1 / 82

What is Science?

"...science is simply common sense at its best; that is, rigidly accurate in observation and merciless to fallacy in logic." Thomas Henry Huxley, 1880 "Scientists are critical realists.” John Polkinghorne

spence
Télécharger la présentation

What is Science?

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. "...science is simply common sense at its best; that is, rigidly accurate in observation and merciless to fallacy in logic." Thomas Henry Huxley, 1880 • "Scientists are critical realists.”John Polkinghorne • "Science is properly described as ‘organized skepticism,’ a realm in which nothing is to be accepted without question.” Philip W. Anderson • …nevertheless, and probably quite accurately, Margaret Wertheim replies with: "Science has always had a huge component of faith." • This latter statement reflects the idea that ultimately not everything (nothing?) can be proven to 100% confidence. A good scientist nevertheless allows that even those things she accepts on faith could very well be incorrect. What is Science?

  2. "…scientists are not a select few intelligent enough to think in terms of ‘broad sweeping theoretical laws and principles.’ Instead, scientists are people specifically trained to build models that incorporate theoretical assumptions and empirical evidence. Working with models is essential to the performance of their daily work; it allows them to construct arguments and to collect data."Peter Imhof • "Science is [best] understood by observing it than by trying to create a precise definition. The word science is derived from a Latin verb meaning "to know." Science is a way of knowing. It emerges from our curiosity about ourselves, the world, and the universe. Striving to understand seems to be one of our basic drives. At the heart of science are people asking questions about nature and believing that those questions are answerable." your text Science, A Few More Ideas

  3. Doing science involves: • Asking Good Questions. • Coming up with Good, Plausible Answers (a.k.a., hypotheses). • Testing These Hypotheses robustly, unambiguously, and honestly (the latter from the point of view of both yourself and that of others). • "Science is a creative human endeavor that involves asking questions, making observations, developing explanatory hypotheses, and testing those hypotheses.“ your lab text • It is "important for you to learn, by example and by practice, how the process of science works." your text • "Anyone going into biology expecting to find the sorts of exceptionless laws that characterize physics will be sorely disappointed." Ernst Mayr Doing Science

  4. Science is a means of timing-wasting avoidance • Doing science poorly (or not doing science at all) results in failing to answer questions efficiently. • Doing science poorly can result in wasting other’s time (with poorly thought out hypotheses or results). • The cost of wasting the time of others is ostracism—nobody wants to have their time wasted by incompetent boobs! • Nevertheless, often there is a fine line between doing difficult science and wasting time—this is one reason the easy-to-solve problems tend to be solved sooner. • The "open-mindedness" that non-scientists often feel comes from lacking a well-developed compunction to answer hard questions rigorously (I.e., robustly, unambiguously, and honestly). Time-Wasting Avoidance

  5. What questions do scientists tend to address? • More often than not the questions that are addressed first are those perceived to be both potentially fruitful and less difficult to answer. • For some questions science is willing to invest enormous amounts of resources (curing cancer, creating weapons of mass destruction during national military emergencies, etc.). • For other questions, science (or, more precisely, funding agencies) are unwilling to invest many if any resources. • The basic questions come down to: • Is the endpoint worthwhile? • Are the resources necessary to solve the problem in excess of the perceived worth? • Is the endpoint likely to be reached? Questions Addressed

  6. Is the endpoint worthwhile? • Are the resources necessary to solve the problem in excess of the perceived worth? • Is the endpoint likely to be reached? Questions Addressed • For conservative, applied research, using established techniques, the answers generally are yes, yes, and yes (even when the scientific questions aren't terribly interesting). • For speculative, basic, or extremely difficult research, the answers can be no, no, no. • Ultimately whether a question is pursued is a function of the amount of resources a society is willing to devote to science. • The consequence is that science does not always work toward its own goals with the efficiency it (or we) would prefer. • Wild card: questions (& means to answering questions) that are interesting in their own right.

  7. "Success in science is rewarded with attention. You gain full membership in the scientific community only by receiving the attention of your fellow scientists. Earning this attention ‘income’ is a prime motive for becoming a scientist and for practicing science. In order to maximize this income, you have to employ your own attention in the most productive way. It does not pay to find things out anew that have been discovered already. Nor is reinvention rewarding in terms of the attention paid. It pays to pay attention to the work done by others." Georg Franck • In school doing science well is rewarded with good grades in science class—the same skills that allow one to do science well will allow one to succeed in biology class: learning, understanding, synthesis, an ability to communicate your thoughts well, etc. Succeeding in Science

  8. Asking Good Questions • Forming Hypotheses • Testing Hypotheses Asking Good Questions • "Items investigated must be well defined, measurable, and controllable. The questions should be reasonable and consistent with existing bodies of knowledge. [Individuals] have a variety of ways to exclude wild speculations." your lab text • A good scientific question is one that may be answered through experiment, observation, or logical inference that is built upon previous experimentation or observation. • Beware of direct correlations vs. indirect correlations (cause and effect vs. “effect and effect”). • Questions are also judged on the worth one or many perceive to be associated with successfully answering that question.

  9. "Does exposure to ultraviolet radiation cause increased risk of skin cancer?" • "Does good nutrition lead to increased intelligence?" • "Why do cacti have spines?" • "Was the malignant tumor found in the lungs of a 70-year-old man caused by his 45-year habit of smoking cigarettes?" • Do good study habits result in good grades in science classes? • Though these are all good questions, they are not necessarily easy to answer. • Was Lee Harvey Oswald possessed by demons? • Bad question: • How do we define demon? • How do we determine whether L.H.O. was possessed by one? Good vs. Bad Questions

  10. Asking Good Questions • Forming Hypotheses • Testing Hypotheses Forming Hypotheses • “A hypothesis tentatively explains something observed.” your lab text • It is a proposed answer to a scientific question. • A good hypothesis satisfies the following criteria: • It supplies a testable mechanism. • It is not unnecessarily complicated (Ocham’s razor). • It conforms with existing knowledge. • It is falsifiable. • if something cannot be demonstrated to be incorrect then it cannot be demonstrated to be correct. • Hypotheses tend to gather favor if they could be but haven’t been demonstrated to be incorrect.

  11. "The test of a hypothesis may include experimentation, additional observations, or the synthesis of information from a variety of sources." your lab text • Remember that: • Hypotheses represent possible causes. • They reflect past experience with similar questions. • Multiple hypothesis should be proposed if possible. • Hypotheses should be testable via the hypothetico-deductive approach. • Hypotheses can be eliminated. • But hypotheses cannot be confirmed with absolute certainty. • Note that in practice hypotheses are a dime a dozen—easy to propose, difficult to prove. • Also, very few are sufficiently comprehensive nor stand up sufficiently to the test of time and experimentation to achieve the status of a theory. More on Hypotheses

  12. A hypothesis becomes a theory following lots of testing (i.e., attempted falsifications), all of which fail to disprove the hypothesis. • An important aspect of this testing is that it is done by more than one (ideally by many) groups using more than one (ideally many) independent techniques. • In other words, a theory is a very robustly supported hypothesis. • Since, by definition, a theory has gone through considerable criticism and attempted falsifications, it is very unlikely that you or me or anyone we know or admire is going to successfully demonstrate that a well-established theory is false. • E.g., Darwin’s Theory of Evolution (which in lay language we would describe as a fact). Scientific Theories

  13. A fact is what is witnessed upon observation. • A scientific fact is only as good as the observer, method of observation, and degree to which the environment is sufficiently controlled during the observation. • Thus, facts are very fallible and must always be considered suspect especially if they are contrary to established theory and are not repeatable under well-controlled conditions. • In other words, extraordinary claims require extraordinary proof. • In the semantics of science, a fact does not have explanatory or predictive power—one speaks of hypotheses and theories as ways of organizing, explaining, and extrapolating from facts. • This is why a scientist speaks of the theory rather than the fact of evolution. Scientific Facts

  14. Scientific Law • A law is "a statement of order or relation holding for certain phenomena that so far as is known is invariable under the given conditions” Webster • In other words, a law, as far as we can tell, is an infallibly robust hypothesis. • In modern science it is considered reckless to call a theory a law.

  15. Asking Good Questions • Forming Hypotheses • Testing Hypotheses Scientific Reasoning • A key aspect of doing science is the reasoning that goes into the designing experiments, something that I'm designating here as scientific reasoning. • To test hypotheses you have to understand how to go about scientific reasoning. • There are two general categories of scientific reasoning: • Inductive reasoning. • Deductive reasoning. • It is the latter that is usually employed in the course of testing hypotheses and designing experiments. • Inductive reasoning involves the gathering of observations and hypotheses into a unifying whole.

  16. Inductive reasoning is associated with great ideas but not necessarily very good experimental design. • For example, Darwin's theory of evolution by natural selection was achieved via inductive reasoning: A great many observations were gathered and a unifying theme was discovered. • While inductive reasoning does not make for good hypothesis testing, the results of inductive reasoning can typically supply fertile ground for hypothesis making. • Another word for inductive reasoning is synthesis. • Synthesis, in general, is analogous to the more specific synthesis observed in chemistry laboratories. That is, synthesis is the build-up of a different whole from smaller parts. Inductive Reasoning

  17. Inductive Reasoning • An example of a synthesis is the "Evolutionary Synthesis" from the middle of the twentieth century, which involved the building up, by inductive reasoning, of a theory of evolution that combined both Darwinian evolution and Mendelian genetics.

  18. "Many people associate the word discovery with science. Often, what they have in mind is the discovery of new facts. But accumulating facts is not really what science is about; a telephone book is a catalog of facts, but it has little to do with science. It is true that facts, in the form of observations and experimental results, are the prerequisites of science. What really advances science, however, is a new idea that collectively explains a number of observations that previously seemed to be unrelated. The most exciting ideas are those that explain the greatest variety of phenomenon. People like Newton, Darwin, and Einstein stand out in the history of science not because they discovered a great many facts but because they synthesized ideas with great explanatory power." your text Inductive Reasoning Except, of course, Darwin did discover a huge number of facts!

  19. Deductive reasoning is what biology is all about. • Deductive reasoning is an assumption of consistency. • Deductive reasoning is the application of generalizations to specific circumstances. • This is hardly a profound statement. It simply means the application of what we generally know to specific things that we don't yet fully understand. • More than anything else, introductory biology introduces students to a sampling of the general themes of biology. • With time you will learn to apply these themes to novel situations to deduce explanations for novel observations. • E.g., once you understand why lipids tend to not dissolve in water, but that carbohydrates do, you will be able to look at organic compounds that are new to you and make specific predictions as to their water solubility. Deductive Reasoning

  20. The process by which science typically progresses is employing a mechanism known as Hypothetico-Deductive Thinking. • This fancy phrase basically means that one understands new observations in light of previously learned or subsequently looked up general knowledge, & then phrases understanding as testable predictions. • I.e., deductive reasoning  hypothesis making. • The catch, of course, is that not all knowledge is correct, knowable, or even necessarily applicable to the new observation. • Furthermore, it isn't always obvious how to apply general knowledge to new observations. • When you have an interesting or important (and repeatable) observation that cannot be explained in detail by current scientific knowledge, what you have is the core of what I would call an interesting scientific question. Hypothetico-Deductive Thinking

  21. A triage is a means of effort-wasting avoidance. • In a wartime medical unit there are three types of patients: (i) those who will survive without medical intervention, (ii) those who will not survive even with medical intervention, and (iii) those who will survive but only given medial intervention. • If you have the resources to deal with only a limited number of patients, then you concentrate first on the latter. • What is being done is prioritizing. • In science usually the first questions answered are the most easily solved or most interesting. • Less-easily solved questions or less-interesting questions are solved next (if ever). • The least-interesting or most-difficult questions tend to be addressed last (often never). Doing Science as Triage

  22. Scientific prioritizing is why questions that many consider important (Why do we exist?) are typically never considered by scientists. • In a world of interesting, solvable problems, few rational individual commits enormous quantities of time and energy to questions that are not readily solved, no matter how interesting they may appear. • Think about your own life. When was the last time you elected to attain world peace and prosperity before dealing with more mundane concerns such as eating lunch or voiding your bladder? • A bad scientific question typically is one in which the one’s potential to answer the question, even given abundant technology and resources, is extremely limited. Not Addressing Important Questions

  23. Skepticism • Attitudes of skepticism derive from desires to avoid wasting one's time on questions perceived to be without significant usefulness. • This is why the burden to answer questions (demonstration of a lack of falsification of hypotheses) is placed on proponents of ideas rather than on the detractors. • Extraordinary claims—one's not consistent with an existing base of knowledge which so far has stood the test of time—typically demand extraordinary proof to be persuasive. • Such proof is found in rigorous, robust, and honest attempts at falsifying the hypothesis in an unambiguous manner.

  24. For many hypotheses existing technology and understanding is not sufficient to supply such proof, regardless of the efforts of proponents. • Such hypotheses are generally discarded by other scientists. • In other words, scientists are typically skeptical of claims that "fly in the face of reason," i.e., that are inconsistent with what is already known. • This is why scientists often come off as fairly conservative in terms of their acceptance of new ideas (a.k.a., have good B.S. meter). • They know how much work is required to test hypotheses—that making hypotheses is far easier than proving them. • Scientists, consequently, are far more interested in results of efforts to test hypotheses than they are in the hypotheses themselves. Liberal vs. Conservative

  25. Sometimes, when technologies and understanding catch up with speculations, speculated hypotheses turn out to be correct. • Proposed future utility of a given hypothesis, however, is no guarantee of present usefulness. • An otherwise empty promise of future utility should never be accepted instead of demon-strated usefulness of a hypothesis in the present. • This is why science fiction can be very cool but nevertheless is still fiction. • The only reasonable predictions of the future is extrapolation of from the past (i.e., that the future in some manner will resemble the present). • This is the utility of science (and history): efficiently and accurately defining the present and the past so as to predict the future consequences of present trends and actions. Predicting the Future

  26. "Another key feature of science is its progressive, self-correcting quality. A succession of scientists working on the same problem build on what has been learned earlier. It is also common for scientists to check on the conclusions of others by attempting to repeat observations and experiments. Among contemporary scientists working on the same question, there [is] both cooperation and competition. Scientists share information through publication, seminars, meetings, and personal communication. They also subject one another's work to careful scrutiny.“ you text • "In science seldom does a single test provide results that clearly support or falsify an hypothesis. In most cases the evidence serves to modify the hypothesis or the conditions of the experiment." your lab text Self Correction

  27. Self correction means that the testing of hypotheses is typically repeated by others so long as a hypothesis: • Impacts on the work of others (i.e., is important). • Is plausible (i.e., people are willing to believe that tests already performed could conceivable support the hypothesis). • Is testable by other means. • Is contrary to other's previous understanding. • Often if a claim is too outlandish then the burden of proof will fall on the claimant, and others will simply reject the claim. This is especially the case if others don't consider the claim to be especially important or plausible: • Look closely at claims of cold fusion? Yes! • Look closely at claims of perpetual motion? No! The Tao of Self Correction

  28. Asking Good Questions • Forming Hypotheses • Testing Hypotheses Experimentation • As you can see by our continued emphasis of this third section on “testing hypotheses,” a great deal of a scientist’s time is spent hypothesis testing rather than hypothesis making. • This is why the caricature of scientists is a person in a lab coat working at a bench.

  29. Asking Good Questions • Forming Hypotheses • Testing Hypotheses Experimentation • As you can see by our continued emphasis of this third section on “testing hypotheses,” a great deal of a scientist’s time is spent hypothesis testing rather than hypothesis making. • This is why the caricature of scientists is a person in a lab coat working at a bench. • "The most creative aspect of science is designing a test of your hypothesis that will provide unambiguous evidence to falsify or support a particular explanation… • …Scientists often design, critique, and modify a variety of experiments and other tests before they commit the time and resources to perform a single experiment."

  30. Variables • To understand how to successful perform experiments, one must understand how to successfully handle (and describe) variables. • A poorly run experiment often is as much a consequence of poor technique as a consequence of failure to properly identify and control variables. • We can divide variables into three general types: • Independent Variables • Dependent Variables • Controlled Variables • Your goal will be to distinguish between and otherwise not confuse these three variable types.

  31. Dependent Variable • Within an experiment, the dependent variable is that which is being measured, the variable that you hypothesize will change as a function of change in one or more independent variables. • Very often varying an independent variable will result in more than one measurable change in the system. • Consequently, it is routine for an experiment to have more than one possible dependent variable, though doing so will often increase the size and complexity of an experiment.

  32. Employing Dependent Variables • If we were determining the effect of sunshine on plant growth, then some measure of plant growth would represent the dependent variable. • The hypothesis in this case may have been something to the effect of, “Exposure to sunshine causes plants to grow.“ • Note that in order for this to be an effective experiment, the dependent variable must be measurable, the more precisely the better. • This need for measurability is what makes the "supernatural" off limits to science since the supernatural, by definition, is not a measurable quantity.

  33. Independent Variable • The independent variable is that measure that is being purposefully varied in the course of an experiment to test whether such variation results in a change in the dependent variables. • For example, if we were determining the effect of sunshine on plant growth, then the degree to which we exposed a plant to sunshine would represent the independent variable. • The dependent variable need not vary with the independent variable. • Such a failure to vary with the independent variable would be termed a negative experimental result.

  34. Using Independent Variables • An independent variable must vary over some range, preferably in a well-controlled or easily measured manner. • Note that it is possible to have more than one independent variable in an experiment, though using more than one independent variable typically increases the complexity of an experiment, or its size. • On other hand, you may be able to save time by essentially doing more than one experiment simultaneously. • Having more than one independent variable typically is used as a productivity trick for those who are well practiced at a given protocol.

  35. The requirement for control or measurement of independent variables limits the types of things that can be used as an independent variable. • The "supernatural" cannot serve as an independent variable by merely natural entities. • "Dinosaurs became extinct because a supernatural power was dissatisfied with their progress" therefore is not a very good hypothesis because, though we can measure the extinction of dinosaurs (the dependent variable), we cannot measure a supernatural power (the independent variable). • We can also reject this hypothesis simply on the grounds that the term "progress" is too ambiguous: • Just what is progress if the dinosaurs, in all of their glory, did not satisfactorily progress? Independent-Variable Limitations

  36. Treatment refers to the independent variable. • Recall that the independent variable may be something that is applied, but also may be something that is inherent to the system and therefore only measured. • When considering level of treatment, be prepared for "treatments" that involve no treatment on the part of the experimenter. • That is, “treatment” is being used in a sense that differs from one’s usual understanding of “treatment.” Similarly, scientists use the word “theory” in a sense that is different from the lay understanding of the word. • Level of treatment can be difficult to grasp due to the prejudices one brings to the table about the meaning of the term "treatment." • Level of treatment is simply the degree or manner in which the independent variable is varied. Level of Treatment

  37. Different Levels of Treatment • Differences in drugs (or a drug versus a placebo). • Differences in dates in the past (e.g., 50 million years ago versus 25 million years ago versus the present). • Nothing at all (e.g., 50 minutes of doing nothing versus 90 minutes of doing nothing). • Differences in what was sampled (e.g., an apple versus an orange). • Different populations, etc.

  38. Independent & Dependent Variables Dependent Variables include, e.g., # pods, # seeds/pod, pod weight What are the Levels of Treatment?

  39. Controlled Variables • The real world is a messy place. • Consequently, there typically exist many more variables within an experiment than just those we designate as the dependent variable and the independent variable(s). • If we would like to see our experiment work (i.e., give us unambiguous results), we had better make sure that the only variables that actually vary are the independent variable(s)… • …and should it vary with the independent variable(s), then the dependent variable as well. • These other variables that we attempt to keep from varying are called controlled variables.

  40. Controlled variables are potential independent variables that, by design, are not varied in the course of an experiment. • Note that it is not always easy to determine and control all of the variables we would like to be our controlled variables. • One of the key difficulties (and challenges) of experimental science is designing and performing experiments in such a way that we keep our controlled variables actually under our control. • "The underlying assumption in experimental design is that the selected independent variable is the one affecting the dependent variable. This is only true if all other variables are controlled." your lab text Controlling Controlled Variables

  41. Accounting for all Variables What are the Controlled Variables?

  42. The Protocol / Procedure • "The procedure is the stepwise method, or sequence of steps to be performed for the experiment. It should be recorded in a laboratory notebook before initiating the experiment, and any exceptions or modifications should be noted during the experiment. The procedures may be designed from research published in scientific journals, collaborations with colleagues in the lab or other institutions, or by means of one's own novel and creative ideas. The process of outlining the procedure includes determining control treatment(s), levels of treatments, and numbers of replications." your lab text

  43. The protocol is the means by which an experiment is run. • The protocol must include all relevant details such that the original experimenter or others can repeat the experiment some time in the future. • In practice, however, often much is left unsaid in a given protocol with much knowledge on the part of the reader assumed. • Consequently, two individuals performing the "same" protocol do not always succeed in repeating each other's results. • In reality there is an art to including all "relevant" detail in a less than fully comprehensive protocol. • Minimally, one should write down what one did in a manner that will allow one to repeat the protocol some time in the future. • The more comprehensive the protocol, the less knowledge assumed of the reader. How Much Detail is Enough?

  44. Replication • "Scientific investigations are not valid if they base their conclusions on one experiment with one or two individuals. Generally, the same procedure will be repeated several times (replication) providing consistent results. Notice that scientists do not expect exactly the same results inasmuch as individuals and their responses will vary. Results from replicated experiments are usually averaged and may be further analyzed using statistical tests." your lab text

More Related