All models are wrong. So what?

All models are wrong. So what?

Innovation Method
Published On:
May 22, 2023

“All models are wrong, but some are useful,” noted the British statistician George Box. I came across this aphorism in a TIPS training workshop with a business group from Switzerland. When one participant commented that a detail in her profile description was wrong, another member responded by citing the quote by Box. Today, let’s talk about the purpose of models, the process of modeling in general, and the modeling of personality and cognition in particular, and what’s wrong and useful about it.

All models are wrong ...

Let’s first talk about the process of scientific modeling. Scientific modeling aims to make a particular aspect of the world easier to understand by linking it to existing, commonly accepted knowledge.

A model is a task-specific, purposeful simplification and abstraction of our perception of a particular aspect of reality. Thereby, the creator of a model simplifies to focus only on those relevant aspects for the specific task (or question) at hand. At the same time, the modeler also abstracts by aggregating information that is relevant but not needed in all intricate details of the actual phenomenon. As such, every model is not a true representation of reality but a simplification and abstraction. It follows that by definition, all models are indeed wrong.

... but some models are useful

So if all models are wrong, as shown above, why do scientists and creators engage in the process of modeling? Because, as Box pointed out correctly, while all models are simplified reflections of reality, some of these approximations can be useful. They allow model builders to debate which is a better model to approximate reality for a given task and visualize, experiment, and intuitively understand the represented phenomenon. As part of this model-building process, more and increasingly better, more useful models emerge over time.

How I arrived at a new “wrong” model for profiling people...

Like most other personal assessment tools, TIPS is a model to capture, describe, profile, and visualize the complex modalities of a person’s cognition and personality that I purposefully created with one question (or focus task) in mind: How can we make everyone contribute to creativity and innovation in harmony with their preferred cognitive styles? (Later on, while refining the model, a second focus question emerged: How can we make everyone perform in connection with their natural talent?)

“Experience without theory is blind, but theory without experience is mere intellectual play,” noted the German philosopher Immanuel Kant. For more than half a decade, I experimented with other existing models to profile personality or cognitive styles to find answers for my focus question. In my fieldwork with companies and graduate students, I noticed that certain theoretical constructs of existing personal and cognitive assessment tools show relevance to my focus question. Other constructs were irrelevant. I also figured that some features of different models were either incomplete or even plain wrong. For example, MBTI (Myers-Briggs Type Indicator) classifies people as either an extrovert or introvert, but I know that I am both at the same time. Finally, I also spotted that certain constructs were missing to explain the dynamic, cyclical nature of business and innovation and what role people play in it.

Hence, I began looking into other domains beyond the psychoanalytic and cognitive realms to find theoretical constructs that allowed me to explain and capture phenomena that I observed in my practical work as a business creativity and innovation facilitator. I came across some interesting concepts from the fields of evolutionary economics and social sciences explaining how technological, economic, and social changes unfold over time, which led me to a new construct of social base orientations.

I incubated on how the different puzzle pieces (= theoretical constructs of the different areas together) fit together for some time. Then one day, the outlines of the TIPS map suddenly appeared in front of my mind, and I immediately knew that this was the right “wrong” model.

All that was left for me to do was to formulate the details of the relevant constructs that conjoined to form the TIPS model. Next, I began to operationalize, test, refine, and optimize the model over various rounds of iteration (using rapid prototyping principles). Being an innovation company, we embrace the iterative nature of model design to steadily improve both the details of the model and the underlying algorithm that we use to compute the profiling results. In the future, we intend to also draw upon artificial intelligence to make the TIPS model less and less “wrong” over time.

... and why this new “wrong” model is useful

“All theory, dear friend, is gray, but the golden tree of life springs ever green,” noted the German playwright Johann Wolfgang von Goethe. TIPS is as wrong as all other models to profile personality and cognitive style. Still, we believe that TIPS is really useful for companies to increase their business and innovation performance. Because of being purpose-designed with an eye on how people can contribute to innovation in particular and business in general, TIPS has lots of concrete business and innovation applications. For example, TIPS gives answers to questions such as:

While other models to profile personality and cognition address some of these points, too, none of them can give you comprehensive answers to all of these questions. So, because of its very wide range of more than 35 applications in business and innovation (and we’re still counting, as we happen to discover new applications in our work with organizations asking us further questions that we haven’t considered yet), TIPS can be tremendously valuable for organizations who want to increase their performance in business and innovation in the era of the innovation economy. (And I use it all the time to run and plan the future expansion of our activities at Thinkergy).

Conclusion: Models are the next best thing to the real thing

George Box was right indeed: “All models are wrong, but some are useful.” The TIPS model is one of the latter because of its extensive applicability in business and innovation.

At the same time, we never forget that TIPS is just a wrong model waiting to be further refined and honed over time. Hence, we welcome feedback and actively ask users to tell us what’s wrong with it. We know that negative feedback is a critical evolutionary principle and an essential aspect of rapid prototyping that we embrace to arrive at increasingly better products quickly. We commit to continue making the TIPS model “less wrong” (and the related profiling concept and algorithm more robust) with each user-data-based iteration that we’re going to release in the years to come.

  • I discuss the conceptual make-up and design architecture of personal and cognitive profiling tools and what’s wrong with many of the models in two sections of my new innovation book titled “The Executive’s Guide to Innovation.”
  • Have you become curious about TIPS, Thinkergy’s talent and innovator profiling system? Check out our TIPS website or download our TIPS booklet to learn more.
  • What type of innovator are you? What’s your talent all about? Get TIPS-ed now for just USD 89 and find out.
  • Care to profile your entire team and train them in a TIPS training workshop? Contact us and tell us more about you and your team.

© Dr. Detlef Reis 2021.