Data Literacy and the Japanese Way of Tea

Data literacy is a pre-requisite for a data-informed culture. However, in most organizations, there is a huge variation in employees’ knowledge, abilities, attitudes, perspectives and career goals. 

The composite nature of data literacy adds another layer to the challenge. Data literacy, in the simplest terms, is the ability to read, work with, analyze and argue with data (Source). It is not one skill, but a set of skills including: 

  • Statistical understanding
  • Data preparation
  • Data analysis
  • Data management (including Data Governance)
  • Effective communication

So, any data literacy training for an organization must take into account:

  • Variation among learners
  • Multiplicity of skills that comprise data literacy
  • Evolving business needs 
  • Advances in data science

You don’t need a personality quiz

An increasingly popular approach to designing data literacy training is to identify “data personalities” using a quiz. It usually looks like this:

Personality/RoleCurriculum Difficulty
Data WarriorLevel 1
Data WizardLevel 2
Whimsical Title 3Level 3
Whimsical Title 4Level 4

This may seem like an improvement over the one-size fits all, but what we really have is one of these 4 arbitrary sizes fits all.

E.g. where does a learner who is excellent at data extraction but lacks a statistical foundation fit? Where does a statistician who knows little about organization’s data management/access practices fit? 

As such, they still don’t address the nuances of adult learning in work place for a fast evolving field. This is why training models cannot be “personality” or role based. They must be skill-based. 

Shu Ha Ri

Shu Ha Ri

It is one of the oldest concepts addressing skill development. It was first introduced by a Samurai named Fuhaku, as a training hierarchy to Japanese tea culture. 

  1. Shu means “to obey” or “tradition”. The learner obeys all that the teacher imparts. The goal is to develop a strong technical foundation.
  2. Ha means “to break free”. The learner begins to question the rules/guidelines while applying them in new ways. He develops his own style
  3. Ri means “to depart”. The learner is now a practitioner for whom the skill is second nature. She has transcended the rules/guidelines and can now extend the discipline, without a teacher.

Apparently, Shu Ha Ri was originally designed with a rather malicious intent to limit the knowledge and culture of tea to the elite. As Peter Merel notes:

“Fuhaku’s idea was simply that Shu training recipients should display obedience to Ha teachers, and pay for the privilege. The teachers would purchase licenses from Ri grand masters (Iemoto) who derived authority by blood or zen from Rikyu. Money flowed up, training passed down, and peer to peer learning communities were swiftly out-marketed.” (Source)

The concept was later adapted to performing arts and martial arts in the 17th century.

Even though it originated as a hierarchy, it is now depicted as concentric circles to symbolize that “there is Shu within Ha and both Shu and Ha within Ri. Thus, the fundamentals remain constant; only the application of them and the subtleties of their execution change as the student progresses and his, or her, own personality begins to flavour the techniques performed.”  (Source)

For me, Shu Ha Ri is an elegant concept to start thinking about how we learn and what it means to become an expert. But it is still too abstract for me to apply directly. So, let’s take a look at another model of gaining expertise.

Dreyfus Model of Skill Acquisition

This was first proposed by Stuart Dreyfus, a mathematician and Hubert Dreyfus, a philosopher in 1980.  They described a five-stage process to go from Novice to Expert. 

Skill LevelContextPerspectiveDecision based onEmotional InvestmentCritical thinking skill
NoviceNoneAll the detailsRecipeNoComprehension
Advanced
Beginner
YesAll the detailsGuidelinesNoProblem solving
CompetentYesSome detailsConceptual ModelsYes, in the outcomeFocus
ProficientYesBig pictureMaximsYes, in the understandingFeedback
ExpertYesBig picture
Pattern matching
IntuitionYesMetacognition

The Five Stages

Novice

Novices have no previous experience with the skill. Their training must begin with foundational facts, without any situational context. They must be given simple and clear rules (“If this, then that” recipes) to guide tasks. In the face of an unexpected situation, they will not be able to troubleshoot. If the rules do not work, they can place the blame on the rules and call it a day. As such they are not emotionally invested in the task.

E.g. A novice learning how to write DAX in PowerBI would use Quick Measures. Goal is to understand the syntax and types of formulae available.

Advanced Beginners

Advanced Beginners start to rely less on the recipe. They aren’t yet able to appreciate the big picture and still have difficulty troubleshooting. As they gain experience and develop context, the specific rules evolve in to general guidelines. If the novice rules and general guidelines don’t work, they may take some initiative but they are still not emotionally invested. 

E.g. An advanced beginner would learn how to write the DAX formula manually. He may use the the quick measure as a starting point and then manipulate it to see how the changes impact the result. The goal is write more complex formulae.

Competents

Competents have gained enough experience to feel overwhelmed by all the details they must keep in mind. Since they are more context driven and no longer have a recipe to rely on, the competent is more emotionally involved in the task at hand. Whether they succeed or fail matters personally. Taking responsibility, without a recipe to fall back on, is crucial for advancement because it leads to better troubleshooting. Their training must help them develop conceptual models

E.g. A competent may feel overwhelmed by all the possible formulas in his toolbox. The goal now is to develop a strong understanding of the syntax and know which formulas to use when.

Proficients

As a result of emotional involvement and reinforcement provided by successes, the proficients are able to recognize and focus on relevant details. They appreciate the big picture and can recognize what does not fit. They are able to self reflect, self correct as well as learn from others’ experiences. However, there is still a lag between seeing what needs to be done and deciding how to do it. To shrink the lag, their training must include case studies and maxims.

E.g. A proficient would benefit from reviewing the work of others and sharing her own work to solicit feedback.

Experts

Experts do not rely on any rules, guidelines or maxims. They use intuition. There is no lag between seeing what needs to be done and how to do it. They seek out opportunities to improve. Their training must expose them to novel experiences and new developments in the industry.

E.g. Experts must stay on top of the new and upcoming features to identify opportunities to improve existing reports and measures in the organization.


The Dreyfus Model and Shu Ha Ri, both posit that :

  • Skill in its minimal form is produced by first learning (obeying) the foundational rules
  • To achieve higher levels and break from the rules the learners must practice and gain experience
  • Expert is someone for whom the skill is second nature, or intuitive

The concentric circle depiction can be extended to the Dreyfus model. It implies that experts should continually deepen their understanding of the basics. For them what was once simple, becomes more nuanced and insightful every time they revisit it. Also note, that the expert level does not have a boundary because learning is never ending. Experts must seek and embrace their own ignorance.

Dreyfus Model for Data Literacy

Adaptation of the Dreyfus Model of Skill Acquisition

A Framework for Data Literacy Training

The concrete descriptions of the Dreyfus levels provide a more adaptable framework to build training on. It allows learner-level customization and acts as a tool to asses gaps in ones’ knowledge. It also highlights the multi-faceted nature of data literacy in an evolving workplace and workforce.

Here is a template:

Skill LevelStatistical
Understanding
Data
Preparation
Data
Analysis
Data
Management
Communication
Novice




Advanced
Beginner





Competent




Proficient




Expert




Agile and Shu Ha Ri – A Lesson

Alistair Cockburn drew the link between Shu Ha Ri and Agile software development. The approach has since garnered some critique because it led to organizations rolling out one-size fits all Agile boot camps. However, it is easy to see that the failure lies in adoption of the concept and not the concept itself. 

After introducing their employees to the basics of Agile and then imitating the agile practices from others, organizations often think the adoption is complete. When in reality they are simply stuck in the Shu-phase.