Test a Hypothesis

Testing a hypothesis is an important part of the scientific method. It allows you to evaluate the validity of an educated guess. The first step is to form a good hypothesis, including the null hypothesis. Next, determine how the results of the study will be analyzed. Finally, after the experiment has been carried out, you can interpret your data.

Steps

Stating Your Hypothesis

  1. Form an educated guess regarding outcomes. An educated guess is formed by making observations or reading literature from past experiments. This guess is informed by legitimate scientific observation. The better supporting information you have, the better your hypothesis will be.[1]
  2. Fashion this guess into a formal statement. Just having an idea of what might happen isn’t necessarily a hypothesis. A hypothesis must be capable of being tested. If there is no way to do an experiment that might prove your statement wrong, it is not a hypothesis.[2]
    • For example, the moon is big is not a hypothesis. However, the moon is bigger than Texas is a hypothesis.
  3. Include the null hypothesis. Your hypothesis wouldn’t be complete without the null hypothesis. The null hypothesis is the assumption that your hypothesis is incorrect. If your hypothesis stands up to scientific investigation, you can successfully reject the null hypothesis.[2]

Setting Parameters

  1. Choose a confidence level. A confidence level refers to the likelihood that your results could have happened by chance. The most common confidence level is 0.05. At this level, there is only a five percent chance that your results could come from coincidence.[2]
  2. Identify the test you will use. The most common test used is the p test. When you calculate a p-value for your experiment, you can compare it to a standard chart. You can then determine if that p-value fits into the significant category (based on your confidence level).[3]
  3. Select an appropriate sample size. A hypothesis makes an assumption that, under certain circumstances, you will always receive the same outcome. For this reason, you may have to test the hypothesis with several different samples. This will ensure accurate results.[2]
    • For example, if your hypothesis is that all children get hyper when they eat sugar, you could not prove your hypothesis by testing one child. Instead, you would have to have a large sampling of children to administer the test.

Gathering and Interpreting Data

  1. Carry out the experiment. Whatever test or experiment has been devised to test the hypothesis must be done according to the scientific method. Record your procedure as you do the experiment. You may have to repeat it several times to ensure validity, and it must be repeatable for other scientists.[1]
  2. Record all results. During each trial of the experiment, you should record your results. Keeping a lab notebook will help ensure that your recording happens accurately and in real time. Your results can be published and reviewed once the experiment is complete.[1]
  3. Evaluate the outcomes. Once the experiment is finished, you can analyze the results. This will be done by administering the p-test (or other statistical test) to determine the significance of the results. If they are significant, you may reject the null hypothesis.[1]

Sources and Citations

  1. ↑ 1.0 1.1 1.2 1.3 https://learn.bu.edu/bbcswebdav/pid-826908-dt-content-rid-2073693_1/courses/13sprgmetcj702_ol/week04/metcj702_W04S01T05_fivesteps.html
  2. ↑ 2.0 2.1 2.2 2.3 http://support.minitab.com/en-us/minitab/17/topic-library/basic-statistics-and-graphs/hypothesis-tests/basics/example-of-a-hypothesis-test/
  3. ↑ https://onlinecourses.science.psu.edu/statprogram/node/138
Retrieved from "https://kipkis.com/index.php?title=Test_a_Hypothesis&oldid=290636"