Designing a rigorous scientific investigation

A large part of the judging on your project will be based on your experimental design. A great idea that is poorly executed will not score well, but any idea that is well thought out and rigorously investigated will do well. The tips below, including several videos, can help you get started on designing your project.

  1. Your problem should include one changing variable (the independent variable).
  2. Your problem should include one measurable variable (the dependent variable) that you will measure as a result of changing the independent variable. This short video summarizes the differences between the dependent and independent variable.
  3. Anything else that might affect your measurements should be carefully controlled as much as possible.
  4. Every measurement should be made multiple times to ensure that it is reproducible. The number of replicate measurements will probably depend on the nature of your experiments. At a minimum, you should have three replicates of every single measurement that you do. If there is large variability in your measurements, you should collect more.
  5. Your data analysis should include required statistical analysis. This usually will involve at a minimum calculation of mean and standard deviation, and often a t-test to compare data sets. For an explanation of a simple t-test for comparing data sets, see this short video.
  6. Usually your data can be presented in tables or graphs. Choose an appropriate graph for your data. Here are some videos that can help you choose an appropriate graph and use Excel to present your data.
    1. Types of graphs and when to use them
    2. Bar graphs vs. line graphs using zombie data