"These Boots Are Made For Walking"
[Introduction] [Procedure] [Hypothesis] [Data] [Graphs] [Calculations] [Conclusions] [Problems]
I set out to find out if there was any correlation between the grade or school a student was in and the steps they took or how far they walked through out a school day. In order to get the data I needed, I chose to use Washington Elementary for grades 1st through 5th, Bloomington Junior High for grades 6th through 8th, and Bloomington High for grades 9th through 12th. In choosing of the schools randomization was not needed for choosing the junior high or high schools because there was only one of each in my school district. The one school choice that needed to be random was the elementary, but it was not use. Randomization was not used in choosing the elementary school because a school that was close to the other two was needed to help in the data collection.
Project Outline and Procedure:
The first step in starting the project was making sure to alleviate all bias by making everything that took place, random. To ensure that randomization was used, the number of students in each grade were obtained. Every student was then assigned a number 1 through however many students were in that specific grade, alphabetically. A random integer producer was used to obtain five numbers for each grade, and the corresponding student from the list made for each grade was the student which was chosen to participate in the research. These students were then asked to wear a Digiwalker for one day of school. More can be found out about the Digiwalkers at www.digiwalker.com. Once the student's parents permitted their student to participate in the research study, the stride length of each student was found in order to be able to calibrate the Digiwalker to that specific student's walk. Once this was finished each student was assigned to a day in which they would wear the Digiwalker. The goal was to have no one student of the same grade to wear the Digiwalker on the same day of the week to help reduce bias and this way an average of a typical week could be found since no one day would have been over represented. Once the data gathering was started each day the data had to be recorded and the Digiwalkers calibrated for the next set of students until all the data was gathered. Eventually all the data was gathered, and it was compiled into a table. From the data table graphs were constructed and the proper tests were run on the gathered data.
Before any data had been gathered I had a prediction as to what the data would look like. I predicted that the steps taken in one school day would decrease slightly by grade, but would have an even more noticeable decrease when grouped by elementary, junior high, and high school. This was predicted because the students as they became older would probably have a larger stride, meaning it would take them fewer steps to get around. For the prediction of distance traveled I predicted to have a reverse of the steps, by having the distance traveled increase when grouped by elementary, junior high, and high school. I did not expect to see any trend grade from grade. I predicted a trend change based on the school they were in and not the grade because all the students in the same school would all follow approximately the same schedule; ultimately leading to similar distances walked in a day.
Data Sheet 1: Has the five students from each grade and how much they walked or how many steps they walked.
Data Sheet 2: Has the averages of the five students of each grade in both categories of steps and distance.
Data Sheet 3: Has the averages of grades 1-5 as elementary, 6-8 as junior high, and 9-12 as high school in both categories of steps and distance.
 |
 |
 |
 |
Calculations: This includes the null and alternative hypotheses for three different tests, which were all solved using chi-square.
I did all these test to make sure that the steps taken and the miles walked by each school was not the same. The first test was done to make sure that the steps for each school was not equal for each school. The second test was done to see if the amount walked by each school was equal. The third was done on the Junior High and High schools because they looked extremely close. I did not do a fourth test on the Junior High and High schools on the miles walked because of the results of the second test. (This will be further explained in Conclusions of Data.)
Upon closer inspection of the data via graphs and chi-square tests, interesting conclusions can be made. When looking at steps taken by all schools, it can be concluded that they do not all walk the same amount of steps in a day of school. This was found by the p-value of 0 produced by a chi-square test. The first test is not the only the test that allows for conclusions on steps taken. After using a chi-square test a p-value of .6997 was found, which leads to the conclusion that students of the Junior High and High schools both walk the same amount of steps in a day. The final conclusion pertaining to the amount of steps taken by the different schools is that students in the elementary walk less on an average day than both junior high and school students, while junior and high school students both walk the same amount of steps.
The graphs of the data and the chi-square test on miles walked in a school day yielded similar yet different results. Because the three averages for miles walked were close, a chi-square test was used to determine if all three averages of miles walked in a school day were the same. After finding a p-value of .9918 for the second test, it was followed by the conclusion that elementary, junior high and high school students all walk the same amount of miles in a day of school.
Problems with Observation Design and Other Bias:
A couple of things lead to problems that could be associated with the observation design. The one problem that seemed to be most prevalent in the observation was that no more than one student from the same grade went on the same say day of the week. Reasons for this problem was the time constraints placed upon the observation and that some students turned in permission slips on time and others had to be reminded several times before they turned it in. Because the student could not wear the digiwalker without a signed permission slip and there was no room to wait for a student to bring in a permission slip, the validity of the observation was forced to be compromised. It was compromised by having more than one student from the same grade wear a digiwalker on the same day of the week. The fact that the grade school was not randomly picked also allows for problems with the observation design.
Bias:
The biggest bias that allows for no way of knowing how much it affected the observation was the accuracy of the digiwalkers. This is because the digiwalkers are able to count the steps taken by the movement of a ball inside, which if the people wearing it were constantly bouncing in their seats or just wiggling around a lot; it would count as steps taken that really were not taken.
How the student wore the digiwalker could have also lead to a bias. In order for the digiwalkers to work properly they had to be worn horizontally on the waist. This allows for bias in the way that the student wore the digiwalker.
Bias would also be able to be found in when the student wore the digiwalker. For instance when the students changed for PE, they may not have transfer the digiwalker to their PE clothes while others may have.
The easiest thing to spot as a potential bias for this observation was that the number of samples for each grade was only five students. This could have been increased, but it was not possible to do because of the limited number of digiwalker available.
The one thing that played a role in my chi-square tests were that there was a low number of degrees of freedom. The problem with this was that for the grades 1 through 12 there are only three school types available, hence limiting the number of degrees of freedom that can be present in the observation. The only way this could be dealt with is by sampling three separate groups of undergraduate students, graduate students, and doctoral students. By doing this it would have raised my degrees of freedom from two to five.
The results have the potential to vary tremendously if taken at a different set of school because of the actual lay out of the school buildings. Some schools could be really spread out while others are small and confined.
[Introduction] [Procedure] [Hypothesis] [Data] [Graphs] [Calculations] [Conclusions] [Problems]
Any questions, comments, or suggestions on my web page please e mail me at alwet2004@hotmail.com.