ENVR 1401 – SCIENTIFIC METHOD BACKGROUND
Revised: December, 2016 1
NAME:
ENVR 1401 – SCIENTIFIC METHOD BACKGROUND
Overview This activity explores the concept and application of the scientific method. In order to distinguish true cause-effect relationships from associations or perceptions, phenomena must be investigated using designed experiments and careful observations that can be repeated by others.
The scientific method is typically discussed as a standardized, linear process that includes the following steps and involves specific skills:
1. Make observations or gather data often leads to a question
2. Formulate an hypothesis which leads to an associated prediction
3. Design an appropriate test/experiment to assess the hypothesis/prediction
4. Conduct test/experiment, record and analyze the results (including mathematical and statistical evaluation)
5. Interpret the results and draw conclusions accept, revise or reject the hypothesis
6. Reporting the results (e.g., laboratory report, formal memorandum, peer reviewed article)
But, as shown in the flow chart, the process is not linear. It is cyclical because good science stimulates further thought and mandates that ideas be challenged and further tested to demonstrate that the results can be repeated (iterative process). A scientific hypothesis is an informed, testable and predictive explanation of a natural phenomenon, process or event. If, upon testing, the scientific hypothesis fails the test, it must be rejected or may be modified
Revised: December, 2016 2
and subject to further testing. Models are mathematical or conceptual hypotheses that provide useful perspectives, though often limited by oversimplification of the process they represent.
If, however, a scientific hypothesis continues to pass repeated tests and the predictions have been verified, then it is considered a corroborated hypothesis. A highly corroborated hypothesis which has been repeatedly tested and is supported by significant reliable evidence is considered a scientific fact or natural law, such as the existence of gravity as a property of all matter.
A unifying and consistent explanation of fundamental natural processes or phenomena that is constructed of corroborated hypotheses and scientific facts is a scientific theory. Scientific theories, such as quantum mechanics, thermodynamics, plate tectonics, evolution, or relativity, are the most reliable and comprehensive form of human knowledge. And, as we gain more knowledge through the application of the scientific method, our understanding of the universe in which we live and our theories on how it functions and evolves must continue to be refined.
Further value of the scientific method is derived from honing and applying necessary skills to develop scientific knowledge:
making, recording and reporting unbiased measurements classifying data translating and analyzing information applying deductive and inductive logic critical, interpretive and creative thinking identifying and controlling variables
The scientific method established an approach to a problem and enforces scientific thought that attempts to eliminate bias in the resulting data and conclusions. Science relies upon empirical evidence which is observable and measurable by more than one researcher. But as humans, we tend to view the world and solutions to problems within our personal framework. For example, if a town along the river is repeatedly flooded, then the structural engineer believes the problem will be solved by building a dam and the politician may believe that any solution is too expensive or unpopular and would be detrimental to being re-elected. In contrast, the natural resource manager may prefer limiting urban development and re-establishing native habitat. The propensity of humans to perceive the world from their perspective is good reason to have a healthy dose of skepticism — to constantly question your beliefs, observations and conclusions.
Scientific methodology has a long history that dates back over 1,000 years, with many cultures and individuals contributing to its development. Ancient Egyptian papyri describe methods of medical diagnosis and empiricism. Empirical evidence found in nature allows us to describe and explain natural processes and natural laws. The first experimental scientific method was developed by Muslim scientists. In particular, Alhazen (Ibn al-Haytham, 965-1039) is credited with introducing experimentation and quantification in his work on optics, among his significant contributions to astronomy, engineering, physics, medicine, and science in general. In Europe the Renaissance resulted in renewed interest in the ideas and science developed during the Greek and Roman empires.
The logic and philosophical approaches of Aristotle and Socrates were improved upon by Francis Bacon in the early 1600s. Descartes formalized the guiding principles for the scientific method, strengthening the link between science and mathematics. Galileo also showed the importance of testing or experimenting to look for the opposite of a consequence to potentially disprove an idea. In the late 1800s, Peirce outlined objective methods using deduction and induction as complementary approaches, as well as outlining the basic scheme for hypothesis and testing that we currently use.
Revised: December, 2016 3
Deduction and induction are types of argument or logical approaches. Deduction, in simplest terms, is the logical process of arriving at a conclusion based on premises (e.g., facts, statements, laws…) that have been verified. Induction, in simplest terms, is the logical process of arriving at a conclusion based on premises that are assumed to be true; therefore, some conclusions of inductive thought processes may be false.
Deduction is generally described as moving from the more general to the more specific. For example, the Law of Gravity expresses the force that attracts objects to each other. In common terms, it explains why objects fall toward Earth or, in the environment, why water flows down hill. Based on the Law of Gravity we would deduce that weathered soil materials would also move down slope.
Induction is generally described as moving from specific observations or details to a general statement. For example, we begin by making observations or measurements, detect patterns or regularities, and develop a general conclusion. In application, if you touch a stove ten times and each time you touch the stove, you burn your hand. You might conclude that the stove is always hot. But this conclusion may or may not be true.
In reality, the scientific method can never “absolutely prove” or provide “truth” to understand. To paraphrase Einstein, “No amount of experimentation can ever prove me right; a single experiment can prove me wrong.” For this reason, and to overcome the human bias for seeing what we want to see, we need to consider that a cause-effect relationship exists (hypothesis) AND also consider that a perceived cause-effect relationship does not exist (null hypothesis).
Step 1 – Make observations or gather data Awareness of our environment may lead to posing a question. Such awareness may result from making observations around us, gathering preliminary data that reflect environmental conditions, or by researching and reading published work of others. This approach often leads to the recognition of a broad problem that warrants further investigation. For example, “Is our water reservoir clean and safe?” Exploring this question by further reading and research may allow refinement of the question. For example, “Is chemical “A” in our water supply reservoir?” The question will provide a basis for further investigation.
Step 2 – Formulate an hypothesis which leads to an associated prediction To formulate an hypothesis, it is important to focus the posed question and to define a specific parameter to be investigated with an expected result. The prediction may be 1) there will be a specific outcome in the experiment, 2) there will be a statistical difference between the tested subject and a control, or 3) there will be no difference between the tested subject and the control. Formulation of the hypothesis is critical because it will help to define and outline the experiment in terms of the specific parameters (independent and dependent variables) that are being assessed.
The hypothesis is a statement of the most likely outcome of the experiment. More appropriately, the hypothesis (designated by HA) should be viewed as a prediction that can be tested, is not ambiguous, and is dichotomous (a “yes” or “no” statement).
HA: Less than 5 mg/l of chemical “A” in the reservoir water will be lethal to bluegill minnows.
But remember that the prediction could be erroneous. Therefore, a null hypothesis (designated by H0) must also be developed:
H0: There is no relationship between less than 5 mg/l of chemical “A” in the reservoir water and lethality to bluegill minnows.
Revised: December, 2016 4
Range of tolerance — In environmental science, we often investigate the relationship between organisms and the abiotic (non-living or physical) characteristics of the environment, such as temperature, moisture, nutrient availability, and nutrient toxicity. Most organisms thrive under optimum environmental conditions. There is a small percentage of each species that can survive under less than ideal conditions despite being physiologically stressed. Such organisms will more readily adapt to changing environmental conditions and thus ensure the survival of the species. Often referred to as the Law of Tolerance, this relationship between population size and environmental condition is reflected in the following figure. This curve may also reflect the relationship between, for example, numbers of predators and the prey population.
Idealized Range of Tolerance Curve
Step 3 – Design an appropriate test/experiment to assess the hypothesis/prediction A well-designed experiment needs to have an independent variable and a dependent variable. The independent variable is what the scientist manipulates in the experiment. The dependent variable responds to the manipulation of the independent variable. Therefore, the dependent variable provides the data for the experiment. Said another way, the independent variable causes a response that is measured as the dependent variable….or…the independent variable may be considered as an action that results in a reaction (dependent variable). Consider, for example, a fire alarm in a building. It is not until the alarm rings that people will quickly evacuate the building. Therefore, it is the ringing of the alarm this is the independent variable that causes the dependent response of people quickly evacuating the building.
Variables that are held constant are called controlled variables. For example, if we wanted to test the effects of varying the level of dissolved oxygen on the survival of fish, then we would maintain other environmental factors (such as temperature, light, availability of food) constant to ensure that they were not affecting whether or not the fish survived.
A well-designed experiment should distinguish between the treatment (the experimental condition) and the control (reference for comparison). All variables are held constant for the control. For the treatment, only the independent variable is changed to determine the consequent outcome (dependent variable). The control is a source of reference and, since no variables are manipulated for the control, then no response or change should be noted. If the control remains constant, then changes that result from the manipulation of the independent variable for the treatment are attributed to the experimental factor.
Revised: December, 2016 5
If, however, changes are noted in the control, then such changes are attributed to one or more confounding variables. The difficulty with confounding variables is that they are not readily identified before or during the study and often leading to misinterpreted or incorrect results.
Step 4 – Conduct test/experiment, record and analyze the results The data may be qualitative (a verbal description of observed outcomes of the experiment) or quantitative (involving the collection of numerical values that can be mathematically assessed or applied to a statistical model). Observation must be thorough and impartial. But, be cautious to present your data in a factual manner. Your interpretations, opinions and conclusions are NOT part of presenting and analyzing the results – they properly belong to Step 5.
Instruments used to collect data should be properly calibrated. Testing equipment and methods should be consistent throughout the investigation. Multiple tests should be conducted on each test parameter, and the existence of outliers (unusual values) should be noted. When outliers are observed, the testing methods and equipment should be checked carefully for malfunctions and recording errors. Finally, all observations must be noted in permanent written documentation so that it can be evaluated at a later time and by others.
Sources of Error – To assess the validity of scientific work, sources of error should be identified and evaluated. Common sources of error may result from measurements or from the sample that is used in the experimentation/testing process.
Precision is a measure of the scatter, dispersion, or ability to replicate the measurements. Low- precision (high-scatter) measurements are referred to as noisy data. Smaller average difference between repeat (replicate) measurements means higher precision. For example, if a sheet of paper is measured several times with a ruler, we might get measurements such as 10.9, 11.0, 10.9, and 11.1 inches. If a micrometer is used instead, we might get measurements such as 10.97, 10.96, 10.98, and 10.97 inches. These estimates show random variation regardless of the measuring device, but the micrometer gives a more precise measurement than does the ruler. If the ruler or micrometer is poorly made, it may yield measurements that are consistently offset, or systematically biased, from the true lengths. Accuracy is the extent to which the measurements are a reliable estimate of the ‘true’ value. Both random errors and systematic biases reduce accuracy. To reduce errors of measurement, a large number of measurements may be obtained and then averaged to attempt to minimize bias and achieve accuracy.
A representative sample is a small subset of the overall population, exhibiting the same characteristics as that overall population. It is also a prerequisite to valid statistical induction, or quantitative generalization. Representative sampling is essential for successful averaging of random errors and avoidance of systematic errors, or bias. With random sampling, every specimen of the population should have an equal chance of being included in the sample. There are standardized techniques for conducting random sampling to obtain a representative sample. Sometimes, however, random sampling is not feasible and the results may therefore not be consistent with the overall population and be free of bias.
Presenting and Analyzing Results — Modern science almost always employs a statistical analysis to interpret data. Reference the Mathematics and Environmental Science Background document from Lab 1 for guidance on selecting appropriate graphic method and mathematical analysis of the data. These interpretations are quantitative in nature and allow the scientist to determine whether the experimental results indicate a consistent trend or condition.
Interpretations may also be qualitative in nature. For instance, an hypothesis predicted that there was fecal coliform in a local stream and, after repeated testing, none was found. This could be
Revised: December, 2016 6
qualitatively interpreted that there is an insignificant amount of fecal coliform in the stream. This type of qualitative interpretation leaves room for errors in judgment.
Step 5 – Interpret the results and draw conclusions accept, revise or reject the hypothesis When statistical evaluation of experimental results is conducted, there will either be a statistical difference between the tested subject and a control group (supports hypothesis HA) or there will be no difference between the tested subject and the control group (supports null hypothesis H0). Lacking statistical evaluation requires that the mathematical analysis and/or qualitative outcomes be interpreted to determine whether the hypothesis is accepted, modified or rejected.
For example, the determination of an insignificant amount of fecal coliform in a stream could be interpreted as the water is safe to drink. But would you feel safe drinking the stream water?
• Hypothesis Accepted – If the hypothesis is accepted, the student will write a formal report and contribute this work to the existing scientific literature. The report will make it possible for other scientists (in this case, other students or the instructor) to evaluate work for scientific soundness and to repeat the experiment to verify the results. This process is known as peer review.
• Hypothesis Rejected – If the hypothesis is rejected, the student will write a formal report and contribute this work to the existing scientific literature. Significant knowledge can be gained from ideas that fail. This makes it possible for the original researcher or other students to investigate the same topic and to build upon the rejected hypothesis from the former investigation. Edison, for example, did not invent the light bulb. Over a year and a half, Edison significantly improved upon a fifty-year old idea. In creating an electric lighting system that contained all the elements necessary to make the incandescent light practical, safe, and economical for home use, Edison found 10,000 ways that would not work.
Step 6 – Reporting the results Laboratory Report – Details and guidance on writing a formal laboratory report are outlined below.
Introduction: The introduction establishes the purpose of the report. A well written introduction should answer the questions, why are we doing this research and what do we hope to find out? You also need to include your hypotheses in this section.
Methods: The methods section should be a straightforward description of what you did in your experiment. This includes describing the materials you used and how you used them. Based on your methodology section, the reader should be able to replicate your experiment.
Results: The results are a summary of your observations. You need to report on all of your results, do not leave anything out. Avoid discussing what your results may mean in this section. That information belongs in the discussion/conclusions section.
Discussion and Conclusions: Based on your results, did your outcomes support your original hypothesis? Keep in mind that science never truly proves anything; it contributes to a growing body of evidence about the way our world works. If your research does not appear to support your hypothesis, be truthful about that. Remember that in science we learn as much from failures as we do from successes. You should also address your possible sources of error in your experiment. If you were to repeat this experiment, what would you do differently and why?
- Overview
- Step 1 – Make observations or gather data
- Step 2 – Formulate an hypothesis which leads to an associated prediction
- Idealized Range of Tolerance Curve
- Step 4 – Conduct test/experiment, record and analyze the results
- Step 5 – Interpret the results and draw conclusions accept, revise or reject the hypothesis
- Step 6 – Reporting the results
