Inductive reasoning involves going from the specific details and observations to the general underlying process that explains them
Deductive reasoning starts with a general explanation which leads to predictions for specific observations supporting it.
Both methods are used in science since the underlying goal is the establishment of new facts and their mechanisms. The fusion of these two types of logic is called hypothetico-deductive reasoning
The scientific process is not set in concrete but a number of steps are constant.
The hypothesis must lead to predictions which are testable. The hypothesis or its prediction should be written as an if-then statement, which points to a possible experiment. (This is an over-simplification but a necessary one for student labs)
Experiments, an organized, systematic method for testing hypothesis. Experiments should be designed so that there can be only one explanation for the results.
Controls are used to filter out other explanations. A control is a replica of the experiment with the independent variable omitted. The dependent variable is what is measured (like how much a plant grows from day to day). The control in human drug tests is generally a sugar pill called a placebo.
The skeptics society has an article discussing the misuse of the placebo effect
Uncontrolled variables often enter the experimental design. These variables must be removed or put under their own controls. In order to eliminate human bias most drug tests are conducted in a double blind fashion. In a double blind experiment researchers and patients do not know whether the pill given is the drug or the placebo until the trial is over.
Conclusion. Given the results obtained by experiment or review of historical data, a scientist will formulate a conclusion. He will decide whether the original hypothesis is supported by or has failed the test. In either case science is advanced.
An excellent review of the scientific method can be found at Fundamentals of Physical Geography
Modified on June 5, 2005