Dunning-Kruger: What is to be Learned?

What is to be Learned?

As we travel along the learning arc described by the Dunning-Kruger graph, some observations:

• At all levels there is a likelihood—a certainty even—of the underestimation of knowledge-to-be-gained and an associated overestimation of knowledge already acquired or inherited from other domains.

• The knowledge-to-be-gained consists of:

(a) Domain knowledge (questions raised and answered to acquire knowledge)

(b) Knowledge-acquisition knowledge (process that exposes questions and sources answers), and

(c) Meta-knowledge (knowledge-about-knowledge, often as context for the questions and answers or other related domain questions and answers)

• The Dunning-Kruger Effect tells us that, in general and as we saw in the How do we Know what We Know? section, it is clear we should always be wary of assessing our own competence in new areas of knowledge. While we might be sure that we are not subject to this confidence / competence disparity, we would be wise to remember that all those other folks who we see clearly are subject to it, also don’t think they are.

• Having a good, context-specific, question is often more important than having the answer since it acknowledges the limitations of our knowledge and helps illuminate the context which may have more and more far-reaching knowledge.

• Being able to only ask context-free questions (eg., What requirements does this system have?) is a strong indicator of a certain level of ignorance. It may also indicate a need to look at the knowledge acquisition process since the answers to vague context-free questions are often themselves vague and may contain little usable knowledge.

• There is a difference between knowledge (what we know) and the knowledge of how to acquire knowledge (how we learn). Of the two, the knowledge and application of learning is probably most important.