Observation and Experimentation
A scientist begins an investigation by observing an object or an activity. Observation typically involves one or more of the human senses—hearing, sight, smell, taste, and touch. Scientists typically use tools to aid in their observations. For example, a microscope helps view objects too small to be seen with the unaided human eye, while a telescope views objects too far away to be seen by the unaided eye.
Scientists typically apply their observation skills to an experiment. An experiment is any kind of trial that enables scientists to control and change at will the conditions under which events occur. It can be something extremely simple, such as heating a solid to see when it melts, or something highly complex, such as bouncing a radio signal off the surface of a distant planet. Scientists typically repeat experiments, sometimes many times, in order to be sure that the results were not affected by unforeseen factors.
Most experiments involve real objects in the physical world, such as electric circuits, chemical compounds, or living organisms. However, with the rapid progress in electronics, computer simulations can now carry out some experiments instead. If they are carefully constructed, these simulations or models can accurately predict how real objects will behave.
One advantage of a simulation is that it allows experiments to be conducted without any risks. Another is that it can alter the apparent passage of time, speeding up or slowing down natural processes. This enables scientists to investigate things that happen very gradually, such as evolution in simple organisms, or ones that happen almost instantaneously, such as collisions or explosions.
Data Collection and Analysis
During an experiment, scientists typically make measurements and collect results as they work. This information, known as data, can take many forms. Data may be a set of numbers, such as daily measurements of the temperature in a particular location or a description of side effects in an animal that has been given an experimental drug. Scientists typically use computers to arrange data in ways that make the information easier to understand and analyze. Data may be arranged into a diagram such as a graph that shows how one quantity (body temperature, for instance) varies in relation to another quantity (days since starting a drug treatment). A scientist flying in a helicopter may collect information about the location of a migrating herd of elephants in Africa during different seasons of a year. The data collected may be in the form of geographic coordinates that can be plotted on a map to provide the position of the elephant herd at any given time during a year.
Scientists use mathematics to analyze the data and help them interpret their results. The types of mathematics used include statistics, which is the analysis of numerical data, and probability, which calculates the likelihood that any particular event will occur.
Formulating a Hypothesis
Once an experiment has been carried out and data collected and analyzed, scientists look for whatever pattern their results produce and try to formulate a hypothesis that explains all the facts observed in an experiment. In developing a hypothesis, scientists employ methods of induction to generalize from the experiment’s results to predict future outcomes, and deduction to infer new facts from experimental results.
Formulating a hypothesis may be difficult for scientists because there may not be enough information provided by a single experiment, or the experiment’s conclusion may not fit old theories. Sometimes scientists do not have any prior idea of a hypothesis before they start their investigations, but often scientists start out with a working hypothesis that will be proved or disproved by the results of the experiment. Scientific hypotheses can be useful, just as hunches and intuition can be useful in everyday life. But they can also be problematic because they tempt scientists, either deliberately or unconsciously, to favor data that support their ideas. Scientists generally take great care to avoid bias, but it remains an ever-present threat. Throughout the history of science, numerous researchers have fallen into this trap, either in the hope of self-advancement or because they firmly believe their ideas to be true.
If a hypothesis is borne out by repeated experiments, it becomes a theory—an explanation that seems to consistently fit with the facts. The ability to predict new facts or events is a key test of a scientific theory. In the 17th century German astronomer Johannes Kepler proposed three theories concerning the motions of planets. Kepler’s theories of planetary orbits were confirmed when they were used to predict the future paths of the planets. On the other hand, when theories fail to provide suitable predictions, these failures may suggest new experiments and new explanations that may lead to new discoveries. For instance, in 1928 British microbiologist Frederick Griffith discovered that the genes of dead virulent bacteria could transform harmless bacteria into virulent ones. The prevailing theory at the time was that genes were made of proteins. But studies performed by Canadian-born American bacteriologist Oswald Avery and colleagues in the 1930s repeatedly showed that the transforming gene was active even in bacteria from which protein was removed. The failure to prove that genes were composed of proteins spurred Avery to construct different experiments and by 1944 Avery and his colleagues had found that genes were composed of deoxyribonucleic acid (DNA), not proteins.