How to Use the Scientific Method

from wikiHow - The How to Manual That You Can Edit

The scientific method is the backbone of nearly all rigorous scientific inquiry. A set of techniques and principles designed to advance scientific inquiry and further the accumulation of knowledge, the scientific method has been gradually developed and honed by everyone from the philosophers of ancient Greece to the scientists of today. While there are some variations on the method and disagreement over how it should be used, the basic steps are easy to understand and invaluable not only to scientific research but also to solving everyday problems.

Steps

1. Observe. It is curiosity that breeds new knowledge. The process of observation, sometimes called "defining the question," is simple. You observe something that you can't readily explain with your existing knowledge, or you observe some phenomenon that is explained by existing knowledge but which may have another explanation. The question, then, is how do you explain that phenomenon--what causes it to occur?
   
2. Research the existing knowledge about the question. Suppose you observe that your car won't start. Your question is, why won't it start? You may have some knowledge about cars, so you'll tap into that to try to figure it out. You may also consult your owners manual or look online for information about the problem. If you were a scientist trying to figure out some strange phenomenon, you could consult scientific journals, which publish research that other scientists have already done. You'd want to read as much about your question as possible, because the question may have already been answered, or you may find information that will help you form your hypothesis.
   
3. Form your hypothesis. An hypothesis is a possible explanation for the phenomenon you observed. It is more than a guess, though, because it is based upon a thorough review of the existing knowledge of the subject. The hypothesis should posit a cause-effect relationship. For example, "My car won't start because I am out of gas." It should suggest one possible cause for the effect, and it should be something that you can test and which you can use to make predictions. You can put gas in your car to test the "out of gas" hypothesis, and you can predict that if the hypothesis is correct, the car will start once you add gas.
   
4. Test your hypothesis. Design an experiment that will either confirm or fail to confirm the hypothesis. The experiment should be designed to try to isolate the phenomenon and the proposed cause. In other words, it should be "controlled." Going back to our simple car question, we can test our hypothesis by putting gas in the car, but if we put gas in the car and change the fuel filter, we can't know for sure whether the lack of gas or the filter was the problem. For complex questions, there may be hundreds or thousands of potential causes, and it can be difficult or impossible to isolate them in any single experiment.
   
   • Keep impeccable records. Experiments must be reproducible. That is, other people must be able to set up a test in the same way that you did and get the same result. It's important, therefore, to keep accurate records of everything you do in your test, and it's essential that you keep all your data. Today there are archives set up which store the raw data gathered in the process of scientific research. If other scientists need to find out about your experiment they can consult these archives or ask you for your data. It's critical that you be able to provide all the details.
     
   
   
5. Analyze your results and draw conclusions. Hypothesis testing is simply a way to collect data that will help you either confirm or fail to confirm your hypothesis. If your car starts when you add gas, your analysis is pretty simple--your hypothesis was confirmed. In more complicated tests, however, you may not be able to figure out whether your hypothesis is confirmed without first spending considerable time looking at the data you gathered in your hypothesis testing. Furthermore, whether the data confirms or fails to confirm the hypothesis, you must always be on the lookout for other things, so-called "lurking" or "exogenous" variables, that may have influenced the results. Suppose that your car starts when you add gas, but at the same time the weather changed and the temperature increased from below freezing to well above freezing. Can you be sure the gas, and not the change in temperature, caused the car to start? You may also find that your test is inconclusive. Perhaps the car runs for a few seconds when you add gas, but then dies again.
   
6. Report your findings. Scientists generally report the results of their research in scientific journals or in papers at conferences. They report not only the results but also their methodology and any problems or questions that arose during their hypothesis testing. Reporting your findings enables others to build upon them.
   
7. Conduct further research. If the data failed to confirm your initial hypothesis, it's time to come up with a new hypothesis and test it. The good news is, your first experiment may have provided you with valuable information to help you form a new hypothesis. Even if an hypothesis is confirmed, further research is necessary to ensure that the results are reproducible and not just a one-time coincidence. This research is often performed by other scientists, but you may also wish to further investigate the phenomenon yourself.
   

Tips

• Note that you do not prove or disprove an hypothesis, but rather confirm or fail to confirm it. If the question is why your car won't start, confirming the hypothesis (you're out of gas) and proving it are pretty much the same thing, but for more complex questions that may have many possible explanations, one or two experiments cannot prove or disprove an hypothesis.
  
• There are many ways to test hypotheses, and the type of experiment described above is just one simple variety. Hypothesis testing can also take the form of double-blind studies, statistical data collection, or other methods. The unifying factor is that all methods collect data or information that can be used to test the hypothesis.
  
• Understand the difference between a correlation and a causal relationship. If you confirm your hypothesis, you have found a correlation (a relationship between two variables). If others also confirm the hypothesis, the correlation is stronger. But just because there is a correlation does not necessarily mean that one variable caused the other.
  

Warnings

• Always let the data speak for itself. Scientists must always be careful that their biases, mistakes, and egos do not lead to misleading results. Always report your experiments truthfully and in detail.
  
• Beware exogenous variables. Even in the simplest experiments, environmental factors can creep in and influence your results.
  

Related wikiHows

• How to Conduct Scientific Research
  
• How to Write a Hypothesis
  
• How to Conduct a True Experiment
  
• How to Conduct a Simulation Analysis
  
• How to Understand Classical Physics
  
• How to Get Kids Interested in Science by Crushing a Can With Air Pressure
  

Article provided by wikiHow, a collaborative writing project to build the world's largest, highest quality how-to manual. Please edit this article and find author credits at the original wikiHow article on How to Use the Scientific Method. All content on wikiHow can be shared under a Creative Commons license.