Subject Matter Experts(SME) Programs : What are the Boundary Conditions?
The hidden things to note while creating Expert Systems
Applying Pareto principle in building organisation expertise translates to “20% of the folks know 80% of the stuff”. The leadership aspires to clone this 20% expertise to the remaining 80%. In this process, the first logical step is to make the experts to create a comprehensive curriculum and then assign it to individual (the remaining 80% crowd) to learn and become an expert. This is a standard industry approach most of the organisation follows. The topic of this blog is to analyse how effective this sort of programs are and what are the caveats we need to be aware while executing these SME program.
Taking Product Support Engineering as an example, we always wants to codify our knowledge into list of “if-then-else” document. These are set of procedures to follow given the well defined problem happens. These procedures work in a well ordered situations where we don’t need to worry about the changing conditions.
These are essential primitive steps to create a knowledge/expert database within the team and it certainly helps in moving forward. The problem lies with the way we create these programs and assessing the impact of the same.
When you start learning a particular technology or a thing, you start with ‘explicit knowledge’ that is available in the books and as procedures. This initial exercise gets you comfortable with the essential grammar and vocabulary of the world and gives fair confidence to face the problem. As you start working towards the more and more problems in that space, You develop experience. This experience includes,
Intuition.
Ability to notice typicality or anomaly.
Build own mental models.
Judgements during uncertain situations.
This above experience is called “Tacit Knowledge”. This tacit knowledge is the most difficult part to extract from the expert. Experts cannot explain what they see, which seems obvious to them. Chip Heath and Dan Heath calls it “The curse of knowledge”. Experts find it hard to imagine not knowing it. The acquired knowledge has “cursed” them. They have difficulty sharing it with others, because they cannot readily re-create the novice’s state of mind.
Tacit knowledge are not something you learn from textbooks. A person know more than what he/she can tell. This is an essential part of our ability to handle the complex situations which the SMEs are known to have that skills.
In a nutshell, SMEs are born out of this equation.
SMEs = STRONG [Explicit Knowledge from Text Books, Procedures, Rules + Tacit Knowledge from Judging Typicality, Perceptual Skills, Mental Models, Pattern Matching]
In complex environments, the “Tacit knowledge” is much bigger than the explicit knowledge. This comes by getting soaked more and more into the area of interest.
Gary Klein says Tacit knowledge is very crucial for the way we design and use procedures in complex situations. Interestingly, we have an area of study that deals with this problem. Its called Cognitive Task Analysis(CTA). Its is an area of psychology which deals with eliciting knowledge from the experts.
There are more than 100+ types of CTA methods which can be used to describe the “Tacit knowledge”. At a high level, CTA breaks down into 3 broad categories,
Perform close Observation and interviews of experts
Work closely with Experts and Talk to them.
Process tracing
Perform “Think-Aloud” Protocol or Subsequent Recall methods to gain insights of a specific problem.
Conceptual techniques
Well structured mind map of the relevant concepts of a particular area.
CTA techniques are gaining more importance as evidence suggests experts are not fully aware of about 70% of their own decisions and mental analysis of tasks.
I would love to hear from the readers if any other techniques has been quite effective in creating expert systems.