Statistical Analyses for Project Fit America

Challenge: Project Fit America (PFA) is a national non-profit organization that supports fitness in education programming for grades K through 8th for schools in impoverished areas throughout the country. Their mission is to get kids fit and to give teachers the tools needed to teach children about fitness and health.  PFA brings exercise equipment into selected schools, along with teacher training, tools and support to maintain a thriving fitness program. PFA to date has approximately 1,000 participating schools in 300 cities and 46 states with over 500,000 kids impacted by the program.  PFA is able to accomplish this with the support of sponsors, predominantly hospitals and other healthcare organizations.  PFA has created ways to measure the success of their program; however, they do not have the capability to analyze the data provided by the participating schools.


Research Approach and Methods:  For the first two years of participating with PFA schools are required to collect data on fitness tests to measure the success of the program. This data is used to determine if there was a significant improvement in the participants’ physical fitness. There are five fitness tests, and these tests are administered by teachers at the school in the beginning and at the end of a school year. Those scores are then given PFA, which gives the data to the Sonoma State University (SSU) Statistical Consulting class to conduct a statistical review. Using SPSS, the data is analyzed to measure if there is an improvement in the participants’ physical fitness after they have participated for one year in the PFA program.

Key Learnings: For each school the data is cleaned up, tested for normality, and then the results are analyzed using the appropriate statistical measure (most often Paired-Sample T-Tests and Wilcoxon Signed-Ranks tests). The results from those tests indicate whether there was an improvement in the participants’ physical fitness. Some schools showed improvements on all fitness tests, while others improved only in some areas. Individual reports were drafted for all participating schools and presented to the client, PFA.  Improvement often was linked to teacher enthusiasm for the program, school location and overall health of their student population. 


Impact of Research Findings: Having analyzed the data for many schools participating in this program, it is evident that there are numerous variables that can affect how valid and reliable a research study can be. Consistent collection of data is essential. In the case of PFA’s data, any difference in how the pre- and post- fitness tests were measured impacted the validity of the results. Valid results also depend on a large enough sample of participants – if there is only data on 10 participants then the results are going to be skewed to show a higher than average success (or failure) rate. For a program like PFA that collecting data every year, being able to analyze the success of your program allows them to determine what is not working and what needs their attention. They can then address those issues and fine-tune their program that brings enriching experiences to so many kids