Category: Innovation
The Marshmallow Maneuver

The Marshmallow Maneuver

By | | Comments 0 Comment

Marshmallows are exquisite probes of the human psyche.

So, here’s a question: what relationship do marshmallows, tape, string, scientific discovery, and uncooked spaghetti all have in common?  (And in case you’re wondering, this week’s feature photo is a bundle of uncooked spaghetti photographed from above.)

The answer to the question above comes from answering another question as posed in the book by journalist Warren Berger A More Beautiful Question: The Power of Inquiry to Spark Breakthrough Ideas.  I came across Berger’s book while doing preliminary research for formulating my discovery cycle (which I’ll log entries on for each phase of the cycle I’m using in early 2019).  It was a welcome reference entry for the first phase of the cycle, asking questions.  Questions are what ignite the process of scientific discovery because they express and focus the desire to know more about something, inspiring us to act.  It turns out that asking questions about how to ask questions was trickier to find information on that I thought it would be.  It’s not something we spend much conscious time or effort on.  We worry more about answering questions well rather than asking questions well.

So back to Berger’s provocative question, which was the following: “How do you build a tower that doesn’t collapse (even after you put the marshmallow on top)?” (p.120)

It turns out that this is what has been asked of a number of groups in various studies and posed as an exercise in design innovation workshops the world over.  In the usual form, participants are asked to build the tallest free-standing structure they can, in an allotted time, using just pasta, tape, and string and with one marshmallow placed on top.  Interestingly enough, among various groups of participants, two stand out in comparison: kindergarteners outperform graduate MBA students on this task.  Part of the reason lies in psychology.

There is a long tradition of marshmallow tests, kindergarteners, and psychology.  The most famous example in popular culture is a study that used marshmallows (among other sweet treats) to investigate willpower in kindergartners and its correlations with later life outcomes.  In that study, kids were given the option to get one marshmallow now or wait for a bit and, in return, get two later.  It appeared that children’s choices between instant gratification (give me one now) and delayed gratification (I’ll wait for two later) were linked to outcomes in adolescence.  Though the jury is still out on exactly how and with what outcomes this test correlates.

I had heard of this story (it’s often in the news), so when I came across marshmallows and kindergartners in Berger’s book, I assumed I already knew the punchline: if you are patient with a question and mull it over it will lead to more positive outcomes.  It turns out I was dead wrong.  When it comes to asking questions, patience is your friend.  But when it comes to answering questions, instant gratification seems to be the way to go.

Here’s what the marshmallow tower studies have found: groups that engage in many trials throughout the allotted time, building, failing, and trying again, on average end up with taller structures.  Kindergarteners jump right in to this approach, preferring a hands-on tactic and prototyping early and often to try and succeed.  In contrast, other groups, like MBA students, spend the majority of their allotted time discussing how they should approach and try to solve the problem.  This results in fewer actual attempts and on average shorter structures (or no successful structures at all!) as a result.

It seems then that Berger’s book not only discusses how and what kind of questions spark breakthroughs (which I’ll cover in a later log entry), but also how best to start trying to answer those questions: trial and error.  If you’ve read many of my log entries on the site, you’ll know favoring trial and error and failure is fast becoming a recurrent theme.  But it’s always good to have reminders.  This is part of the intent of the ARTEMIS virtual reality software being built: to give you a way to build mental models of what you are trying to discover fast and often.  And if you read much in the startup (like Eric Reis’ Lean Start Up), software (like Jeff Sutherland’s Scrum: The Art of Doing Twice the Work in Half the Time), or entrepreneurial arenas (like Jake Knapp’s Sprint: How to Solve Big Problems and Test New Ideas in Just 5 Days) then you will know that rapid prototyping to test out answers and learn by getting immediate feedback is all the rage right now.

So, trying out the marshmallow maneuver, with office supplies and uncooked food to build my own tower, may be the way to remind myself of the value of fearlessly trying out answers to big, weighty, scientific discovery questions.  A great scientific discoverer, Thomas Edison, inventor of the light bulb, once said in an interview with Harper’s Monthly Magazine (1890):

“I speak without exaggeration when I say that I have constructed three thousand different theories in connection with the electric light, each one of them reasonable and apparently to be true.  Yet only in two cases did my experiments prove the truth of my theory.”

(Thomas Edison, Harper’s Monthly Magazine, 1890)

He’s talking about theories, not experiments.  Three-thousand-theories.  As a theoretical particle physicist that really resonates by “quantifying” the “degree of try” it might take to even think up a good answer to a good question.  Besides, maybe if there’s a marshmallow at the end of every attempt, I’ll get better at generating my own 3,000 theories to find the 2 that work.  And if I’m smart, I’ll go after that marshmallow today and not wait until tomorrow.
At Discovery’s Edge

At Discovery’s Edge

By | | Comments 0 Comment

The balancing act between theory and practice, qualitative insight and quantitative assessment, is a tough one.  In my quest to develop a repertoire of skills and practices targeted at scientific discovery, theory and qualitative insight have dominated the body of literature I’ve read so far.  Until I came across a magnificent pair of papers published by a group of sociologists and a theoretical biologist.  Their goal was to analyze a tension often discussed among scientists: stick with tradition or pursue innovation?

In these recent papers, the authors devise a living map of “what is known”, represented as a series of nodes and links between them, on a network graph.  They use biochemistry as their scientific use case; nodes represent molecules and links between nodes represent published connections between molecules.  They do this using a massive network mapping of molecules and connections appearing in abstracts of published articles in journals—around 6.5 million abstracts.  Ah, the glorious face of big data.

So, in this little microcosm of knowledge about discoveries in biochemistry, what can we learn about community-wide research strategies?

The first thing we learn is that there are techniques to map “what is known” and “how was it discovered” in a way that make them amenable to quantitative interrogation.  This is no small matter because in these two papers the authors pursue two fascinating questions: (1) what balance does a scientific community strike between pursuing tradition and pursuing innovation as the knowledge network grows; and (2) what can be done to maximize the exploration of such knowledge networks?

The answer to the first question is given in the longer of their two sociology papers (heavy reading for a poor physicist, but worth every ounce of effort).  As the knowledge network grows, research becomes more intensive and localized on already well-explored nodes and well-explored links, i.e., research favors tradition.  Innovation, exploring or seeking new nodes and links, is marginalized and receives less attention.  The authors connect this leaning in to tradition and leaning away from innovation to numerous factors, including some of the usual suspects like pressure to achieve high publication and citation rates for job security and job advancement.

In their second, shorter, paper they examine their newly quantified knowledge network from the perspective of maximizing discovery, defined as discovering new links and nodes in the network.  They find that when the knowledge network is young the approach of tradition, a localized search moving outward from central nodes (important molecules), is efficient.  But as the knowledge network grows this approach becomes more inefficient, even though this is the strategy that becomes more favored and represented in the published literature over time.

They suggest a number of policy remedies that would trickle down to individual discoverers by enacting change at the community level:

“Thus, science policy could improve the efficiency of discovery by subsidizing more risky strategies, incentivizing strategy diversity, and encouraging publication of failed experiments…Policymakers could design institutions that cultivate intelligent risk-taking by shifting evaluation from the individual to the group…[Policymakers] could also fund promising individuals rather than projects…Science and technology policy might also promote risky experiments with large potential benefits by lowering barriers to entry and championing radical ideas…”

[Rzhetsky et al., PNAS vol. 112, no. 47, p. 14573 (2015)]

As always though, I remain most concerned with how the individual can take action: how, with my own two hands and one mind, can I weave outward and affect change in the shape and size of the known web of knowledge, especially in my own field of neutrino physics?  If I combine what I’ve read in these fascinating sociology papers with my thoughts in “A Good Map is Hard to Find”, then I formulate an idea: my own two hands and lone mind can make one PowerPoint.

Now, I’ve been invited to attend a workshop to discuss possibilities for discovering new physics in a newly observed reaction called coherent elastic neutrino nuclear scattering, or CEvNS (i.e., a neutrino bounces off the nucleus in an atom as if it were one solid unit, instead of bouncing off of one proton or one neutron in the nucleus).  Workshops to produce agendas, devise long-term strategy, and draft roadmaps and white papers are ubiquitous in physics (and other sciences).  It’s how communities foster consensus on “what to do next.”

To me, an agenda-setting, roadmap-writing workshop seems like the perfect time to field test the idea of a “discovery call”: a voluntary, open-science call to action to trial scientific discovery strategies.  A “discovery call” is something you can talk about with colleagues, add to a website, or put on a PowerPoint slide.  The discovery call I’ll be pitching is as follows:  in physics, particles are analogous to molecules and particle interactions and mechanisms are analogous to connections between molecules.  Can we build a network map of published trends in our area of interest, CEvNS, and consider new strategies to maximize our network coverage with minimal experiments?  And can we take this a step further and build two other deeply analogous maps to use for comparison: one for neutrino neutral current interactions (i.e., where a neutrino bounces off another particle) and one for neutrino charged current interactions (where a neutrino bounces off of another particle, changing particle type in the process)?  It would be a way to provide a roadmap with a greater degree of informed choice about how, and how well, we’ve explored a given microcosm.

It seems to me that we have an opportunity to leverage our own history to help point our compass toward discovery, and to be able to see where untried paths have been neglected but might now be the roads best taken.  Perhaps today is the time to map what is known, with greater awareness and more practical purpose, so that tomorrow we can stand at discovery’s edge.

 

Interesting Stuff Related to This Post

 

  1. Jacob G. Foster, Andrea Rzhetsky, and James A. Evans, “Tradition and Innovation in Scientist’s Research Strategies”, American Sociological Review, volume 80, issue 5, pages 875-908 (October 1, 2015).
  2. Andrea Rzhetsky, Jacob G. Foster, Ian T. Foster, et al., “Choosing experiments to accelerate collective discovery,” Proceedings of the National Academy of Sciences of the United States of America (PNAS), volume 112, issue 47, pages 14569-14574 (November 24, 2015).

 

Related Content on The Insightful Scientist

 

Blog Posts:

 

 

How to cite this post in a reference list:

 

Bernadette K. Cogswell, “At Discovery’s Edge”, The Insightful Scientist Blog, September 21, 2018, https://insightfulscientist.com/blog/2018/at-discoverys-edge.

 

[Page feature photo:  A dewy spider’s web in Golcar, United Kingdom. Photo by michael podger on Unsplash.]