Scientific thinking involves being critical or skeptical about claims made by people. Environmentalists, government officials, corporate spokespeople, news media reporters, advertisers, politicians, and just the average citizen often make statements about an environmental issue that prove later to be unfounded, untrue, or based on incomplete or wrong information. In this chapter, the authors attempt to describe the methods used by scientists to establish knowledge and give some examples of scientific, pseudoscientific, and non-scientific thinking. In order to illustrate how the public can be misled, consider the Case of the Mysterious Crop Circles (Case Study). The media and many people were fooled into believing that the designs that appeared overnight in agricultural fields in England were due to various forces, including: aliens, electromagnetic fields, and whirlwinds. After much investigation, it was concluded that the designs were made by people with boards who dragged them through the fields, created the intricate patterns, and carefully covered their tracks as they went. Scientists are often some of the most skeptical people in the world, often not believing reports and explanations in the media of incredible events like the crop circles. "Tabloid" newspapers and television shows do little to improve the credibility of the news media, which often report phenomena based on incomplete data or a few non-scientific sources. Scientists are skeptical people, because they rely on the scientific method to aid them in finding the truth. The scientific method uses an inductive logic tree, in which successive hypotheses are eliminated one at a time using experiments. Most people use everyday thinking to evaluate statements about the environment, which differs from the scientific method in that it is an informal method based on trial and error experiences, hearsay evidence, and common sense, not careful observation and controlled experimental data. Scientists are slow at reaching "the truth" and often state that there aren't enough data to reach a certain conclusion; what they mean is that they have not ruled out many alternative explanations for an event. In fact, "the truth" is impossible to know for certain, for science advances by disproof of hypotheses, and the disproved explanations are eliminated from being "the truth". Any remaining hypothesis or explanation is considered to "be true" until eliminated by an experimental test. The only exception to this rule is that a hypothesis or explanation, to be considered scientific, must also be falsifiable, that is the phenomenon or proposed explanation must be observable by people and the hypothesis must be phrased in a way that it is possible for an experiment to show it to be false. Because of the falsifiablilty rule, most explanations involving unobservable events (such as aliens or supernatural beings) are eliminated from consideration by scientists immediately.

What is science?

• Science is an organized human activity, based on observation of nature and the application of the scientific method, that has been devised as a way to assemble falsifiable statements known not to be false.
• Scientific statements have a high probability of being true, although they are not known to be true without a doubt.
• Science represents all of human knowledge and this is typically written down in scientific books and journals.

What is the difference between science and technology?

• Science is a search for understanding of the natural world, whereas technology is the control of the natural world for the benefits of humans.
• Science depends on new tools and inventions to improve the ability of scientists to make observations of nature.
• Technology often advances because of scientific discoveries that were made in pursuit of knowledge for its own sake. For example, scientists using the light microscope were limited previously to seeing things no smaller than 1/10 of a micrometer and magnifying them 1000 times. Now, with the invention of the electron microscope, they can see things that are much smaller, because the electron microscope can magnify objects 100,000 times. Because an understanding of electron beams and focusing them required scientific investigation, the technology that improved microscopy was dependent upon science.
• Thus, science and technology are interdependent human activities, and one promotes the other. They are often used interchangeably by people in discussing environmental issues; however, they are different but related areas of human activity.

What isn't science?

• Statements about the ultimate purpose of life, the existence of supernatural beings (including God), statements on human values, beauty, good and evil are not considered part of science.
• Scientists often have opinions on such topics, but that does not make their opinions scientific fact.
• Statements about supernatural beings and religious belief systems are very important to humans, they just are not part of science.

What is the scientific method?

• The scientific method consists of four steps:

What kind of reasoning is used in everyday life?

• Most people reason using common sense, which is more than adequate for daily decisions of what to buy, what to wear, what to do in a given situation.
• These decisions are guided by previous experiences and an expectation that a future event will be similar to what occurred in the past.
• However, this everyday knowledge and reasoning is tolerant of imprecision, is acquired by trial and error methods, is not validated by formal methods of testing, is often taken on faith or assumed based on what others say, (especially parents or experts) and is based on subjective values or beliefs.
• Most of the time, these limitations do not significantly influence the outcome of the decision, or people don't care if they do.

What are the assumptions that all scientists make?

• Events in nature follow patterns
• There are universal patterns or rules
• General conclusions can be drawn from specific events; this type of reasoning is called inductive logic.
• These general conclusions can be tested; tests can falsify or disprove the generalizations (or hypotheses).

What is scientific "proof" and how does it differ from "proof" in mathematics and in everyday life?

• Science cannot provide definitive proof; scientists can only make statements that have a high probability of being true.
• Science advances by disproving false hypotheses: any hypotheses left after testing are considered to be true for the time being.
• A proof in mathematics involves using deductive logic to show that given initial starting premises, a certain condition follows. This kind of proof is definitive, but it is not the same as scientific proof.
• When someone says that something has been proven scientifically, it is better to think of this as a statement that is likely to be true 95 - 99% of the time.

What is a hypothesis?

• A hypothesis is a statement (not a question) that might explain some event or natural phenomenon.
• Usually, hypotheses propose a cause for the event, in a conditional sense: "If it is warm, clams will grow faster than if it is cold".

What is a theory?

• A theory is a collection of unfalsified hypotheses.
• A theory thus is a collection of statements with a high probability of being true, which are put together as a narrative (or a story).
• Theories are never proven in science, but they can be disproved if certain key hypotheses that are central to the theory can be disproved.

What is an experiment?

• An experiment is a controlled comparison study among experimental groups in which all factors that could influence the outcome variables are the same among groups being compared except for one or two factors that are intentionally varied.
• For example, in an experiment to test the hypothesis that growth rate of a certain species of clam is faster at high water temperatures than at low temperatures, an aquaculturist could intentionally vary water temperature and the growth rate of clams could be monitored among groups of clams exposed to different temperatures.
• For this to be a controlled experiment, all other factors that could affect growth (food availability, sunlight, dissolved oxygen in the water, salt content or salinity of the water, substrate type, etc.) must be the same across the groups of clams; only temperature can be different among groups.
• Experiments may be done in both the laboratory and the field, but normally it is easier to control temperature and salinity in the laboratory.
• Field experiments are often more variable, but are required to demonstrate that the relationships among variables discovered in laboratory studies are valid in the natural world.

What is the difference between inductive and deductive reasoning?

• Inductive reasoning is used when one makes many observations and draws a general conclusion from them, usually a statement in the form of a hypothesis to be tested in an experiment. This is the kind of logic that scientists most often use to make observations and formulate hypotheses.
• Deductive reasoning is a form of logic in which general premises or statements are taken to be true and specific conclusions can be made based on the premises. Deductive logic does not require that the premises are true, but rather that the logic is sound in reaching a conclusion.
• An example of each type:

Aren't all scientific measurements made without error?

• No. All scientists realize that every measurement is an approximation.
• All measurements in science need to have an error or uncertainty value associated with the measurement.
• Some measurements are more precise than others, but they are all estimates of the true value.
• One source of potential error is called measurement error, in which the scientist's ability to measure is limited by the tool used in measuring, the scientist's eye, or other senses used to make the measurement.
• The true value will always be unknown, because of measurement error and other random errors. The true value is approached as the sample size (or number of measurements) increases. As sample size increases, the associated random error gets smaller.
• Example: The temperature at which the Space Shuttle O-rings fail is -1 ^oC (30 ^oF) + 1 ^oC or colder. If the temperature at launch time is 0 ^oC (32 ^oF), then the shuttle should not be launched. Even though 0 ^oC is above -1 ^oC, the temperature is within the error range in which it will fail, according to the engineers that built the O-rings.

What is the difference between accuracy and precision?

• Accuracy is the degree to which a measurement approaches the true or known value. This can be assessed for measurements in which we have a standard true value that is agreed upon by all scientists. For example, seawater has an agreed upon standard value of 35.0000 parts per thousand salinity or salt content. At sealevel, the boiling point of water is 100.000 ^oC (212.000 ^oF). If you measure the temperature of water boiling and your thermometer reads 99.000 ^oC (210.200 ^oF), then your thermometer is inaccurate.
• Precision is the degree of exactness with which a measurement is made. If your thermometer is marked in 0.1 ^oC (0.2 ^oF) increments, then it is less precise than another thermometer which reads in 0.001 ^oC (0.002 ^oF) increments.

• This is the error associated with running an experiment (see what is an experiment?).
• The experimental error is normally estimated by measuring the disagreement within a group treated the same way in an experimental treatment (statistically, this is done by calculating its variance).
• Thus, in the example above, all clams in a temperature treatment group should grow at exactly the same rate, but they rarely do in actual experiments. The reason is that there is error in measuring the growth rate (measurement error associated with measuring the size or weight of the clams) and there is random or unassigned error (the temperature varies ever so slightly for each clam, the clams are all unique individuals with their own metabolic rates, etc.)
• The experimental error, which is the combined measurement and random errors, can be seen in the fact that all the clams will be measured growing at a slightly different rate within a treatment, but the variation within the treatment group will be less than the differences in growth rate between groups of clams grown at different temperatures. So, experimental error gives the variation among individual clam responses as well as the error associated with the temperature effect.

• A dependent variable is the response variable or the measurement that will most likely respond to an experimental treatment. In our clam growth experiment, this would be the growth rate measured as g of body mass (excluding the shell) added/day.
• The independent variable is the variable in an experiment that is intentionally varied. In the clam example, the independent variable is temperature.

What is meant by an operational definition of a variable?

• When scientists are describing their methods for conducting an experiment, they must produce a list of the variables that were measured and a definition for how the measured variables were obtained. These are the operational definitions.
• In our clam example, the way that growth was measured (g of shell-less body mass added/day) must be specified: the clams were grown at each of the temperatures for 14 days, and the biomass increase for each treatment group was measured by weighing the shell-less biomass of the clam, then subtracting the average shell-less biomass of a random sample of 10 clams in each group taken at the start of the study, and finally dividing by 14. This is an example of a quantitative variable.
• A qualitative response variable would be to use categories such as small, medium, or large to group the clams, and count the number of clams that fell into each group for each temperature treatment.

• Probability is the number of successful trials divided by the number of attempts. Success is defined as the outcome of interest.
• For example, we can estimate that 50 % of the clams in our experiment will die if the temperature gets close to 30 ^oC (this is called the thermal maximum, and now that we know of its existence , we would be forced to reevaluate or reword our hypothesis stated above. How might it be changed?). We could define successful outcomes as those in which the clams survive.
• Thus, the probability of a clam surviving would increase as we decreased the temperature below 30 ^oC, but the probability would decrease as we approached 30 ^oC. If 30 ^oC is exceeded, all the clams die (0.00 probability of survival):

What is a computer model?

• Find article in your local newspaper, or a magazine, that discusses a scientific study of an environmental issue.
• What hypothesis is being tested by the scientists?
• Did the scientists eliminate or disprove any of the alternative explanations for the discovery?
• Can you think of any alternative explanations?
• If no hypothesis is stated in the article, write a hypothesis and a series of alternative hypotheses down. How could you test each alternative?
• Please respond to these questions or send your thoughtful examples and comments to:

• Gleick, J. 1987. Chaos: Making of a New Science. Penguin Books, New York. 352 pp. A book about the developments in Chaos theory, an emerging field of study in physics, chemistry, biology, and earth sciences. This book explains why computer models can never be perfect predictors of what will happen in nature.
• Feynman, R., R. B. Leighton, M. Sands. 1963. The Feynman Lectures on Physics. Volume I. Addison-Wesley, Reading, MA. See Pages 2-1 - 2-2.
• Ideas about physics, nature, observation, description, and experiments using the scientific method.
• Richard Feynman was a physicist who lectured extensively about science and the scientific method.
• As an example of what scientists do, Feynman used the following scenario, imagine that you have never played chess before and you are watching a match being played (this could be any game you are familiar with). How could you figure out the rules from simply observing the players moves? When you have watched enough games and made careful notes, you probably would have a good idea of what each piece can do. At some point, someone will "castle" - a move that you could never have anticipated from previous moves of the king and the rook - and this will puzzle you. In a similar way, scientists must find patterns in nature and explain unanticipated results. If a scientist is fortunate, he or she will perform an experimental manipulation - move a chess piece - and predict the result for each piece. If it is as anticipated, great! If a castle follows - well, that's science! You need to make more observations, make more hypotheses, and conduct more experiments.
• See also the BBC Videotape: "The Pleasure of Finding Things Out" about Feynman
• Harte, J. 1985. Consider A Spherical Cow. William Kaufmann, Inc. Los Altos, California

• Note: If any of these links are not working, please see if alternative links are available at the Ecolink Update Site.

1. What caused the mysterious crop circles in corn fields in England?

2. Scientific statements must:

3. Scientific statements always have a high ___________________.

4. ____________________ is the control of the natural world for the benefit of humans.

5. Based on the data from a laboratory clam growth experiment given in the table below, which of the following is a non-scientific statement?

6. Based on the above clam growth data, at what temperature would you expect a clam farmer to grow clams to market size the fastest?

7. At which temperature is the experimental error the smallest?

8. In the above experiment, the __________________ is + 0.1 g/day for clam growth rate.

9. Which of these is the independent variable in the clam growth experiment described above?

10. Calculate the probability of a clam surviving when the water temperature is 29 ^oC, if you were given the following data: 50 clams were held at 29 ^oC and 28 of them survived.