How do Scientists Design Valid Experiments?
Science is Based on Experiments That Yield Reliable Data
Jul 13, 2009
Science may be defined as the search for natural truths (facts and information) undertaken by exacting individuals (scientists) using precise and reliable methods. What are these scientific methods? How do scientists design and conduct experiments to get at the truth?
Experiments are a series of components (steps) usually carried out in sequence.
The Problem is Established
Component 1: The Problem. Curiosity leads to questions to be answered and problems to be solved. The first step must be to specify exactly what is to be solved. This is accomplished through the problem question. The problem question should:
- clear and concise language.
- he specific kinds and amounts of materials and supplies that would be required to conduct the experiment.
- ample size and sampling methods. The rule of thumb on sample size is: the larger the sample, the more accurate the results. Would you be suspicious of a new medicine that was advertised to have miraculous powers if you discovered that the drug had only been tested on 20 people? 200 people? If not, you should be.
- <li style="l
- hat data will be collected and how the data will be organized. In the rat growth experiment example, one would carefully weigh the rats in both the test group and the control group over a given period of time organizing the weights obtained into a data table.
Data is Analyzed
Component 4: Analyze the Data. Once scientists collect the raw data from a well-designed experiment what happens to it? How do scientists make heads or tails out of what the numbers (data) may be trying to tell them?
One method of analysis is to graph the data. Graphs turn numbers into graphics which is useful because pictures are more easily understood than lists of numbers. Another analysis scientists perform is to run statistical tests on their raw numerical data. Statistical tests are mathematical processes that help determine if the data collected are significant and if so, how significant.
Component 5: Draw a Conclusion. Analysis of the data will hopefully reveal whether the original hypothesis was correct (supported by the data) or incorrect (not supported by the data). That in turn should answer the original problem question. In the conclusion component, scientists attempt to put the results in perspective, establish the significance of the results, and explain how the experiment fits into existing knowledge.
If science is the search for natural truths, then scientists must be very exacting in that search to ensure that truths are what result. Well-designed experiments are the key to the reliability of the entire scientific enterprise.
 |
