).

Advertisements
Link | Posted on by | Tagged | Leave a comment

What Every organization Owes its Employees

What every organization owes its employees

1. Answering the ‘Why’.
The organization must give each employee a reason behind its passion. It must provide a rational for its existence, a higher purpose for each to commit to. To continually engage in some of the following corporate dialog.

  • Why are we doing this rather than that. What satisfaction are we providing for people? What are we contributing to our world that it cannot do by itself, with its current skills and knowledge?
  • What would we do even if we could make the same amount of wealth by investing our money in an interest bearing fund?
  • What we do satisfies our sense of meaning, and that meaning is a shared meaning as evidenced by the satisfaction given to others by our product .
  • If an organization does not continually, and effectively answer the ‘why’ of its existence, it can only satisfy level 1 of an employees needs; money as a means of satisfying hunger, shelter, and physical survival. If it satisfies any of the higher levels it does so by accident.
  • we don’t try to do everything, because we can’t. But what we can do, and do well, we do, even if we knew that it would cost us dearly to do it.
  • profit is not the goal, it is only a constraint on our attempts at achieving our goal. It can be seen as a test of our goal, a limiting factor on our activities. It is not the maximization of profit but the achievement of sufficient profit to cover the risks of economic activity and avoiding the loss of what we love to do. P. Drucker
  • We do what we do because we really believe it. We do what we do because we cannot image doing anything more important with our life. It is not that we spend all of our waking life thinking about it, its just that given 8-10 hours of every day, its how we want to spend those 8-10 hrs.

2. Answering the ‘What’, and the ‘Where’.

It must explain how and where each employees ‘part’ fits within the ‘whole’. In effect, it must help each employee answer the question of what justifies their being on the payroll. It must ensure each employee knows what part of the organization is downstream from their part, as well as what part of the organization is upstream from their part. Finally, an employee must know how the organization as a whole cannot carry out its ‘why’ (i.e. mission) if their part did not exist, or if they did not performed the right tasks in the right way.

  • You must continually seek to instill a focus on contribution not efforts.
  • You should ensure that the employee ‘see’ the results of his/her job. It need not be the complete whole, but it must be the completed part. They must be able to say with pride. ‘that is my contribution, I am responsible for that.’

3. Answering the ‘How’

It must continually work with each employee to ensure that their efforts are effective in achieving the organizations goals.

  •     Making sure that nothing done at the local level negatively impacts the global level (i.e. local optimization resulting in global     sub-optimization), and conversely, everything done at the local level serves the goals at the global level, (i.e. even if global optimization results in local sub-optimization).
  • Reward skill, but do not punish innovation failures. Encourage each employee to view organizational growth through the lens of ‘jobs to be done’ model. http://hbswk.hbs.edu/item/5170.html
  • It is at the ‘How’ level that measurements need to be developed. You need to ensure that each employee be involved in the design, testing, and application of both the tools and resulting measurements by which he/she will be held responsible to.

The measurements must be able to ‘tell’ the story of employee contribution to the ultimate goal (the ‘why’, or mission) of the organization.  A good place to start thinking about what measurements count (i.e. predictive of achieving the organizations ultimate goal). is to read Chapter 4 of Michael Lewis book, Moneyball. It is here one is introduced to Bill James who after thinking long and hard about the goal behind baseball’s statistics states. “a hitter should be measured by his success in that which he is trying to do, and that which he is trying to do is create runs.” For until the measurement actually tells somebody whether or not their efforts are or are not contributing to the goal, they are worse than useless, for they can actually do harm.

Posted in musings | Tagged | Leave a comment

Why ‘Why’ must come before ‘What’, and ‘How’

One of the great lessons in life is when one realizes that something is no longer a means but is in fact an end. the ability or skill to recognize when your thought process is seen within a larger whole is wisdom of the highest order. in fact as one educator has remarked, ‘education is the science of relationships.’ for human beings it has been generally believed that men and women pursue higher needs that are related ordinally. 
Abraham Maslow developed the theory of hierarchy of needs to explain why humans do what they do from a motivational standpoint. See http://en.wikipedia.org/wiki/Maslow%27s_hierarchy_of_needs
in effect he was the first to use the principle of asking the 5 why’s. http://en.wikipedia.org/wiki/5_Whys
the primacy of answering the ‘why’ before any how, or what is needed now perhaps more than ever in a world where one performs work with, or on only parts of the whole within the organization to which they are employed. whenever one’s efforts, and tasks are focused on piece work we are at risk of tunnel vision and may contribute harm to the organization all the while thinking we are doing something useful. additionally, when one stops asking ‘why’ at level one or two of Maslow’s hierarchy of needs pyramid, one is at risk of living an unfulfilled life, and impairing their ability to lead a more flourishing one. http://www.theguardian.com/lifeandstyle/2012/feb/01/top-five-regrets-of-the-dying
who hasn’t asked at one time or another ‘why am I doing this?’ we all have. in fact as John Medina wrote in his book, Brain Rules: 12 Principles for Surviving and Thriving at work, Home, and School.’….if you want to get the particulars correct, don’t start with details. Start with the key ideas, in a hierarchical fashion, form the details around these larger notions.’ In other words, ‘Meaning before details.’ in an earlier blog on the Novice to Expert Continuum https://spaine.wordpress.com/2013/12/05/on-the-novice-to-expert-continuum/
It was noted that experts think in terms of patterns and context, where novices think more in terms of steps and rules. Becoming an expert does not happen overnight. However, two exercises that have shown to speed up the process are a combination of whole-part-whole study, and applying the principle of the 5 why’s. With the whole-part-whole (WPW), one forces themselves to mentally step outside their world by asking where in the whole does their part ‘fit?’ By building a model on paper of how their ‘part’ is connected to the next ‘part’ both upstream and downstream. Asking why my part is connected with that part, for what purpose, and how, all the way up to the ‘end’ is crucial in ensuring personal competency and meaning. I remember years ago when I was employed by a Telecommunications company tasked with troubleshooting digital transmission circuits. Whenever a communications circuit was experiencing a service impacting fault it would either generate or pass along a specific signal to indicate whether or not the fault was upstream, downstream, or originated within its own circuitry. These signals were designated with both colors and names. Red, Blue, Yellow. Loss of Signal, All one’s, and Remote alarm. It wasn’t until I drew out a circuit path (i.e. the big picture) and replicated a fault at each piece of terminal equipment (i.e. the part of the whole), all the while asking ‘why’, was I able to fully understand and take appropriate steps to correct the problem. In fact for a long time I had a drawing with me so I could refer to it when troubleshooting (I was at that time at the novice stage). I also remember the damage I could cause whenever I attempted to ‘fix’ the circuit before I understood the correct relationship between the alarm and the location of the fault. It took months and hundreds of trials before I felt comfortable enough to ‘see’ the whole circuit in my head and know what steps to take quickly. If you find yourself lacking in motivation, or confused as to why, or what you are doing makes any difference. A good strategy to start with is keep trying to draw out where your ‘part’ fits, how it fits, and why. do not stop the search to early, keep going until you can no long envision another connected part. Read the history of your ‘part’ to help you see the how, and why. You will know you have reached (or at least come extremely close) to this when you realize that whenever your efforts result in sub-optimizing the whole even though you optimize your ‘part’, you do harm. Whenever your work does not contribute to the goal of the system and you realize this, you are becoming more of an expert. When your efforts and the measuring of those efforts focus on local optimization without understanding how they will impact the ‘whole’, you can be sure that you do not understand the why behind your work. You can be sure that you are closer to the novice end of the continuum.

| Tagged | Leave a comment

The Expected Value of Luck’s Sample Size

Expected Value is defined as, The sum of all possible values for a random variable, each value multiplied by its probability of occurrence. Also called: mathematical expectation. If events are independent (if you don’t know assume that they are), as the number of trials goes up, so does the probability of success. The following comes from Amir D. Aczel’s informative little book: Chance: A Guide To Gambling, Love, The Stock Market, & Just About Everything Else. I will quote the relevant passage from Chapter 6.
“Suppose that my probability of getting a job is a mere half a percent, that is, 0.005. My claim is that  if I can just persevere and apply at a very large number of firms, I can bring my probability of getting at least one job offer to a virtual certainty as well. Lets try for, say, two thousand jobs. If I apply for two thousand jobs, my probability of getting at least one job offer is: 1-0.995(2000) = 0.999956, or 99.9956%. So even someone with a tiny probability of success a single trial can bring up my probability of success to a virtual certainty by trying enough times (and then, the probability is of getting one or more job offers!).

With independent trials, keep trying and eventually you will succeed!”

Ok  you say, but I am not going to take the time to submit 2000 job applications. Fair enough. I only used the above as an illustration of the principle for the need to persevere long enough to satisfy a sufficient sample to “tip” the odds in ones favor. If you want certainty in an unlikely field (i.e. job hunting) your sample of “attempts” will need to be quite large. If all you are looking for is an edge (i.e. +EV, or positive expected value) a very small sample will do the trick. Recall how this was discussed  in an earlier blog of becoming minimally informed .https://spaine.wordpress.com/2013/10/18/on-obtaining-a-minimally-informed-opinion-mio/  As human culture has become increasingly complex, the role of randomness (i.e. luck) has increased as well. Therefore, to properly exploit its effects in our decisions requires us to increase our “at bats”. To tip the odds in our favor. The best way to do this is simply increase our trials. So, to do this a best practice that has worked for me is to assume that whatever sample size I have selected is too small and increase it by 20%. Next, to preserver until the sample size has been reached. You will also need enough resources (time, money, personal sacrifices, etc.) to stay in the game (to stay in the game is simply another way of saying don‘t quit until you have reached your sample size). Again, the best way I know is to use a modified version of the Kelly Criterion. This principle of proportional “betting” is an optimal way of ensuring that you can run enough trials an never go bust. Finally, in any realm where chance, randomness plays a role, make sure you incentivize the risk takers by rewarding process over outcomes. Only reward outcomes if the realm is primarily skill based. See Michael Mauboussin’s article. http://blogs.hbr.org/2013/01/skill-luck-and-the-nhls-shorte/

“You’re not a failure if you don’t make it. You’re a success because you tried.”- Susan Jeffers

In this life our primary competition is not another person or system, it is our own fear.

Posted in musings | Tagged | Leave a comment

My personal journey along the novice to expert skills continuum

In an earlier blog on the novice to expert continuum I reviewed and elaborated on the skills continuum written by Hubert & Stuart Dreyfus in their book, Mind Over Machine: The Power of Human Intuition and Expertise in the Era of the Computer. I would like to add to their ideas by sharing my own journey along that continuum. However, the continuum that I will be referring to is a cognitive rather than a manipulative ( thinking rather than manual) skills. Additionally, I will only be making the claim that I have just “budged” off the novice category and still have a severely long way to go before I am in sight of the advanced beginner stage.
I calculate that this process began in earnest back in 1983, thirty years ago! One book started that journey entitled General Philosophy, by D. Elton Trueblood. The importance of that book was as a springboard to a great chain of Authors who subsequently and ever so gradually influenced all my future thinking. Below and not in any order of importance is a listing of my own personal cognitive journey along the skills continuum.

  1. Avoiding premature cognitive commitment by way of three mental habits:
  •   Recognition of the error of inerrancy 1
  • Recognition of the method of comparative difficulties 2
  • Force myself to include vigorous and serious attempts to think up at least three ‘tests’, or thought experiments to unearth anomalies with my favored opinion. To seek to falsify my favored opinion, all the while reminding myself that my mental ‘default’ is to find ‘evidence’ to confirm my views, and that it would be intellectually dishonest to only seek out these.

The error of inerrancy follows from the fact of our finite predicament, for even if there exists and infallible guide, how are we to know this? For infallibility necessarily includes not only an infallible revelation but the infallibility of the human mind to judge that revelation, which we know by every fact at our disposal we are incapable of. Regarding the method of comparative difficulties. To quote again from Trueblood. “The mood of philosophy is not that of the person who waits until he finds a solution utterly free from difficulties; it is rather the mood of one who knows very well that the best available is itself imperfect…acceptance of something which is less than perfect, because the alternative is worse.” Our minds seek a sense of certainty and assurance, if we short cut the method of intellectual ‘due process’ we too often gain the feeling of intellectual safety at the cost of intellectual honesty and more accurate opinion.

2. Continual reminder of the ever present risk of emotional hijacking 3

  • Related to D. Kahneman’s system 1 vs. system 2 intuition and thinking.
  • I have learned painfully that at the novice stage of cognitive development my thinking is influenced more often by the ‘pull’ of strong emotions, than by the ‘still small voice’ of reflective thought.
  • My best practice suggestion is to avoid like a virus inflammatory speaking and writing. If you can’t avoid it, force yourself to ‘cool down’ at least 30 minutes to allow system 2 to become engaged.

3. Recognition that my experience base will always be too small, yet only through great effort will I be able acknowledge it.

4. The importance of daily practice of remembering the biases I have been most susceptible to so they can be more easily and quickly recognized.

  •           Availability bias – whatever is most recent and salient tends to be more easily retrievable (works in both positive and negative directions).
  •           Priming – Wikipedia (works in both positive and negative directions).
  •           Framing – Wikipedia (works in both positive and negative directions).
  •           Confirmation bias – My default cognitive/emotional attitude is to seek to confirm my beliefs rather than to test their ability to avoid falsification.
  •           Law of small numbers – recognizing that my sample space (either measured in time, or observations) will always be too small.
  •           Contamination bias – Remembering that my default biases ‘do not smell.’ 4
  •           Representative or Base Rate Neglect bias –  Wikipedia

5.  Avoidance of speed in reaching conclusions – At the novice stage this is a recipe for increasing the likelihood of being wrong. Rather I have learned that thoroughness is the ‘better part of valor.’

6. The Unity of Truth 5 –  Something cannot be true in one discipline, yet false in another This implies that accurate and reproducible knowledge is interdependent. Religions are the biggest violators of this principle.

  •           I like to label this ‘ethical due process’. I have learned over time that I am to think, write, and say nothing that would hinder or impede future efforts by myself or others to advance knowledge claims in a given field even though my present views are wrong.
  •           I also like to remind myself of what it means to have legitimate regret.The difference between legitimate vs. illegitimate regret is that with legitimate regret, I only regret the outcome, whereas with illegitimate regret I regret both the outcome and the process used in reaching a conclusion. See Errors of Judgment vs. Errors of Technique
  •           See also Ben Goldacre’s book, Bad Science, chpt 4. On the principles of a fair clinical trial.

7. Spend three times the effort on improving your feedback tools and conducting post mortems after every opinion you feel is actionable. See also the principle behind deliberate practice, JustSevenThings web site.

_____________________

1 See D. E. Trueblood’s book, Philosophy of Religion, chap. 3
2 See D. E. Trueblood’s book, General Philosophy  chap. 4
3 Phrase comes from Daniel Goleman’s book, Emotional Intelligence, Bantam Books, 10th ed. 2005

4 See Timothy Wilson, David Centerbar, and Nancy Brekke, Mental Contamination and the Debiasing Problem, Heuristics and Biases: The Psychology of Intuitive Judgment, Cambridge University Press, 2002
5 Truth in Religion: The Plurality of Religions and the Unity of Truth, by Mortimer J. Adler, Touchstone Press, 1992

| Tagged | Leave a comment

Mental Habits

Mental habits that are helping me navigate through the last quarter of my life

I just turned 58 this year. As I appear to be living in the last quarter of my life I was thinking about the reoccurring thoughts that have become habits of both my thinking and sometimes actions. Here goes:

1. The base rate filter. When encountering anything numeric I find myself asking what’s the base rate? Ratios tell one more than just absolute numbers do.

2. Karl Popper’s dictum. Our ignorance is infinite while our knowledge must necessarily be finite. We always know less than we think we know.

3. My sample size (as measured in time, observations, or experiences) is always too small. Yet I tend to generalize (to my detriment) by referring to it.

4. Everything, and I mean everything in this life is temporary.

5. My emotional, intellectual, and behavioral “defaults” are often maladaptive to our modern age. Cultural evolutions speed coupled with my latent ability to adapt to it requires structural compensators, or “Nudges”.

6. I am better served if I seek first to become progressively less wrong, than by seeking to become more right. (see point 2. above)

7. I have found that one can leverage “luck” if I just increase my at bats.

8. The three two’s of decision making*
1.Types of knowledge required

  • On matters of general knowledge

          Use surrogates

  • On matters of specialized knowledge

          Use the consensus of experts

2. Time frames in which to make the decision

  • Long fuse
  • Short fuse

3.Decision consequence (risk potential)1

  • If it really matters
  • Chances of it being wrong x consequence of being wrong

9. A positive expected value (+EV) in evaluating my decisions is two fold

  • Pragmatic – did it work?
  • Viewing the world through the probabilistic lens –   Was it a good bet?

10. Once a +EV (i.e. best practice) decision model has been found, force yourself to value process over outcome.

*   More on these three categories in a future blog
1 My thanks to Douglas Hubbard. See his How to Measure Anything: Finding the Value of Intangibles in Business, Wiley, 2010

| Tagged | Leave a comment

On the Novice to Expert Continuum

On the Novice to Expert Continuum

Have you ever felt fear or anxiety when attempting a novel task, taking on a new job, incorporating a new skill? I guess we all have at some time in our lives. For those who see life as a series of ever expanding experiences it can be unsettling(at first). I recently re read an old favorite book of mine dealing among other things the process of developing expertise.1 Chapter 1 entitled ‘Five Steps from Novice to Expert’ is worth the price of the book. The Dreyfus’s lay out a compelling case for seeing the development of expertise as a process along a continuum. Their list  is as follows:

1. Stage 1 – Novice. Characterized by rule based decision making within a context free environment.

2. Stage 2 – Advanced Beginner. Still rule based, but less and less context free due to the accumulation of concrete experiences. Agent now relying more and more on prior experiences to help in making decisions.

3. Stage 3 – Competence. The gaining of the ability to prioritize among competing multiple sense data.

4.  Stage 4 – Proficiency. Experienced based intuition decision making. Automatically prioritizing in rapid fashion.

5. Stage 5 – Expertise. Pattern based with high level of personal responsibility attached to decisions.

I have used the above approach in to help me emotionally deal with every novel situation I have found myself facing since I read the book some 15 years ago. I have also found the works of Gary Klein, Sources of Power, Gerd Gigerenzer, Gut Feelings, extremely valuable in aiding my transition from novice to proficient (not sure if I ever qualified for expert). I want to add my comments based on personal experience of what has worked for me by modifying slightly the Dreyfus’s list.

Stage 1 – Novice – it is at this entry level of a new job, new skill acquisition, etc. that the use of checklists in invaluable why technicians also need check-lists, see also Atul Gawande wonderful book Checklist Manifesto. It is also the time that the organization must continually instill the concept of acceptable errors see Jerome Groopman’s article on the making of errors of judgment and errors of technique, Errors of Judgment vs. Errors of Technique. In fact organizational investment at this stage is crucial in aiding the employee in moving successfully to the advanced stages. Additionally, I have found that mentally running your own postmortems and asking yourself repeatedly what you would do differently next time in the same circumstance helps. Also, a trick that I learned in the military was to teach a problem solving technique to a colleague. Preparing for this unearths numerous uncertainties that help one to target your weak areas with additional research
Stage 2 – Advanced Beginner. At this stage I have found that what works best in advancing to higher skill levels is making sure you are “teaching” your emotions to overcome natural fear and anxiety so you can grab as many opportunities of novel experiences that you otherwise would let your more experienced colleagues tackle. In fact the only sure fire way to gain experience is to volunteer for as many and as difficult tasks as possible. A wonderful book by the late Susan Jeffers Feel the Fear and Do It Anyway, has always been a source of inspiration to conquering the fear and anxiety that so often prevents one from personal growth. It is also at this stage that an organization should encourage “fail often, fail fast” mindsets along with incentivising an experimental approach to skill level advancement. A “killer” is deincentivising a learning environment by rewarding only successes, while punishing failures.
Stage 3 & 4 – Competence & Proficiency – The rise of searchable and convenient databases has been for me a deciding factor in what I like to describe as signal to noise detection, or increasing the mean time to detect to mean time to repair ratio KPI Library. In fact the lack of convenient , simple to use, relevant, and searchable databases is in my view one of the main stumbling blocks in organizations to help their employees advance in experience and skill level while minimizing ET errors along the way.

1 Hubert L. & Stuart E. Dreyfus, Mind  Over Machine: The Power of Human Intuition and Expertise in the Era of the Computer, The Free Press, 1986

| Tagged | 2 Comments