The basis of all inferences and decisions is data, which are simply facts and figures that bear directly on the problem we wish to solve or the question we wish to answer. Such facts or figures are collected by either observation or experimentation. Observation is passive in which the observer wishes to record data without interfering with the process being observed. Experimentation is active in which the experimenter attempts to control completely the experimental situation.
The difference is illustrated by the works of Tycho Brahe and Galileo at the beginning of the Scientific Revolution. Brahe devoted his life to recording precisely the positions of stars and planets and left records from which Kepler deduced his laws of planetary motion. This was observation.
Galileo studied motion under the influence of gravity by rolling balls of various weights down inclined planes of various lengths and angles. This was experiment.
The advantage of experimentation in an experiment is when researchers actively intervene by administering a treatment in order to study its effects. The great advantage of experimentation is that we can study the effects of the specific treatments that interest us, rather than simply observe units as they occur “in nature”. Imagine the frustration of a researcher trying to study the effects of prolonged sleeplessness on reaction time by finding persons who just happens to have been awake for 48 hours. Instead he or she performs an experiment of keeping volunteer subjects awake for 48 hours and then measuring their reaction time. He or she can even keep the subjects awake for 36, 48, 60, and 72 hours and measure their reaction time for each duration of sleeplessness.
In principle, experiments can establish causation. If we change the value of an independent variable with no other changes in the experimental conditions, any resulting changes in a dependent variable must be caused by the changing independent variable.
In experimentation, it is important to know some vocabularies used such as: Units, defines as the basic objects on which the experiment is done. Variable, defined as the measured characteristic of a unit. Dependent variable, a variable whose changes we wish to study, a response variable. Independent variable, a variable whose effect on the dependent variable we wish to study. An independent variable in an experiment is called a factor. Treatment, any specific experimental condition applied to the units. A treatment is usually a combination of specific values (called levels) of each experimental factor.
The first goal of experimental design is to obtain valid results, to discover the true effects of the treatments. The second goal is to do this as efficiently as possible, to use as few units as possible for a given degree of precision.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment