Coffee Science methodology Episode 1: The vision

This is the first episode of a series going through theory of science and research design to support the global coffee community with solid ground when building theories and practices for education, product development and quality control. This is a pretty long and deep series of podcast and only the most patient and eager listener will stay through it all and reap the fruits of their patience. We will provide some very specific guidelines and also be completely honest when going through all the theories in roasting and sensory education that we find both wrong and misleading. 

Transcript

Welcome to this podcast series about coffee science. It is for the patient listener who wants the full story of CoffeeMind’s approach to a science based education which is developed during our involvement with research at the University of Copenhagen and our work with SCAE education and research activities as well as teaching thousands of students at London School of coffee and CoffeeMind academy in Copenhagen and consulting CoffeePreneurs worldwide since 2007. This podcast series might be long and with a lot of details but I have decided to hold nothing back, which means that a patient listener will get rewarded with a deep understanding of how we in CoffeeMind work with science based theories for education and consultancy purposes. Something you might find interesting if you work with education either under some of the established organisations such as SCA, CQI or other coffee education systems or if you work internally with education in a company. Or even if you have coffee as a hobby and settle for no less than the full story. If you find this podcast too long and detailed it is not because I’m boring. It is because you fall outside the intended audience. Our podcast called ‘Coffee-Science for CoffeePreneurs’ is for the coffee professional who already has experience and knows the fundamental concepts of coffee roasting and sensory methodology. If you don’t and still want to learn you can still listen to this podcast and learn but you might have to plan to pause a few times and browse the internet, books and ask a coffee professional or two in the process to clarify concepts and ideas.

CoffeeMind’s mission is to bring scientific methods to the coffee community to help provide clarity for CoffeePreneurs to support them getting on the quickest, cheapest and least risky road to the success they dream about. This podcast hopes to clean up the mess when it comes to some of the most fundamental concepts used in the coffee business but also to start a conversation about it so that we can get to the next level on these subjects together.

Firstly, I have to admit that it comes out of a purely personal frustration Ida and I experience in the very first hour of all of our sensory and roasting classes. As trained scientists we have to spend a lot of time dismantling a lot of misleading ideas and concepts that have been gathered in the specialty coffee community over the last 20-30 years. It is quite a waste of time and a lot of repetition which by the years is becoming more and more frustrating. Outside our classes we often talk about something we know to be true from a scientific point of view only to be met with an attitude of ‘well that is just your opinion”. What we say is not just our opinion but our best efforts to be aligned with the scientific method as we see it which we don’t mind to discuss but for this to happen we need to go a bit deeper. It seems that the global coffee community itself has not really started this deeper conversation about scientific principles which is needed if we as a community want to be better at distinguishing between opinion and qualified knowledge. Often this kind of qualified knowledge would be called evidence based knowledge, but I avoided this word as that might imply that only knowledge published in a scientific paper would be good enough and that is not my opinion. As long as the fundamental scientific principles are correctly applied and aligned with established theory of science and research design it is qualified knowledge even if it is not published scientifically. Most experience based opinions are not bad at all as they could be a good point of departure for a hypothesis to be tested but before it is tested it is just a qualified opinion and not true knowledge! I feel there is a specific need for the content in this podcast series because much of the new research projects coming out from universities these days seem more focused on generating new knowledge than dismantling the misleading concepts that have flooded the specialty coffee community for many years and causes a lot of daily confusion in product development and quality control processes. It is great with new research but we also need to address the misunderstandings of fundamental concepts and methods used daily in the global coffee community.

This podcast series is not only addressing the vague general concepts used in the community and education systems but also the direct misuse of scientific concepts, methods, and scientific conclusions used daily by CoffeePreneurs in their businesses. The purpose of the podcast is not just to tear down but to build up by clarifying scientific methodology with the purpose of brining a new foundation for the community on which we can build precise theories that does not waste anybody’s time and are precise and specific enough for the community to be results oriented when developing products and controlling quality across relevant process parameters. Particularly our focus is to provide a specific and clear foundation for the education systems that seem to be burdened by decades of opinion driven models and claims that just don’t keep up with scientific principles. We do not rule out hand craft and intuition as an important factor as a coffee professional but using scientific theories and concepts wrong is not leaving more space for intuition and handcraft. It only creates confusion. And a confused mind is not more intuitive nor better and carrying out hand craft. My coffee friend Stephen Lee says this beautifully in a interview (link to YouTube), where he uses his experience as a musician to provide a great model of the role of the scientific/technical aspect of art: You need to be so skilled in the technical aspect of what you are doing that at some point you forget about the technicalities because your technical skills has now turned into habit and only at this level you can focus on the art itself. We need clarity on the scientific/technical level before we can elevate ourselves to artists and as long as we are confused about this we struggle. The art is the master and the technicalities and science the slave. That is the goal. But as long as we don’t master the science, we are slaves of our bad and imprecise use of what we think is good science.

After almost two decades working with coffee education from a scientific perspective it  appears to me the only two fundamental misleadings approaches to education are responsible for an impressive amount of confusion in our community. 

1. The first is the failure to mix technical technicalities with preferences. Hunting for a sweet spot, a universal superior quality or very specific technical recommendation as if there is only one obvious or true optimal flavour of coffee we are all hunting for. Technicalities should be about control and how to navigate the general terrain and once you master this, you can go hunting for different flavours. And once you master the technicalities you can create different flavours to be able to strategically create different products to different customer segments.  First you learn how to drive and then later you can go to different places given a map and a destination. In education focus should be on how to control flavour diversity and how to map this to different customer  segments. Companies can choose broad or narrow product ranges and a broad or narrow customer segment depending on the business vision but education has to be broad and inclusive when teaching technicalities and not paint students into a preference corner as part of the education! Once you master the technicalities you can go narrow. But education has to keep the big picture in mind.
2. Obsession with small details of process parameters. Small differences are only relevant for people who have worked so much with coffee that only them and a few other people in the world would be able to pick up the nuances. Ok. Fair enough. There is nothing wrong with that in itself. The problem is that these detail-obsessive theories often set out the whole global coffee community chasing unicorns because in reality many of these alleged small differences are just good stories. If you describe a theory that predicts an extremely small difference that is vaguely defined nobody could really catch the theory of being wrong. Therefore it survives as a convincing narrative even though nobody has ever really gained from it because it did not make a difference in the first place!

I know that the above sounds obvious and that you perhaps don’t expect that this criticism will be relevant for most of the practices we have in the specialty coffee business world with, but the patient listener, who are willing to go through the whole podcast series, will be surprised how most of the theories in roasting and sensory education worldwide are based on these fundamental errors!

It is a bit ironic that everybody wants science on their side, and I really think that some people are quite unscrupulous claiming to have built their knowledge and education concepts on scientific principles when somebody with a scientific background quickly can see through this in seconds. And the loser in this game is the hard-working CoffeePreneur who are undertaking a huge risk in an uncertain adventure where error and time wasters are a big problem. Just because you know how to jot down some numbers, use a calculator, set up simple calculations and can name drop a few molecules does not make your methods and concepts scientific!

My purpose of this podcast is NOT to scare everybody without a science degree from using scientific methods. I’m not trying to create a situation where we should only trust statements from people with a minimum of a Bachelor of Science! In fact, often because scientists are educated as specialists in a field they often make mistakes at the level of scientific theory when trying to understand and collaborate with other scientists or even worse: if they don’t understand other fields of science they have a tendency to look down on or underestimate the rigorousness and complexity of other sciences such as we see way to often when chemists are either collaborating or making assumptions about sensory properties of coffee just by knowing the chemistry, which so often has been demonstrated not to be possible! You need to taste the coffee to say anything about how it tastes so you can’t skip the sensory tests just because you know the chemical aspects of the coffees and some chemists are a bit reluctant to get this. So this podcast might be as useful for scientists to get a better grip of the foundation for all sciences so that they can better collaborate with other scientists. Scientists often go straight to the core material of the science and theory of science and research design as such is often not dealt with other than a small class during their education they never really found interesting. So my hope with this podcast is the opposite of trying to scare non-scientists into using science! I’m trying to make the scientific methodology simple and specific enough for everybody to start using it correctly regardless of degree or not. I have done more than 30 research projects of which 9 are published in scientific papers and there will be two more in the first quarter of 2022. My point is, that I don’t have a PhD degree, which is the standard academic education to become a publishing researcher but that does not stop me from doing scientific research. Normally a PhD project ends up with 3-4 articles and I have 9 and counting so I’m an example of the fact that you can do science without the full formal training to do so. My opinion is that the scientific principles are useful for everybody and when looking at them they are more like a series of common-sense principles when it comes to the effort of making an experiment and making sure that what you do is correct which is necessary to trust the conclusions in the project. My opinion is that these principles can be understood and applied regardless of educational background and therefore it is my hope that they can be spread wide and far in the global coffee community regardless of educational background. I have done research projects with both Tim Wendelboe and Rob Hoos who are not scientists and I don’t recall that we had a single controversy about methodology during any of those projects and I really felt we were perfectly aligned when it came to what to do, how to do it and correct distinctions between hypothesis and facts. If you are just honest about what you know and don’t know and also are careful to set everything up correctly to not make conceptual mistakes along the way, the rest is common sense. It is really my hope with this podcast to make fertile ground for future discussions not to be about if I’m right as a person or not but if I’m using the right concepts and methods to build my theories and claims. I’m just so tired of being in discussions where people think that I’m going personally after somebody else just because I criticize a theory. It is my opinion that the global community does not have time for this type of ego bumping. The question is not who is right or who is wrong but rather which principles and theories are best at helping the global coffee community to maximize its purpose: to create amazing social aesthetic experiences for as many as possible through good coffee experiences. Anything that supports this mission is right and anything hindering this is wrong and the questions of who is right and wrong is hindering dialogue and clarity. I’m stressing this now because what you will hear later is very harsh criticism of some of the most established concepts used in the fields we are working in namely coffee roasting and sensory methodology and we don’t have time for imprecise or directly wrong concepts. We are here to support the community and time wasting is the worst type of waste so whatever we can do we will do to avoid this. I have compiled a long list of references for this podcast series that you can access in the podcast notes for everybody to pursue the concepts on their own and take me out of the equation. Of course I have made mistakes and wrong conclusions in this podcast series but by providing a comprehensive reference list I hope that we can start talking about right or wrong concepts and not about who's right or who is wrong when feedback starts coming my way based on this podcast. I have really tried to take myself out of the equation and just play the role of the messenger for classical theory of science and established research design methodology but of course I really try to expand all this into all corners of what we are doing in education, consultancy and research so I look forward to getting feedback on where I cross lines and make wrong conclusions but please keep your feedback aligned with the fundamental scientific concepts you find here or other places in other literature. Let’s try to develop the concepts we are working with and not waste time on ego bumping.

The fundamental concepts you operate by are important because the wrong concepts could lead you to waste a lot of time focusing on aspects and activities that do not lead to useful results because you would have to look and make an effort elsewhere to get useful results at all! It is my experience that thousands of people are wasting time looking in the wrong place for a result and I feel that we are amongst the few telling people that they are looking in the wrong direction because the fundamental concepts they are using are giving them a wrong focus. Sometimes we feel like the policeman in the classical “Streetlight effect” story where a policeman sees a person crawling under a streetlamp because he has lost his keys and even though he has been looking for a long time they just seem to be lost. After looking for a while the policeman asks the person to specify where he lost them and to his surprise the man points to a dark path leading to the streetlight! “But why are you looking here, then, if you lost them over there?” the policeman asks! “Because the light is better here” the man replies.

My point is that it is important where your focus is and who is guiding your focus. If you are not careful about the method used by your guide you might end up wasting a lot of time in areas that are completely fruitless for you and given the risk and resource scarcity for CoffeePreneurs this is not a small problem! It seems like the policeman could provide a better method: His torch could light up the dark path where you lost your keys where I think you will get a directly useful result to your problem.

As mentioned before, science is not just a few calculations and molecules taken out of context. Science is the continuously refined method behind the historical scientific breakthroughs. To those of you who don’t know, it might be useful to mention that often theory of science and research design is done by people with very different professions and are therefore two different subjects. Theory of science is generated by philosophers and research design comes from the mathematical tradition. Many philosophers have a bit of a skeptical approach to science claiming that science is ‘not a unified method anyway as different sciences use different methods and don’t agree which method is the correct one and history always proves that surprises are lurking around any corner for any theory anyway’. It will become clear through this podcast series that I’m really excited about the natural sciences with research design and statistical methods as foundation, but I also strongly believe that there is a limit to what can be explained by the natural sciences when it comes to social dynamics, aesthetics, the nature of consciousness and other experiences originating from the inner dimensions of the human domain. But I think for the purpose of clarifying the scientific methodology and how it is best used - and not used - by the specialty coffee community worldwide to improve the quality of the global dialogue about best practices in production and education it makes sense to assume that, for the natural sciences, there is a well-established body of fundamental concepts and methods that we can use to handle the physical and chemical entities we process as CoffeePreneurs. This is what I want to flesh out in this podcast series and for this perspective I’m without any hesitation excited regardless of what skeptical philosophers might say!

I know that this introduction seems a bit arrogant, but my point is exactly that this is not about you, me, or anybody else’s ego. The reason that Ida and I believe that we are often right and others wrong in the aforementioned discussions is not because we want to be right based on the title we got from our education, but rather some really specific and historically founded principles on which science as such is based. The purpose of this podcast is to put these principles out there so that our personalities can be taken out of the equation and the principles can be discussed and applied beyond anybody’s ego. And even though I want to take my ego out of the equation I still think it is useful for the listener to get a bit of background on why I’m qualified to even talk about these principles at all. Therefore I have made this quick intro to my background for understanding scientific methodology which also explains why I’m so obsessed about using and defending scientific methodology against misuse:

• My father was a high school teacher in physics and mathematics and taught me a lot about the evolution of classical physics into quantum mechanics and relativity theory and I did my high school dissertation about quantum physics.
• At the university I studied biology I got a thorough introduction to a lot of really basic scientific disciplines from basic molecules to advanced molecules, cellular structures, tissues and ecology as well as research design and statistics. I really loved this and for the first time in my life I really went for it and got good grades.
• After taking the basic education in biology I switched my major to philosophy where I got a general introduction to western philosophy and I specialized in comparative studies between western cognitive science and Tibetan Buddhism. I had to go really deep into the theory of science in order to make any sense of this and I used Husserl’s and Heidegger’s Phenomenology and Hermeneutics, classical western theory of science, Francisco Varela’s Neurophenomenology and Ken Wilber’s Integral theory to build the bridge between the very different cultural traditions.
• For two years, after studying biology and while studying philosophy, I had a student job teaching theory of science for biology students at the university
• From 2007 to 2012 I was teaching medical research design and statistics to students of medicine at the university of Copenhagen. Basically we had a pile of articles to go through where I was teaching them to understand the fundamental research design and statistical concepts and results in each article.
• I started my passionate pursuit of coffee science as early as 2003 where I did a project at the university explaining the chemistry of espresso brewing from the theory I found in Illy’s “Science of Quality” book. In 2004 I made a theoretical model of cappuccino foam together with Dairy Professor Richard Ipsen which I presented at Nordic Barista Cup in Iceland in 2004 which led to my first research funding which again resulted in my first scientific publication in 2011 in “International Dairy Journal”.
• I was running smaller projects at the university to kick-start my career as a self-taught scientist. In 2014 this led to SCAE funding a small part time position at Food Science in Copenhagen. Here my job was to convince Danish Bachelor and Masters Thesis students to work with coffee which gave the students some really relevant projects and we got a lot of relevant and cheap projects done funded by the Danish government. This was running for four years and led to around 20 projects of which 6 were scientifically published between 2014-2019.
• Ida Steen’s industrial PhD in CoffeeMind about sensory learning is currently my most intense involvement in science. My role in this project is to be the company-supervisor which involves a lot of project design, planning and execution in close collaboration with Ida, the university and our partners around the world running the Sensory Performance course associated with the research project. If you live in either South Korea or Saudi Arabia you would be able to participate in this course with Acts29 in South Korea and Arabian Coffee Institute in Saudi. 

The good theory

I will explain my approach to theory by addressing it at three different levels from short to long format and the first level is more like a short sentence capturing the most important parts. The sentence might be more correct than elegant but here we go:

Level 1

“A good coffee theory is capturing the most self-critical and simple way of saying something specific about a relevant, expected sensory difference which is specific in both how the event it is created, how it is evaluated and the audience for which the expected difference is expected to be relevant”

Level 2

is a longer and more specific list of features of a good theory:

1) Choose simplicity over complexity whenever possible

2) A small cause typically has a small effect unless there is a really well-established theory that explains otherwise. Expect small outcome differences from small differences in input parameters.

3) Form follows function: Design follows purpose. The type of method you choose needs to directly satisfy the purpose of doing the research project in the first place which is given and specified through the audience for whom the expected difference is both perceivable and relevant. 

4) Extremely specific in the description of the circumstances and the setup of the experiment (the input parameters)

5) Extremely specific and narrow in the description of the expected outcome of the experiment as explained by the basic scientific theory underlying the hypothesis of the research project. 

6) The everything else equal principle: Only change one input parameter at the time for the different samples and make sure these differences are the relevant differences to answer given the research question.

7) Has precise concepts for a non-reductionist model for relating first person human experience with the physical/chemical outer world.

7a) For sensory data keep preference (quality) data and intensity (quantity) data separate!

7b) Any claim of optimum for process parameters has to be related to a specified consumer segment and never just justified in a technical aspect itself. 

8) Systematically self-critic when it comes to making a possible wrong conclusion due to either personal interest in a certain outcome, confounding factors or coincidental outcomes.

Level 3

The following is a thorough explanation of each of the features in the aforementioned list:

1) A good theory is the simplest possible model of the system: If two explanations explains the same we will choose the simpler explanation over the more complex explanation. 

• 1a) From a theoretical perspective the simpler is seen as more ‘true’ because it better captures the ‘essence’ of the phenomena described
• 1b) From an execution perspective ‘complexity is the enemy of execution’
• 1c) a simple explanation is easier to make specific because you can focus on getting deep in a few concepts 
• 1d) Simplicity takes you faster to a level of intuition because you only focus on the fundamental concepts and are not confused by the more superficial emergent concepts. If you are a linguist, knowing about grammar makes you ‘see through language’. If you are a computer scientist, knowing about how binary logic is expressed in transistors and how simple logical circuits are built on top of that makes you ‘see through' the software coding. If you are a doctor, it is handy to remember anatomy by heart because you can derive understanding of different symptoms and possible cures from this fundament.

2) There seems to be an obsession about finding magic bullets in small aspects of process parameters from individual organic acids, rare and specific sensory descriptors (is this the inside or the outside of the mango peel?) to really small aspects of roasting processes. These theories sound like amazing stories but are often grounded in the blue air rather than science because the differences they work with are really questionable to have any hold in reality at all.
A sense of relevant size of magnitude of difference is where I feel that most education is most off when it comes to creating relevant content for education and consultancy. People seem to obsess about details which leaves little time – if any – to ask the bigger picture questions that relate what you do to the expected audience for what you do.

3) Has a specific purpose which translates into a specific benefit for somebody specific: somebody either gets a different (perhaps even better) experience, saves time or gets increased clarity over a situation. This is key in choosing a narrow research question and making sure the method chosen is designed specifically to answer the question in the most relevant way for the decided purpose. If you are not specific in who is expected to perceive (and gain from) the differences you can’t design your samples set as the magnitude of differences between the samples would have to be designed to be relevant for that particular audience.
This point has three sub-points:

• 3a) If you are looking to find small differences with certainty you need to budget for a lot of data points whereas if you expect huge differences in the dataset you can get away with a much smaller dataset and still get convincing results. If one person out of 10 reports a less intense headache from a pill it could be a coincidence and you would probably estimate that you need closer to 1000 people in the experiment group as 100 might still not get a clear result. Whereas if all 10 out of 10 told you that the headache completely disappeared by taking the pill, you could conclude that it works and you might just make some more experiments to see if different there could be subgroups of people who react differently than the 10 random people in the first group, but a basic trust in the mechanism of action is pretty clear even from a small dataset.
• 3b) The expected effect in the dataset of the experiment is important to estimate as it would answer the question of relevance as a small effect could be below relevance threshold even if the small effect is very well documented. Again, if all 10 in the group reported a decrease in headache, but it was on average only a 1 point reduction on a scale from 1 to 10 the effect was certain but few would pay anything for a 10% reduction of headache even though the effect is certain from a scientific and statistical point of view. The difference is simply too small. Also remember, that if it is ‘expensive’ in data points to document a small effect it is even more relevant to ask the question if the expected effect is practically relevant at all and therefore worth doing.
• 3c) If the expected difference is small, it is a relevant question if it exists at all since there can be a lot of personal bias in claiming it exists at all. A lot of valuable time can be lost if a lot of time is spent on something that might not even exist or be relevant or only relevant in a few situations.

4) A good theory is SPECIFIC both when it comes to describing the circumstances and the input parameters of a theory and the output parameters. This point is elaborated much more in the coming episode about Empiricism and Logical Positivism.

5) A good theory has a specific, very simple and with a very narrow predictive outcome. If a theory either describes an extremely small difference and/or is vague in specifying the outcome parameters thoroughly and clearly nobody would ever be able to really catch the theory being wrong because it did not really predict anything in the first place! If the theory does not dare to predict a certain and obvious outcome and in the same breath exclude other outcomes it does not say anything specific about any situation and hence can’t really be used for anything anyway. More about this in the later episode about Karl Popper’s Critical Rationalism.

6) When setting up an experiment you decide which question you would like to explore and design samples accordingly so that they are different exactly to the extent that they explore the central question. But here people often forget to keep everything else equal. I often see roasters experiment with drum speed, airflow, start condition and other subtle aspects of the roasting process and when I enquire about color measurements they seem surprised and tell me that they did not measure color because that was not the subject of the study. If you don’t plan your projects with color consistency between samples it will most likely lead to wrong conclusions on the cupping table because the difference in color will be a confounding factor driving the sensory differences on the table and not the more subtle roasting conditions you also changed and thought you were actually testing. Since you failed to keep ‘everything else equal’ you have created a mess on the cupping table that can’t answer your fundamental research question you actually pursued.

7) Clear boundary and correct boundary between the inner and the outer world of human experience. 

• 7a) First step is to respect the fundamental non-reductionist approach to this where we will simply give up on any psychologist or social relativist trying to explain away science as just biased or superficial opinions as well as we won’t accept any neurologist or chemist trying to explain away the human experience as something that is not interesting or reliable in itself. Psychologists and chemists can work together by respecting and using both methods in parallel without any of them trying to reduce one domain to the other. I can’t count the times I have been talking to scientists or read scientific articles written by chemist who does not understand how broad and correct sensory science is when used correctly and any scientist who has accepted the use of the SCA cupping form as an objective method in a research project should be arrested and punished by the science police. Not because of the nature of the data in the form but for the fundamental violation of scientific principles and methodology where the whole proud and rigorous tradition in sensory science is ignored. Anybody more interested in this should study Ken Wilber’s Integral theory and Francisco Varela’s Neurophenomenology.
• 7b) It will also show up as correct use of ‘subjective’, ‘objective’, ‘opinion’, ‘preference’ and ‘method’. I think it is about time to clean up the mess created by an old and outdated way to talk about ‘subjectivity’ and ‘objectivity’ as if what happens in the inner world of a human is filled with bias and bad judgement and only what happens in the outer world can be certain. I will suggest talking about ‘the inner’ and ‘the outer’ world and to reserve the term ‘objectively true’ to observations grounded in a correct use of ‘method’ and not with reference to the origin of the data. You can just as easily get lost in bad observation and bad application of method when you observe the outer world as you do the inner world so that is no reason to relate bias particularly to data coming from first person data. You can have data from the outer world and the inner world and both domains can be explored correctly in which case you have objective data or they can be explored incorrectly in which case you will have bad and biased data. Keep referring to ‘subjective data’ when talking about sensory data (as they have done in the new “SCA Sensory Coffee Sensory and Cupping Handbook”) only keeps sensory science often miscredited as ‘not a real science’ which is a completely wrong perception often still upheld even by other scientists who never had reasons to really reflect on the difference between the inner and outer world and methods for data collection in each. 
• 7c) In the domain of data from the inner world there is an extremely important distinction we still have not fully embraced in the coffee community which is the distinction between preference and observed intensity of a first person experience (could be flavour but also loudness of a sound or intensity of pain and so on). From a method perspective preference data and intensity data are objective data! Consumer data is not subjective. They are objective data about preferences. Adding the term subjective here would be outdated and confusing. Failure to distinguish between technical data and preference data often makes coffee professionals assume universality of quality and lead them to have a over simplified model for quality (such as is the case with the SCA Cupping form) or looking for small blueprints in roast logger software with curve pattern recognition or arbitrary calculations as if there is a really small window for ‘quality’ and only if you stay within this narrow parameter sweet spot you are doing it ‘correctly’. Good theories keep the technicalities separate from preferences and will show you the possible flavour modulation for each process parameter and then you can leave the question of preference to later to not risk missing out the complexity of this step of the product development process.

8) Self-critical to what can go wrong and how you can be wrong in your interpretation of your own data. How could I have designed this wrong? How can I calculate my data correctly without making mistakes? From a mathematical perspective you have the challenge that all measurement of samples has naturally expected variation which is why a good theory always is critical to individual outcomes and collects data to assess variation and the risk of claiming a difference when really it could just as well be a coincidental outcome. The best theories and experiments use statistics such as 95% confidence intervals for parameter estimates and p-values for claiming any sample differences.