GRAND CHALLENGES AND INDUCTIVE METHODS: RIGOR WITHOUT RIGOR MORTIS
Science is built from basically two pillars: inductive and deductive approaches. The differences may change slightly from area to area, but fundamentally: deductive approaches use first principles to build knowledge, whereas inductive approaches start from experiments to make sense of reality, building up theories.
Science is built from basically two pillars: inductive and deductive approaches. The differences may change slightly from area to area, but fundamentally: deductive approaches use first principles to build knowledge, whereas inductive approaches start from experiments to make sense of reality, building up theories; here, it is defended case studies to build up theories, as an inductive method to grand challenges/complex problems. They are both important, and they may be seen as two opposite but complementary engines of any scientific branch; even though different branches may tend toward one approach, they are both important and must coexist in a health discipline.
The editorial assay presented herein makes the case that inductive methods may be the best option when it comes to working with grand challenges. As we are going to see, essentially, those methods are not constrained to any specific set of theories, or framework. This makes them ideal to tackle gran challenges.
One examples of deductive reasoning in science is the general theory of relativity (physics): built from first principles, and then proven true by experiments with the eclipse of the sun. One example of inductive reasoning are the laws of Newton, built from experiments. One interesting case is psychology: started out as a mainly deductive and speculative science, and now is also an inductive science, as so a Nobel prize was given to Daniel Kahneman; several psychologists built their theories from interviews, case studies and similar.
The authors of the editorial bring some interesting points: from now on, we are not going to mention directly the paper, since this commentary is about this paper; just accept as default that we are mentioning the paper.
Grand Challenges (GC) may be discrete with a clear endpoint or broad and open-ended. The difference is basically that the former has a final solution, well-defined, whereas the latter has an ill-posed solution. Their examples are, respectively, landing a rover on Mars or developing a Zika vaccine, curing cancer or eliminating poverty; we are going to mention AI as an open-ended GC, which may be seen as open-ended since it is here to stay.
Another interesting point is regarding that some BC “beginning as primarily technical problems and then shifting to social concerns”. Take the case of Artificial Intelligence (AI). For quite a while it was just a technical problem. As brings up Pires (2021a), neural networks were limited for quite a while, until new methods were developed, what gave rise to deep learning. Nowadays, it is to some extent amazing what those models can do, from voice recognition, medical diagnosis to self-driving futuristic cars. Nonetheless, several authors brought up the need to concern about artificial intelligence as a social issue rather than just computer science; Amershi and Vorvoreanu (2021) talks about research on Microsoft about AI centered in humans, and O’Neil (2016) alerts about several social issues with AI in case it is not taken seriously; Pires (2021a) highlights that it may be taking decisions for us soon.
One interesting point the authors bring is the application of induction by using case studies, to create theories. Indeed, psychology and social sciences gained a lot from those approaches. Brené Brown built a whole career studying shame, using those approaches.
When the authors bring up ethnography, and point out “Observation can reveal what people cannot or will not express”. Malcolm Gladwell discussed on his book Outliers how different cultures deal with authorities; that may influence a lot what we see when studying systems inside these systems, e.g., organizations. It seems that in South Korea planes were going down since their culture does not value direct confrontation, which made it hard to communicate when something went wrong. If we bring that to an organization, it means that people will not speak up when needed to their superiors. Charles Duhigg in The Power of Habit highlight how inefficient habits can destroy organizations, and that may have cultural factors. Malcolm Gladwell also mentions that in some cultures, the receiver is responsible for understanding the message, whereas in others, the speaker is responsible to be clear. Taking this observation say to an organization, that may influence how people receive feedback; Thanks for the Feedback: The Science and Art of Receiving Feedback Well by Douglas Stone and Sheila Heen discusses those possible issue with feedback.
Regarding the remark “Its use of observation illuminates rituals, non-verbal cues, artifacts, and the use of physical space”, an interesting case was the attempt to adapt the Japanese model of Toyota to Brazil: one of the biggest issues was the value the Japanese give to formal education, discipline and “the whole”, Brazilians share opposite values, that may have influenced negatively when importing this industrial culture; Brazilian culture is closed to United States. See curious accounts from Benjamin (2008).
When the authors make the remark “Inductive methods are particularly able to address these substantial problems.” Not even sure we can handle otherwise. The biggest issue of GC is that they “are complex problems with significant implications, unknown solutions, and intertwined and evolving technical and social interactions.” This is already well-known is several areas, in some areas they are called complex systems, complex networks, systems biology and more. What they all have in common is a complex nature, different from complicate: they exist from interaction of a huge number of independent agents, giving rise to emergent properties/behavior. Their inner workings can be overwhelming to be grasped, as so computer simulations gained considerable attention in the last decades, being “super” in “computers” one of the highest terms in the last decades, alongside cloud computing.
How inductive methods can support BC solutions
This excerpt “Creating novel ideas that can contribute to solving and explaining grand challenges is well suited to inductive methods” is the core of startups, wherein new approaches and ideas are used to solve big problems, most like, they can all be classified as GCs.
Regarding the observations that the demand for “personalized medicine persists” and “Innovation processes are especially relevant to grand challenges because many of them… include significant technical problems”, even though Pires (2019) does not mention GC directly, the idea was similar: innovation by startups in healthcare may be the way to solve this specific GC related to medicine: how to keep the treatments personalized without elevating costs?
Still on the medicine issue, “science-first” vs “clinically driven”, regarding the former being apparently more efficient, the authors bring up another study:
“…she developed a provocative theoretical argument that the ‘science-first’ innovation process favored by funding sources for more than 30 years may be less effective for dealing with the complexity of human disease than the more holistic ‘clinically driven’ innovation process that previously dominated…”
Two points can be made. Somehow medicine already recognized that. P4 medicine is a new trend in medicine, and one of the Ps stands for Participatory (Fiala et al, 2019). Another point is that medicine itself is evolving constantly, and the difficulties to understand medicine can be seen on how hard it is to create models. We also have the concept of evidence-based medicine, which is in some level inductive; maybe the best example of deducting approaches comes from physics, outside that, inductive approach most likely is the best approach.
Regarding the excerpt “Addressing grand challenges requires thinking ‘big’ and thinking ‘new.’”, science was grounded for quite a while on “thinking big.” It has already been reported in several independent and out-of-academic real writings the need to think big and out of the box. Would Einstein have thought big should he had been stayed in the academic world? Some writings suggest that not, it seems he gave a declaration claiming that most likely he would have caught trapped on endless trivial publications. Another issue that we face on the same direction is just reporting what goes right, it goes against thinking big, and may create what some calls the confirmation bias; see interesting discussions from Moosa (2018).
One interesting point that the authors bring when using the concept of “rigor without mortis” is the fact that the canned style of writing can be quite limiting for some areas: it seems to favor certain type of researches, and punish others; this may happen even with areas such as biomathematics/bioinformatics that requires the research to be innovative, and it may require some freedom to decide the best way to present and defend their findings; see previous paragraphs. It may be the case that this format-forcing culture was created by publishing houses, each time more and more present and dominating the scene, see interesting discussions from Moosa (2018); some argues that now they indirectly work as professorship selection board, since accepted papers on big publishing house will decide whether or not someone is accepted as professor/research on big institutions.
In conclusion, on this commentary we have discussed the editorial paper “Grand challenges and inductive methods: rigor without rigor mortis”, our goal was to present some criticism on the aforementioned editorial paper, focused on Grand Challenges (BC). BC are problems with a high level of complexity and no theory can grasp their inner workings, they can fall into the general category called complex systems. The inductive methods can be the best bet you can make on making sense of their nature, it happens because inductive reasoning is not constrained to any predefined theoretical framework; it does not mean they do not follow scientific rigor. Furthermore, evidences have been showing that case studies can be a promising path to build knowledge for BC, from specific cases to general ones, a sort of bottom approach, focused on data. BC can appear in several areas, including medicine and social problems.
O’Neil, C. Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy. Crown: 6 setembro 2016.
Amershi, S.; Vorvoreanu M. Create human-centered AI with the Human-AI eXperience (HAX) Toolkit webinar. Microsoft Research. Aug 6, 2021.
Pires JG. ALGUNS INSIGHTS EM STARTUPS: VENCENDO O DILEMA DA ‘PERSONALIZAÇÃO VS. CUSTO’ DA MEDICINA DE PRECISÃO?. Rev. G&S [Internet]. 3º de junho de 2019 [citado 11º de janeiro de 2022];10(2):261–75. Disponível em: https://periodicos.unb.br/index.php/rgs/article/view/24842
Clare Fiala, Jennifer Taher, Eleftherios P Diamandis, P4 Medicine or O4 Medicine? Hippocrates Provides the Answer, The Journal of Applied Laboratory Medicine, Volume 4, Issue 1, 1 July 2019, Pages 108–119, https://doi.org/10.1373/jalm.2018.028613
Imad A. Moosa. Publish Or Perish: Perceived Benefits Versus Unintended Consequences. Edward Elgar Publishing (26 janeiro 2018)
Benjamin Coriat. Pensar Pelo Avesso. Revan; 2ª edição (26 março 2008)
 The authors define Grand challenges as “highly significant yet potentially solvable problems.”
 Whole-cell model | in conversation with Jonathan Karr. https://www.youtube.com/watch?v=EGpq49y33dg&t=1023s
Commentary | assignment | 2021
GRAND CHALLENGES AND INDUCTIVE METHODS: RIGOR WITHOUT RIGOR MORTIS
Jorge Guerra Pires, PhD, firstname.lastname@example.org
Requested by student as a paid service
Do you need help with you PhD, just let me know!