Mrs. Giannantonio's Science Fair Headquarters

Statistics are what set the great projects apart from the good projects. While complex statistical analysis isn’t compulsory at the tenth grade level, science fair judges are interested in projects that include some basic analysis

Statistics Guidelines:

• At the very least, the mean (average) of all data trials should be calculated. (In fact, I require this!) You should already know how to do this in Excel. Take this a step further by calculating the confidence interval. Are your data points close enough that your mean is significant?
• Over and above that, you may want to investigate whether or not your data is statistically significant when compared to your control or when compared to a given standard. There are two tests that are commonly used at this level to determine significant differences: the T-test and the Chi-squared test. Luckily for us, the complex calculations needed to conduct these tests are performed automatically if your data is stored on an Excel spreadsheet. Should you decide to perform one of these tests, it is not critically important that you understand all of the mathematics involved in the test, but it is of UTMOST IMPORTANCE that you understand the underlying principles behind the test! The tests help determine to what degree chance played a part in your different data points. There must be a 5% or less probability that your data differences were due to chance in order for your data differences to be considered significant.
• The University of the Sciences in Philadelphia’s Introductory Biology course has an excellent website that helps explain Confidence intervals, the T-test and the Chi-squared test and also helps you to calculate these tests on Excel and graph your results using trend lines and error bars (so you can really look professional!). Click on the following links:
  • For normal distribution and confidence interval
  • For T-test
  • For Chi-squared test
  • For information on correlation tests
  • For help with graphing and error bars
• Having trouble deciding which statistical test to use? Click on the following link, which will take you to a site called“Biology 4 All”. Robert Smith, the professor from USP took much of his information from this site, so you will see some overlap, however there is also a great chart that helps you determine which type of statistical test would be best for your data. http://www.biology4all.com/resources_library/43.asp

Refer to these links:
Science Fair checklist
Elements of a Good Research Project
Good Hypotheses
About Journals
About Abstracts
About Graphs
Using Statistics
About Discussions and Conclusions
About Backboards and Displays
About Forms
List of past projects (Microsoft Word format)