Tuesday, June 27, 2017

The Broader Scientific Method

The scientific method is the whole process whereby science generates, testsvets, and evolves ideas. Most descriptions of the scientific method are at best partial representations focused on the actions of individual scientists, typically leaving off the subtle distinction between testing or experimenting and vetting, which includes collective actions of many people interacting in conversation as part of the scientific process. Here is one example list of steps in the scientific method:

which states:

1. Ask a question.
2. Do background research.
3. Construct a hypothesis.
4. Test your hypothesis by conducting an experiment.
5. Analyze your data and draw a conclusion.
6. Communicate your results.

This list and most lists like it are missing some of the most fundamental steps in the scientific process. Although part of science is conducted at the level of the individual scientist, the rest of it is conducted as an open conversation between different people. All scientists have biases, because they are people. Any of our conclusions could be wrong. They could be wrong because we did not ask the best question. They could be wrong because the background research neglected important relevant points. They could be wrong because our hypothesis was poorly posed, included logical flaws, or because it left out fundamental concepts. They could be wrong because the experiment was poorly designed. They could be wrong because of errors in our analysis or because we applied inappropriate analysis techniques. In the end, our conclusions could effectively be wrong because we miscommunicate them, using language that confuses the listeners.

Given the various pathways we could be wrong, science depends on having other people review our work, including people with biases different from our own. The formal peer review process we follow to publish results is an important first step, but by itself it is woefully inadequate and often results in wrong conclusions being deemed to be right, or correct conclusions being discarded. In a more general sense, science vets its results through an open conversation in which anyone who wishes to do so can challenge or debunk our conclusions for any reason. Similarly, just as anyone can debunk a scientific claim, the counter claim is also up for open discussion and refutation by the original scientist or by anyone else who wishes to participate. Science reaches consensus through years or decades of this type of open debate. The debate takes place in formally published papers, at scientific conferences, in the public media, on the Internet, and in informal conversations. In spite of consensus, no published conclusions are ever considered final, even if they are held at high confidence. For science to function properly, any idea must be open to being debunked at any time.

Modern science functions as an open marketplace of ideas. Most new ideas and some "tried and true" ideas are wrong. Often when we are wrong, we think we are right, so we need discussion with other people to protect us from our own folly. Just as science says that anyone is free to express any idea, anyone must be free to debunk any idea with evidence and logical argument. All ideas should be challenged, including those that are new and probably crazy as well as longstanding broadly accepted theories. As such theories are tested and attacked, if they are true, they can stand the test of time and our confidence in them increases. As ideas are challenged, improved, and communicated better as a result of being challenged, we come to understand them better and to see them better in their context. This open conversation causes our ideas to evolve with time, usually (but not always) toward something more correct. In that sense, we can think of scientific ideas as similar to Darwinian evolution of life. Ideas evolve in the face of criticism and further evidence, changing over time into something more fit for survival, or they get discarded as new ideas take over in the same context.

In a healthy scientific culture, even a child can suggest an idea, have it welcomed, and then have it challenged. In that case, a parent, a teacher, or peers may do the challenging. If being wrong is not disparaged (but is challenged) this approach might have positive impact on the psychology of most children. The school system trains children (and thus the adults that children become) that being wrong is bad, and that they need to seek truth from unquestioned authorities to determine what is correct. Instead of the logical pathway toward an idea being the most important aspect of education (honing the thinking process), whether the final answer is right or wrong is the priority. This educational philosophy has it backwards. Children need to learn to develop and process ideas, to accept challenge of those ideas, and to learn to discard and replace ideas when they are shown to be wrong. They also need to learn that people tend to hold on to wrong ideas, so we collectively depend on vigorous, unrestrained debate. Children need to expect to sometimes be offended as part of that debate, and they need to learn to accept that offense and move on with the conversation focused on the topic instead of the offense.

Teaching the processes that most likely lead to correct answers will ultimately help students to arrive at more correct answers in the future, even on questions that we have not yet thought to ask or questions whose answers are presently unknown. Emphasizing only the correct answers yields people who cannot think effectively. In the healthy open marketplace of ideas, people are constantly discussing, debunking, testing, and seeking good understanding of reality. Sometimes people will be wrong and will want to stay wrong. Sometimes their biases or priors mislead them. This stubbornness is normal. Scientists are not objective people. Science itself is also not completely objective, because it is done by scientists. The reasons science works so well are not that science is simply true or that scientists or experts are right all the time. It's not even that scientists tend to be really great at what they do (even when they are, their greatness is always insufficient). Science works because ideas are freely presented and vigorously tested and debated by other people, and because this debate influences the direction future science takes.

A broader restatement of the scientific method including both the scientist and the scientific community follows: 

You, the scientist should: 

1. Ask a question.
2. Do background research.
3. Construct a hypothesis.
4. Test your hypothesis by conducting an experiment.
5. Analyze your data and draw a conclusion.
6. Communicate your results.
7. Accept and respond to challenges from other people, and revise your questions, hypotheses, experiments, analysis techniques, conclusions, and communications where facts and logical arguments merit.

The scientific community should: 

Engage with the scientist's ideas, discussing each step in the process, criticizing, debunking, or adding clarifying ideas where appropriate. Different scientists from the first should repeat steps 1-7 to validate or refute the original work, then, they should modify questions, hypotheses, experiments, analysis techniques, and conclusions where evidence and logical argument dictate.

Saturday, March 11, 2017

The Quest for Diversity in Science and the Climate of Climate Change Science

I recently had the excellent opportunity to present a Teen Science Cafe' at the MiSci museum in Schenectady NY. The audience was great, and I enjoyed the visit. I joined a conversation before the event between a group of brilliant teenage girl participants and the museum curator. He provided them advice he gives to women in science: Stick together. I echo that view in the sense that any young scientist needs several things to be most successful: A good mentor/advisor, friends and colleagues for support and encouragement, and challenge from people who have different points of view.

Gaining support and feedback from people who are similar to us can provide  the courage to go on and the motivation to succeed. As such, university departments specializing in science need to be friendly and welcoming to all types of people.

However, young scientists also need to seek and receive often critical feedback from people who are different from themselves.  Scientific peer review can be brutal. Young scientists may thus benefit from exposure to diverse types of people and diverse points of view. The conversation about diversity in science often rightly includes motivation to increase the number of students and scientists from underrepresented groups. In some fields, especially physical sciences, women are underrepresented, as are Africans and latinos. In some other fields, men are underrepresented (women with PhDs in psychology outnumber men three to one).

It is not my view that gender parity is needed for the success of science. The free choices of individuals matter to gender balance in different fields. It is difficult to refute the notion that more men like mechanical engineering while more women prefer psychology or medical sciences, even after accounting for effects of any sexism. Regardless of the initial causes of these personal preferences, they really are personal preferences, and I think personal freedom for people to choose to follow their passion is more important than achieving gender balance. The causes of these different preferences could be social, biological, or some combination of the two, and it is not within my expertise to try to explain why, and it does not influence my argument. We cannot force people to choose to do that which does not interest them. Efforts to increase interest of girls (and boys, for that matter) in science are well intentioned but have limited success. When I manage a search committee to fill a job opening, I cannot start with a fifth grader. In my position, I want to include all people who are qualified and who want the job, and I think that where any systemic resistance to inclusion occurs, it should be corrected through positive action, but exclusion is not the only reason for gender disparities.

Even when gender parity is not achieved, scientists may gain advantage by seeking conversations with scientists of the opposite sex and from people from different demographic groups.

Now, don't get me wrong, progress in most of the physical or natural sciences, such as in understanding the nature of a thunderstorm, doesn't require information about whether the scientist is male or female, or black, white, or even purple, for that matter. Any differences between interacting scientists might catalyze opportunities for new insights. Women should thus seek collaboration with men, men with women, and we should all seek interaction with those of other races or ethnicities. I'm not suggesting that every time a male latino scientist has an idea that he should consider it necessary to march down the hall to seek an opinion from an Asian female. Instead, I think that simply having access to and occasionally working with different types of people can enrich the quality of our science overall. Different perspectives don't change the facts, but they might lead us to find previously unrecognized facts and ideas simply by nature of their different experiences. For example, a biologist who grew up in poverty as a farm worker might have insights relevant to the science because she worked in the fields as a youth. For the betterment of science, all types of people from all backgrounds should feel welcome to participate.

The most important interaction we need, though, is with scientists who might disagree with us, regardless of their demographics. This viewpoint diversity is probably the most needed type of diversity for the advancement of science. When there may be multiple views consistent with existing evidence, and perhaps only one or none of them are correct, science can ill afford blanket rejection of viewpoint diversity. Science stagnates when we exclude the perspectives of scientists with points of view that are different from the mainstream. In the end, it is the facts and evidence that lead to a conclusion, not points of view, but different points of view can lead to different ideas of how the facts fit together, and these ideas can mold the direction of the conversation. Occasionally a scientist makes a profound discovery that overturns the mainstream view.

Study of climate change induced by human activities including burning fossil fuels is one field in which some forms of viewpoint diversity are being shut out. I am aware of few climate scientists who argue that increasing carbon dioxide in the atmosphere would not increase the earth's temperature. Nearly every climate scientist agrees that human activities are warming the lower atmosphere, especially in the Arctic (myself included). A large majority of climate scientists agree that more than half of the warming in recent decades is a result of human activities (see the IPCC reports). This type of broad agreement is supported by multiple distinct arguments and analysis techniques. Yet even though the idea of human induced climate change has broad evidential support, the public discourse on this topic is rife with fallacies, especially the argument from authority and argumentum ad populum.

Beyond the points on which most climate scientists agree, there is substantial discussion regarding how much it is likely to warm and how that warming might influence changes in extreme weather events. I think that the most important unanswered questions remaining in climate science relate to how the weather will change with the climate, and how changes in the weather might feed back on the climate. This problem is immensely complicated and the science is still young. People who argue that discussion about climate change and extreme weather events (such as tornadoes, hurricanes, blizzards, and flooding rains) is over or that consensus has been reached on these issues are being dishonest. Scientists are making progress in these areas (such as uncovering evidence suggesting that in many parts of the warming world, both the amount of time between summer rainfall events and the intensity of the rainfall probably will increase), but the problem of the impact of climate change on extreme weather events is far from solved.

Sometimes behaviors of people who fight scientists who express different points of view are similar to behaviors of schoolyard bullies (one such point of view is expressed here). When people argue that "the discussion is over", they shut out conversation where it may be far from over. Some activists and scientists demonize and attempt to exclude from the conversation people who agree that climate change is occurring and is influenced by human activities, but who do not argue that all types of weather are getting worse, or who do not agree on solutions to the problem. These activists often create strawmen descriptions of the contrarian scientists' views in an effort to discredit them.

Rather than address the arguments of scientists who might cast doubt on prevailing views about extreme weather events, detractors label them as deniers or shills for the fossil fuel industry. The ad-hominem attack rears its ugly head. This behavior is inappropriate and unhelpful. Members of the public who remain on the fence on the climate change issue may be turned off by this behavior, and those who do not believe that climate change is occurring may take this behavior of scientists as evidence of a politically motivated conspiracy. Avoiding the appearance of such behavior requires us to patiently discuss errors where we see them, with respect and in light of evidence.

Those opposed to climate change science are often guilty of the same behaviors I describe above, but in the context of the scientific community, the opposite side holds most of the power and needs to be called out. To me both types of negative behavior are reprehensible. Surely we can do better.

Sunday, February 19, 2017

Consequences of Overspecialization in Science

Isaac Newton is famous for new ideas that transformed nearly every branch of science after his time. Known best for physics, his development of mathematical concepts set the stage for most of the advancement of the modern world. Other scientists before and after his time provided foundations of their own fields. Back in that time, little was known about complicated natural systems, including the structure of matter, electromagnetism, light, astronomy, biology, weather and climate, and nearly every other field of modern science.

With so little known, it was possible for a few brilliant minds to contribute deeply to our understanding of what ultimately became many different fields. Knowledge has now been accumulating in these fields for decades to centuries.

Today, a student might specialize in biology as an undergraduate, focus on marine biology for a Master's degree, then specialize for a PhD in understanding a type of skin growth on a rare species of sea slug. The newly minted PhD might be the world leader in analysis of that type of skin growth, but the same scientist might know relatively little about other branches of biological sciences. With so many different subfields that have developed out of many different past research projects, the days are long past when anyone can learn the whole of biological or even marine biological sciences.

Many remaining scientific debates relate to concepts that align in between fields. Although scientists working in individual fields may think that their fields are well understood, understanding of concepts at the boundaries between the fields may be less so.

As a professor, my first research proposal I submitted for funding to the National Science Foundation was aimed to better understand how weather signals interact with El NiƱo in the Pacific Ocean. Observations suggested that the condition of the ocean at a given time influences weather events, including windy periods, and that the winds then cause changes in the condition of the ocean.

Since my field was closest to atmospheric science, I submitted my proposal to the branch of the NSF that oversees atmospheric dynamics. However, the oceans component of the work necessitated that the proposal also get reviewed by a panel of oceanographers. The atmospheric science reviewers ranked the proposal well, but the oceanographers prevented it from being funded, because they thought my proposed analysis of observational data was not exciting.

Sometimes the subculture of scientists in one field resists changes initiated by scientists who started out in neighboring fields, and it can be difficult for scientists trained in one subfield to gain access to the community in a neighboring field.

Alfred Wegener was a prime example: He was a meteorologist who presented the concept of continental drift to a community of geologists who were clearly not ready for the idea and who on the whole did not accept it until decades later.

Bottom line: In order to make more efficient progress, scientists need to work to speak the languages of neighboring fields when they attempt to solve problems in those fields. They also need to not be quick to reject new ideas suggested by scientists from neighboring fields just because the authors do not know all of the jargon and the history in their own fields.

Sunday, January 22, 2017

Seattle Conference, Inauguration, and Protests

I arrived in Seattle Washington today to attend the annual meeting of the American Meteorological Society. Friday I watched footage of the rule of law, managed democracy in action, in the inauguration ceremony of Donald Trump. Yesterday I saw freedom of expression in action as people swarmed the streets of Seattle as part of the set of nationwide "women's marches" protesting the inauguration and Trump's rhetoric, history, and proposed policies. As this is an education blog, my point in discussing these issues here is to encourage people to listen to diverse perspectives different from their own. My purpose is not to side with one view or the other on the Trump presidency.

I took some time to mingle with the crowd, listening to their conversations and perspectives. Most of them aligned solidly with the political left. Yet, aside from their chants and signs, most of their conversations were about things they saw around them, their experience of the day, and their decisions for their next meals. Really similar, actually, to what I saw in Tea Party rallies. Also similar to my observations of videos of Palestinian and Israeli birthday parties posted on Youtube. As different as aspects of our views are, most of us are surprisingly similar. Some readers my trivialize my perspective here--I mean, obviously both groups are human, but that is precisely my point. It is becoming common practice for each side to dehumanize the other, and it does not make things any better. 

I recently posted on Twitter how I think that every person on the political right needs a friend on the left. Every person on the political left needs a friend on the right. We all too often listen just to people (and sources of news) with whom we already agree. Besides being intellectually weak, this behavior amplifies political polarization. We construct for ourselves echo chambers, where we tend to hear progressively more extreme views with which we see no reason to argue. People on the left and the right have also segregated themselves geographically. This separation leads many of them to conclude that basically everyone is like themselves, and that the distant strangers who disagree are simply deluded or even evil. Many people on the urban east and west coasts do not have many opportunities to meet conservatives. Many in rural America do not know people on the political left. 

In comparison with the past, I think there is relatively little hate left in the United States that is motivated by race, sex, or other characteristics of identity. Most of us mingle well with the opposite sex, other races, and people of other religions or belief systems. Of course hate motivated by those things still exists and still impacts people's lives, and it is an important problem for the individuals seriously impacted by it. I'm just claiming that it does not dominate our culture. 

Recent events suggest to me that most of the hate left in the United States is now aligned with politics. Abhorrent attacks on Trump supporters, including a man with a mental disability, along with other attacks going the other way support this view. 

People on the left, for the most part, did not hate George W. Bush because he was white. They disliked his conservative rhetoric and policies. Similarly, most people on the right did not dislike President Obama because he was black. They hated his left-leaning rhetoric. The truth of this point becomes obvious when the same people voice support for Clarence Thomas. I think we can all do better than this. There is no good reason to hate any person because of either their skin tone or their politics. 

Although it may be true that some ideas on either side really are simply wrong or harmful to humanity and human dignity, most of the time when we see the views of the opposition that way, it results from our own restatement of their position. The straw-man fallacy shows up frequently in both Tea Party protests and the ongoing anti-Trump marches. In the straw-man fallacy, we put up a false caricature of the opponent's view, then attack it as if it were their actual view. Sometimes we even do it while thinking our representation of their view is correct. Being wrong and not knowing it still feels like being right. Personally I would rather be less wrong. 

Obviously not everything either side claims about the other is a straw-man, but we owe everyone the common decency of an honest representation. Our politics can do better. My invitation to each of you is to seek out people with views different from your own. Question yourself and each other. Don't attempt to silence people with differing political views. Celebrate free expression, consider each other's views, and then stand up for what you think is right.