Scientific Method Lecture.pdf

Scientific Method “To do science is to search for repeated patterns, not simply to accumulate facts.” - R.H. MacArthur (1930 - 1972) “Scientific knowledge is a body of statements of varying degrees of certainty -- some most unsure, some nearly sure, none abso...

Scientific Method “To do science is to search for repeated patterns, not simply to accumulate facts.” - R.H. MacArthur (1930 - 1972) “Scientific knowledge is a body of statements of varying degrees of certainty -- some most unsure, some nearly sure, none absolutely certain.” - R.P. Feynman (1918 - 1988) Science as a Way of Knowing How do you know what you know? Science is a system for trying to help people figure out what they know, to discover new knowledge, and to evaluate how confident they can be about their knowledge. Often described as “a system of thought that is based on reasoning, logic and evidence”.  Scientific Method The Scientific Method The “scientific method” is used to try to figure out how things work; to gain new knowledge/insight. There are many descriptions of the scientific method, but most are variations on similar themes. There are also philosophers of science that suggest there is no universal scientific method, and others that suggest that the existence of a scientific method is an idealization that few scientists actually follow. There is a well-known book: “Against Method”; -take home message of the book is often summarized as “anything goes” -but the “anything goes” summary is considered to be an oversimplification But even people who disagree that there is a scientific method do agree that reasoning, logic and evidence are extremely important.  Scientific Method Philosophers of Science often Describe Two Types of Science: 1) induction-based science (also called discovery science) 2) hypothesis-based science Induction (inductive reasoning) Predictive generalizations that are based on a large number of observations. Classic example: we predict that the sun will rise in the east tomorrow morning, based on past experiences of the sun doing just that. Sunrise over Wascana Lake. https://farm2.staticflickr.com/1612/24647710976_b11013fe98_o.jpg  Scientific Method Induction-Based Science (= Discovery Science) Makes predictions/generalizations based on past experience. An example of induction-based science is the development of “Cell Theory”; -a series of generalizations about cells based on observing many cells (Cell Theory is covered in more detail in the Cells portion of the course). One aspect of Cell Theory: All organisms are composed of one or more cells. No one has ever demonstrated the existence of an organism that is not composed of cells. Cell Theory came about because a lot of different scientists working in different places at different times were making the same observations.  Scientific Method Hypothesis-Based Science Hypothesis-based science gets most of the attention from philosophers of science. Is the basis for most formulations of the “scientific method”. Hypothesis – A tentative explanation for an observation; an educated guess at an explanation; sometimes referred to as a “conjecture”. Philosophers of science debate whether inductive or deductive reasoning is used to generate hypotheses; -I tend to favour the idea that inductive reasoning is used -but I’m not a philosopher Deduction (deductive reasoning) Reasoning from more general statements to a conclusion that must be true. A deduction is a “necessary inference”; e.g. 1) Jill was at a Saskatchewan Roughriders football game on Sept. 9 at 7:00 pm. 2) Jack was at a Saskatchewan Roughriders football game on Sept. 9 at 7:00 pm. Therefore, Jill and Jack were at the same football game (this must be true).  Scientific Method Hypothetico-Deductive Method = One label/description of the method of hypothesis-based science. Involves hypothesis testing; hypotheses are tested by the use of experiments. The outcome of the experiments is predictable based on the hypothesis, e.g. if the hypothesis is correct, then the outcome of the experiment must be predictable. This is where the deductive reasoning definitely comes in: “If the hypothesis is correct, then the results of the experiment must be......” You are deducing the experimental outcomes, based on the hypothesis. If the outcomes of the experiments turn out to be inconsistent with the predictions, then the hypothesis is rejected (it is wrong, refuted). Alternatively, if the outcomes of the experiments are consistent with the predictions for the outcomes, then we have support for the hypothesis (but this ≠ “proof”). There is a major focus on hypothesis testing in the hypothetico-deductive method.  Scientific Method The Scientific Method Deals with Testable Hypotheses Testable hypothesis - Hypothesis which can provide testable predictions for outcomes of experiments. A testable hypothesis has the potential for being shown to be wrong/incorrect (i.e. it can be rejected/refuted/falsified). This is the criterion of falsifiability – Can the hypothesis be demonstrated to be incorrect? Non-Testable Hypothesis A hypothesis which cannot provide testable predictions. It cannot be shown to be incorrect. Such a hypothesis is not necessarily incorrect, but the scientific method simply cannot deal with it (might be wrong, might be right; who knows?). Science is good at disproving things – science does not really “prove” things (but can provide support for ideas/hypotheses).  Scientific Method – An Algal Example You have been investigating algal growth, and you’ve noted that some algal species can exist as either unicells or clumps of cells, i.e. the algae are “unicellular” or “clumped”. Your hypothesis (tentative explanation) is that “unicellular” is the standard state, and that “clumped” is a defence response against crustacean grazers (there are a lot of aquatic crustaceans that eat algae). Design an experiment in which you add crustacean grazers (Daphnia) to an algal culture. Lürling M (1993) Annals of Limnology Daphnia magna. 31: 85-101. Daphnia Dieter Ebert, Basel, Switzerland, Creative Commons Attribution-Share Alike 4.0 International Prediction about the outcome of the experiment: Addition of a crustacean grazer (Daphnia in this image) should cause transition from “unicellular” to “clumped” algal cells.  Scientific Method – An Algal Example Based on this hypothesis, you’ve designed an experiment whose outcome is predictable if the hypothesis is correct. A simple experiment would be to have two beakers of algae growing under identical conditions. One beaker is the “control” (nothing done to algae in this beaker), and the other beaker would have crustacean grazers added to it (this is the “treatment” beaker). You predict, based on your hypothesis, that the treatment beaker would show a shift from “unicellular” to “clumped”: two beakers with algal suspensions Experimental Design Two beakers containing the same volume of algal suspension. Control Treatment (nothing added; (crustaceans added) a negative control)  Scientific Method – An Algal Example So now you’ve run the experiment, and tabulated the results. One possible scenario is that addition of grazers had no effect on the algal culture (no clumping was observed); → your hypothesis was incorrect (or “refuted”), i.e. you were wrong and it’s time to come up with a new testable hypothesis Alternatively, perhaps the algal culture with added grazers did become clumped. This provides support for your hypothesis but does not prove your hypothesis (because there might always be a different explanation). “No amount of experimentation can ever prove me right; a single experiment can prove me wrong.” - Albert Einstein (1879 – 1955)  Scientific Method – An Algal Example Assume that your prediction was correct. What’s next? The next step is to conduct further experiments, for example, what if the grazers themselves are not directly causing the clumping, but rather grazer-caused turbulence in the culture is leading to clumping? To test that latter hypothesis, you could compare a stirred algal culture (treatment) with a non-stirred culture (control), with the prediction that only the stirred culture would show clumping. two beakers with algal suspensions If the stirred culture did not show clumping, that would lend more support to the original hypothesis that it’s the grazers that induce clumping. Control Treatment (nothing added; (stir bar added) a negative control)  Scientific Method – An Algal Example But it’s also important to keep an eye on the “control”, and there always must be a control. In the absence of a control, there is no way to judge the validity of your experimental results. For example, if in the control beaker of algae there is a transition from unicellular to clumped (i.e. in the absence of grazers), that means that the transition from unicellular to clumped in the presence of grazers does not actually provide support for your hypothesis (i.e. the transition was likely independent of grazers). But without a control beaker, you would never know that. H. Weger The green alga Desmodesmus sp. This strain was isolated from Pasqua Lake in Saskatchewan.  Scientific Method – Controls Experiments always need controls. Two types of controls: 1) negative controls 2) positive controls Negative controls are always part of experimental designs. In the algae example, the negative control was the beaker to which nothing was done; it simply sat there. If the negative control ends up acting like the treatment → you cannot say that the treatment caused the observed effect In other words, if the treatment is not different from the control, you cannot conclude that the treatment had an effect. A negative control Control Treatment (nothing added; (crustaceans added) a negative control)  Scientific Method – Controls Positive controls are often, but not always, part of experimental designs. A positive control is a known/characterized treatment (often a standard treatment used in the sub-discipline) that is known to produce an effect similar/identical to what you predict your experimental treatment will produce. E.g. you are testing a new type of plant fertilizer, and you want to see if the new fertilizer is as good as the existing fertilizer: positive control – existing fertilizer “treatment” – new fertilizer negative control – no fertilizer  Scientific Method – Controls Positive controls are typically very important in chemical analyses. Example on the right: A simple “colorimetric assay” https://www.pharmaguideline.com/ that detects the presence of a certain molecule (analyte); in this example, the presence of the analyte causes the solution to turn red. These assays should have both a negative and a positive control. The positive control is the addition of a known amount of the analyte in question. If the positive control does not turn red, then you have a problem with your assay → the lack of redness in the sample means nothing If the negative control turns red, that’s also a problem.  Hypothetico-Deductive Method Flow Chart Hypothesis (= tentative explanation) Design experiments - Outcome is predictable if the hypothesis is correct Run the experiments, tabulate the results Outcome not as predicted Outcome as predicted Hypothesis is rejected (you were wrong) Support for the hypothesis ≠ “proves” the hypothesis Need a new hypothesis More experiments Design experiments to test the new hypothesis  Scientific Method – Scientific “Progress” “Progress" in science is often greatest because of rejected/falsified hypotheses. This is especially true if there has been a particular explanation for a certain process that has been widely accepted because of much supporting evidence. New experimental results that are inconsistent with the widely accepted view produce a different or more refined model. This is how a lot of progress takes place; -old ideas are rejected, and new ones formulated -this is why falsified/rejected hypotheses are so important -rejected hypotheses lead to new ideas/hypotheses, which are then tested Sometimes one hears about hypotheses being “powerful”; -means that have a great deal of “explanatory power” -the greater the amount of explanatory power, the more important the hypothesis -if the hypothesis suggests a model for a general process, rather than one specific example of a process, it has great explanatory power  Scientific Method – Scientific “Theories” Hypotheses that attempt to explain a large variety of phenomena, i.e. are of large scope and are generally applicable, are known as “theories”. Theories have a great deal of explanatory power. Because theories attempt to explain so much, they only become accepted if supported by a large body of evidence; -e.g. Einstein's theory of relativity, and the theory of natural selection -natural selection is widely accepted because of the accumulation of a very large body of evidence in support of it Theories may be superseded by other theories E.g. Newton's theory about the relationship between the motion of a body and the causes for this motion (study of dynamics) was superseded by Einstein's theory. Einstein's theory could explain everything that Newton's theory could explain, and also explained things that Newton's theory couldn't handle e.g. motions of objects moving close to the speed of light; → greater explanatory power  Scientific Method – Replication & Reproducibility Replication has various meanings/contexts; one meaning means that sample size ≠ 1. Algal example: if this is the complete experiment (one control beaker and one treatment beaker; → no one will be convinced that crustacean grazers lead to clumping (they will say “yeah, maybe”) Greater chance of convincing someone of the validity of your results if you had “replicates.” E.g. five (5) control beakers and five (5) treatment beakers; → did all of the control beakers remain unicellular, and did all of the treatment beakers transition to clumped?  Scientific Method – Replication & Reproducibility Reproducibility refers to the idea that if an experiment is repeated, the results should be the same. If you repeat the experiment, do you see the same results every time that you do the experiment? If someone on the other side of the world does the exact same experiment, do they see the same results that you did? This is a key point with respect to science: the same results should be observed for the same experiment, regardless of where or when the experiment is conducted. The above is a very superficial discussion of replication and reproducibility; there is a lot of thought and writing that has gone into these two topics.  Scientific Method William of Occam (c. 1287-1347). Occam’s (Ockham’s) Razor Centuries-old principle of logic. If several explanations are compatible with the evidence at hand, the simplest should be considered the most likely. Alternative version: explanations should be no more complicated than necessary. Named after William of Occam (who was not the inventor of the idea, but who popularized it). The razor “shaves off” unnecessary details. Ockham - from a manuscipt of Ockham's Summa Logicae, MS Gonville and Caius College, Cambridge, Another quote from Einstein: 464/571, fol. 69. Public Domain “It can scarcely be denied that the supreme goal of all theory is to make the irreducible basic elements as simple and as few as possible without having to surrender the adequate representation of a single datum of experience.” usually paraphrased as: “Everything should be made as simple as possible, but no simpler.”