How Rapid Prototyping in Schools can Fail

- Posted in Uncategorized by

How does your school solve problems, make changes, or figure out what works best? In my previous post I wrote about how important it is for schools to get used to the idea of conducting controlled experiments to generate new knowledge for how make decisions and solve problems. In this post, I am going to be talking about a design model that I’ve seen used in schools to facilitate the testing and revision of new ideas: The Rapid Prototyping model.

Screen Shot 2020-04-10 at 5.09.42 AM

As the name suggests, rapid prototyping is a procedure consisting of the development of incomplete or draft versions of instructional initiatives, aka prototypes, and testing them in rapid succession to verify their effectiveness (Nixon & Lee, 2001). The theory is that by producing multiple iterations rather than one, and by emphasizing the process of revision, formative evaluation and testing early in the design process, an effective final product will likely emerge. While rapid prototyping is more of a problem-solving model than a scientific model, if the controlled component of experimental design is maintained during prototyping by creating two conditions, an experimental and a business-as-usual group, and only one variable is manipulated at a time, the Rapid Prototyping model can be a powerful grassroots tool for innovative schools.

Rapid Prototyping vs. Traditional Change Management

What have teachers traditionally done when they want to solve a problem? In my experience, teachers figure out ways to solve small problems by themselves, but for big problems they will usually go to their supervising principal. Since this can be an intimidating experience, and because principals are usually stretched thin, a lot of big problems never make it to principals’ desks. When problems are brought to a principal’s attention, either a top-down solution is formulated through introspection, or meetings are called to give the impression of collaboration and a solution is chosen by consensus.

I find it unlikely that the introspective, meeting-based approach that most schools use will lead to the best solutions. When principals and teachers sit around tables and think, think, think, the solutions that are generated are, at best, partially informed and inherently governed by instinct, bias, and belief. Instead of discussing what we believe will work, deploying it, and then having no way of measuring its success, we might be wise to design and deploy a variety of mini-experiments and test their effectiveness based on measurable criteria, keeping the initiatives that rise to the top. Essentially, I am describing the Rapid Prototyping model. When done well, I believe it can have a powerful impact on the culture of an organization, and is a more reliable method of generating effective and efficient learning solutions than the status quo.

Ideally, the rapid prototyping process should look something like this:

  1. Prototype
    • Teachers identify a problem through active “problem-seeking”
    • A team of experts is formed and mock-up/prototypes are designed
  2. Review
    • The prototypes are reviewed and then deployed for testing
    • Data is collected and the successes and failures of the prototypes are evaluated (compared to a control group)
  3. Refine and Iterate
    • The prototypes are revised (preferably one element at a time) based on the feedback and redeployed for further testing.

To me, this is an exciting alternative to how schools are typically run. Rather than passively receiving top-down initiatives from ill-informed leaders only to watch the solutions decay and disappear with the leaders who championed them, teachers can be empowered to craft their own context-specific solutions with the guarantee that they will be rigorously tested and refined at all stages of development. 

When Rapid Prototyping Fails

Like all instructional design models, there is always a way to totally screw up the rapid prototyping process to the point that it becomes nothing more than a formality and leads to flawed solutions. For the rest of this post, I’d like to describe some of the common pitfalls that I’ve seen firsthand with a rapid prototyping approach, and what we can do as intentional educators to preserve its integrity.

Lack of knowledge or expertise

While the Rapid Prototyping model provides the iterative cycle that describes the generic process for creating and refining school initiatives, it doesn’t include the context-specific recommendations for how to do each of the steps. Rapid prototyping can fail to generate effective solutions when members of the design team do not possess the a) expertise for what a good initial prototype for the project might look like, and b) the specific knowledge for how to refine the prototypes to make them better. For example, let’s say that a school is trying to adopt a model for teaching math. A successful design team would readily have multiple options already in mind, and a bit of research would produce a few more. Once a couple of selected prototypes are deployed for initial testing, a successful design team would assign experts in math instruction to review the success of the prototype against the objectives, and their expert knowledge would enable them to recognize multiple approaches to refining the prototypes in ways that novices simply cannot. Since one cannot think well about nothing (no, one can’t just Google it), a design team without deep wells of organized knowledge about the problem is doomed to failure from the start.

Lack of research skills or opposition to research

While a design team of teachers without knowledge of the problem at hand is seriously disadvantaged when using the Rapid Prototyping model compared to a more knowledgable design team, a design team can develop their expertise through research on the problem. I’m talking about using a research database and reading peer-reviewed articles in scholarly journals. I described my evolution towards becoming a more research-informed educator in a recent podcast I was featured in, The Research Question. Unfortunately, I’ve been met by opposition to this approach in the past. The main arguments against using research are that a) teachers cannot possibly learn to be research literate (absurd!), b) that research is constantly changing so we’ll never know what’s true and what’s not (short answer, that’s science, but there are, of course, things that have an overwhelming evidence base to the point that they are unlikely to be challenged), and that c) experience and intuition, aka craft knowledge, trumps what those elites say in academia (maybe, except that so much of the knowledge that science produces is counterintuitive and flies in the face of our experiences). For a rapid prototyping approach to work in school settings, those doing the prototyping must draw from the research literature, including choosing between others’ ready-made solutions, to expand their understanding of the problem they’re meant to tackle.

Taking a non-scientific approach

Deploying a prototype can look different depending on the nature of the problem and the context, but the part that tends to be forgotten is the controlled part. Whenever we test something in our schools, we should do our best to organize a business-as-usual group, and we should change only one variable at a time. To continue the example of prototyping in order to adopt or construct a math instructional model, one class section could be given XtraMath to see if it would be a efficient way of improving math fact fluency, while another class section continues learning math facts with the usual method for a pre-determined amount of time. If we roll out XtraMath for all the grades all at once and keep changing the conditions for those involved in the experiment, how can we expect to learn anything from the results?

Risk-adverse or time-constrained faculty

Rapid Prototyping can fail when teachers feel uncomfortable taking risks. Without building a culture that places value on scientific experimentation, which can produce failed experiments or inconclusive results, teachers may prefer instead to go through a more traditional top-down problem-solving process. Equally as debilitating to the rapid prototyping approach is when time for prototyping simply doesn’t exist. Teachers should feel excited to contribute creative solutions, so prototyping shouldn’t be added on to our workloads without taking something else away.

Lack of buy-in from leadership

Leaders that decide to adopt a rapid prototyping approach to school improvement have to walk the walk. When a deployed prototype is well-intentioned, but ultimately fails, it has to be understood by the administrator that this is a normal part of the rapid prototyping process. Administrators can ensure that rapid prototyping succeeds by carving out time in teachers’ timetables to allow them to do this sort of work – No adding without subtracting! –  and use professional development sessions to improve the prototyping methodology. Finally, administrators should model the prototyping process by carrying out their own experiments. If teachers do not have a model for what effective prototyping looks like, we shouldn’t expect that they will be successful at replicating the process themselves.

Rapid prototyping is a process for school improvement in which teachers propose instructional initiatives and test them. Even the most well-thought-out experiments can fail, which is both exciting and scary. Still, when teachers leverage their collective knowledge and expertise, integrating research and theory in the process, and receive the support they need from administrators, I think the Rapid Prototyping model can be a highly effective way of designing both instructional and non-instructional solutions for a school.

How does your school solve problems and generate knowledge? Does it look anything like this?

– Zach Groshell @mrzachg


Nixon, E. K., & Lee, D. (2001). Rapid Prototyping in the Instructional Design Process Classic Instructional. Performance Improvement Quarterly, 14(3), 95–116.


Help! I’m Trying to Teach My 9-Month-Old How to Crawl and it isn’t Working

- Posted in Uncategorized by

I have a 9-month-old daughter who still cannot crawl. I’ve tried having her build up her strength through various leg and abdominal exercises. I’ve shown her interactive diagrams and YouTube videos of babies crawling, and I’ve read her the definition of crawling from the dictionary. I’ve modeled the correct way to crawl so many times I’m wearing holes in the knees of my jeans. The doctor told me that this is totally normal and not to worry, but today when I put her down in the crawl position, she was totally hopeless. Please help!

– Zach Groshell

Before this blog gets flooded with comments either a) kindly informing me that she is well within the normal age range to not have crawled or b) making fun of my parenting, let me clear that I am being facetious. I, like most people, understand that no amount of instruction will teach my baby how to crawl; It will come naturally, when she’s ready. It would be just as silly to try to teach her how to coordinate her lips, tongue, alveolar ridge, and breath in order to try to get her to speak (Paas & Sweller, 2012). As a typically developing child, she will learn to speak at about the same time as all other typically developing children – naturally, effortlessly, through simple membership in English–speaking society.

Why is this the case? The basic answer is that we’ve evolved to be able to crawl, walk, and speak, and acquire a number of other abilities naturally, without instruction. In this post, I would like to share an especially useful theory of knowledge categorization that was featured in one of my recent workshops, but that seems mostly unknown to teachers; Biologically primary and biologically secondary knowledge (Geary, 2008). While it is not a “finished” theory, it can help teachers and leaders in education to distinguish between that which is best learned unconsciously and naturally, and that which requires formal instruction.

Biologically Primary Knowledge

Imagine the most remote, isolated tribe of peoples, inaccessible by plane, train, or automobile, cut off from modern-day amenities like TikTok and Amazon Prime. We can be certain that, despite there not being any formal system of schooling or access to instructional texts in this society, every member of this tribe will learn to walk and speak their native language. We can also expect that children in this society will quickly learn to discriminate between faces and organize themselves into cooperative groups (Geary, 2008). We have evolved to be able to rapidly and effortlessly acquire this kind of knowledge, biologically primary knowledge, because without it the chances of survival, stability, and reproduction would be greatly reduced. As teachers, we can expect that a policy of “just let kids be kids” will be sufficient to teach biologically primary knowledge for all typically developing children. In effect, the classification of knowledge into the two categories of biologically primary and biologically secondary knowledge provides the ultimate justification for play-based learning in the early years of schooling, because by depriving children of opportunities to socialize with each other and play with tools, we may be stunting their acquisition of non-instructable foundational knowledge.

Humans across all tribes and all societies, regardless of their level of modernization or economic development, have evolved to acquire, and will acquire, biologically primary knowledge. What, then, is biologically secondary knowledge?

Biologically Secondary Knowledge

While it would be pointless, and frankly, bizarre to attempt to teach biologically primary knowledge to children through formal instructional methods, such as through lecture or by providing worked examples, a great deal of knowledge that we require to function in developed society must be taught (Tindall-Ford, Agostinho, & Sweller, 2019). This sort of knowledge is known as biologically secondary knowledge, and examples include practically any academic skill, such the knowledge of how to read, write, and solve math equations. Biologically secondary knowledge, while essential for getting a job or comprehending a newspaper in today’s world, constitutes knowledge that was only relatively recently invented in human history and isn’t guaranteed to be present across all cultures and societies; Some societies have never developed a system of writing for their language and many cognitively able adults are illiterate. Unlike biologically primary knowledge, biologically secondary knowledge isn’t strictly necessary for our survival, so we have not evolved to acquire it effortlessly through immersion or osmosis.

Acquiring a strong grasp of the correspondence between letters and their sounds in an alphabetic writing system (i.e. phonics) is a prime example of something that we have not evolved to learn, and therefore must be taught. We know that simply placing texts around a child’s environment and modeling the attitudes of a reader is not enough to teach a child to decode words; We must pass on the knowledge of letter-sound correspondences of a written language to our students through instruction. We cannot expect that the same strategies that work for learning to speak a first language fluently (i.e. play, discovery, and social interaction) will work to teach students to read, write or solve math problems fluently.

The role of schools is, therefore, to teach biologically secondary knowledge so that the next generation can inherit that which we have not evolved to learn naturally. Instructional design is effectively a field concerned with the design of “unnatural” instructional procedures for the teaching of biologically secondary knowledge that comply with what we know about the capabilities and limitations of the human cognitive system (Paas & Sweller, 2012).

As teachers, we should be content with our role in teaching domain-specific biologically secondary knowledge, and reject the arguments from snake oil-peddling education “gurus” that academic learning should be natural and “when they’re ready.” If we haven’t evolved to learn an element of the curriculum naturally and without effort, then the strategy of letting “kids be kids” so that they might magically pick it up through osmosis will not work. It would be just as awkward, bizarre, and pointless to attempt to teach a student advanced algebra through free play and unguided discovery as it would be to try to teach my daughter to crawl through instruction. 

– Zach Groshell, tweeting @mrzachg


Geary, D. C. (2008). An evolutionarily informed education science. Educational Psychologist, 43(4), 179–195.

Paas, F., & Sweller, J. (2012). An Evolutionary Upgrade of Cognitive Load Theory: Using the Human Motor System and Collaboration to Support the Learning of Complex Cognitive Tasks. Educational Psychology Review

Tindall-Ford, S., Agostinho, S., & Sweller, J. (Eds.). (2019). Advances in cognitive load theory : Rethinking teaching. Retrieved from