2 The Process of Science
Learning Objectives
Understand the nature of science as a way of knowing.
Understand what distinguishes science from non-science
2.1 What is Science?
If you want to study the effect of bloodletting on a condition, divide the patients into two groups, perform bloodletting only on one group, watch both, and compare the results.
Does this quote, more than 1000 years old, represent the scientific method? You are likely aware of the basic components of the scientific method:
- Make an observation
- Pose a question
- Develop a testable hypothesis
- Design an experiment
- Collect data
- Analyze data
- Report your findings.
That’s a lot of steps and this short quote doesn’t capture all of them, but it’s author implicitly recognizes some crucial pieces. The quote is most directly related to Designing an experiment. We can easily envision the basic structure of a medical trial that seeks to understand how bloodletting affects the average condition of the patients. Some of those patients would undergo bloodletting and some would not (…“divide the patients into two groups, perform bloodletting only on one group…”). Then, we would collect data from all patients and analyze them (…watch both, and compare the results.”).
I included this quote because I am always amazed by how relatable it is to the current high-tech world of science. We do the same things in our labs every day, in every discipline. So, yes, this is most certainly a representation of the scientific method. But, by itself, the quote would get an F if proposed by an undergraduate student in a science course. Why? Because it is only an invitation for more detail. For scientists, nearly every word sparks curiosity for more detail. Let’s pick out some examples:
“If you want to study the effect of bloodletting…” — I know nothing about bloodletting. Does it involve a scapel? A tube? How much blood is let? Is it a fixed amount or does it change based on a person’s age, weight, size?
“If you want to study the effect of bloodletting on a condition…” — What condition will you test? Will you measure how a single condition responds or multiple conditions? Does it matter how long they have had the condition? How do you know they have the condition to begin with? Is it reported from another doctor? Is it self-reported?
“If you want to study the effect of bloodletting on a condition, divide the patients into two groups, perform bloodletting only on one group…” — Will the other group have a placebo? Will the bloodletting group know that they’re part of a research trial? Will that knowledge affect their response? Who will “perform the bloodletting”: the same doctor each time? If different people will perform the procedure, how will they be trained to maintain consistency?
“If you want to study the effect of bloodletting on a condition, divide the patients into two groups, perform bloodletting only on one group, watch both…” — How? Will patients be in the hospital the whole time? If they’re allowed to go home, how do you monitor recovery?
“If you want to study the effect of bloodletting on a condition, divide the patients into two groups, perform bloodletting only on one group, watch both, and compare the results.” — What will be compared? Recovery time? Whether they make a full recovery, regardless of time? The groups won’t be exactly the same no matter what you do, so how will you decide that bloodletting worked?
Phew. We managed to get through the whole sentence, but this is exhausting. Imagine that your professor is responding this way about something you wrote. You may want to scream “OF COURSE THE CONTROL WILL GET A PLACEBO!!” or “BLOODLETTING IS 3,000 YEARS OLD, JUST LOOK IT UP ALREADY!”. But science is written precisely to address potential weak points. It can be frustrating, because you feel like some things are just obvious, so they can be implied, or they’re just common knowledge, so you don’t need to cite anything to support them. This is where science writing differs from other more casual forms of writing. Most of science writing can feel like you’re spending a large amount of cognitive effort explaining things that are obvious.
2.2 Science versus Non-science
Scientific disciplines can seem stilting. Each branch of science is full of jargon and reading and conducting science often feels like learning a different language. Scientific discoveries are not made with sweeping insights generated geniuses pondering the depths of knowledge near the flickering light of a candle. They are made through lots of iterations of ideas that are continually subjected to the criticisms shown in the previous section. Science is a systematic way of discovering and organizing knowledge. It proceeds by examining testable hypotheses. Systematic and testable are key concepts that distinguish science from non-science.
Systematic means that science is done according to pre-specified plans. Each scientific experiment takes months or years of planning. Consider the development of the COVID-19 vaccines. They represent perhaps the fastest, most well-funded, and most comprehensive effort to complete a scientific study that has ever been conducted. Thousands of people were dedicated to the single goal of developing and testing a vaccine in a systematic fashion. Even with all of this effort, it still took months to formulate a testing plan, followed by months of conducting trials, and months of analyzing the data. That is not a critique. It is one of the most exciting scientific endeavors of our life-time. Now consider the rest of us. Underfunded, working on projects when time permits. That kind of “regular” science easily takes years. Without a systematic plan of action, their is no guarantee that the science itself would ever be completed.
Testable means the hypotheses could be tested. For the vaccines, the testable hypothesis was that people who were vaccinated would get infected less often that people who were not vaccinated. That is testable, because we can follow each person and collect data on their infection status. Now suppose someone offered an alternative hypothesis that the ancient god of Covidea intervened to induce immunity in trial participants. How could we test that? We couldn’t, because we can’t measure the supposed causal agent (Covidea). It doesn’t mean that it’s impossible or even wrong, it just isn’t possible to know using scientific inference.
2.3 Science Communication is Hard
Because science requires us to focus on the nitpicky details, science writing can often feel bland, formulaic, and overly detailed. But there is a benefit to having to constantly explain and justify your decisions as a scientist. First, it helps you to understand what you’ve actually done. As a scientist, I (Wesner) am a poor note taker. I always think that I’ll remember the temperature that I just recorded in tank 5. No need to get my notes wet now just to write down 23.5 degrees C. But I also know that a) I will most certainly have forgotten that number five minutes from now, and b) I will then have to explain why that number is missing from my data sheet in the paper that I’ll write 5 months from now. How can someone trust that I collected the data rigorously if I have to write: Data from tank 5 are missing because I forgot to write them down..
The other benefit of the details is that science writing has an egalitarian appeal to it. No matter how famous the scientist is, she cannot simply claim something as true without explaining the details. And on the flip side, no one is too junior to ask for proof. This requirement to document the details helps to prevent science from becoming overwhelmingly dictated by a few gatekeepers. That is not to say that science represents a utopia. Far, far from it. But requirements to “show your work” is one way to prevent it from becoming just another place where prestige and power matter more than truth.