Maria n Phillls Website

 

Statement of the Problem...

 

For our final project, we chose to test whether the mean purse weights for girls who regularly carry food in their purse is higher than the mean purse weight for girls who don't carry food.  We got this idea after looking around in our own high school.  We noticed that girls purse sizes vary extremely.  They range from the size of a wallet to the size of backpack able to carry textbooks.  What girls carry in their purses also varries, from make up to meals and a whole change of clothes.  Seeing this, we wanted to see if the weight of the purse was related to what the girls carreid in them.  To test this, we decided on our project, testing if the weight was greater for those who carry food than for those who don't.

 

 

What is a Handbag???

A Handbag is not just a fashion statement any more. Today’s handbags must be large, sturdy and functional to support the many needs of today’s lifestyles. Today many women carry more than one bag. This has become a vital necessity for some of today’s women. One bag is not enough. The importance for a woman to have the right bag has also made the need for business cases, totebags, sportbags, cell phone bags, clutch purses as well as the handbag, to be a basic part of a woman's everyday toting essentials. One  thing that never changes is the secrets that lie beyond the closed handbag. (What is a handbag?:)

 

Abstract...


For our statistics project, we decided to test whether the mean purse weight for girls who carry food in their purse was higher than the mean purse weight of girls who don’t carry food.  We decided on this topic after seeing how large girl’s purses in our school were and the multitude of different things they carried in them.  After deciding on a study and creating a random sample of 100 high school girls from our population of interest of all the girls attending North Olmsted High School, we were then faced with the decision of how to measure the purse weight.  We decided to go to each of the girl’s study halls, or when those weren’t available, classes, and measure their purse weight ourselves with our own scale to minimize the amount of measurement bias.  
    After weighing each girl’s purse, we then set ourselves to the task of interpreting the data and seeing if there was anything of statistical value in it.  The first thing we had to do was get rid of the response bias.  We had quite a few of these as some of the girls in the sample didn’t carry a purse or decided against being in the study.  We decided to throw these people out of the sample as to not skew any of the data.  After we did this, we found something rather interesting and that could possibly endanger our results.  We measured girls in 2 periods: 3rd, the period right before the lunch periods begin, and 10th, the period directly after the lunch periods end.  We found that of the girls measured in each period, there were a higher proportion of girls who had food in 10th period.  This could very well skew our data and make any assumptions we may make useless.  Nonetheless, we carried on.  
    We did a 2 sample t-test on our data to see if the mean weight of girl’s purses whom carried food in their purse was higher than the mean weight of girl’s purses who did not carry food.  As our alpha value, we used 0.05.  After carrying out the test, we obtained a p value of 0.0136.  This showed that at the 0.05 level of significance, the mean weight of girl’s purses whom carried food in their purse was higher than the mean weight of girl’s purses who did not carry food.  We were very pleased with this result as this was what we had hypothesized would happen at the beginning of our study.  
    Although we did get the result we were anticipating, there were a few weaknesses of our study that could have skewed the results.  First is the fact that if the girl had food or not could be dependent on time of day.  This could alter our data depending on what time we measured the girl’s purse.  Also, to successfully measure the purse, we had to weigh myself holding the purse and then subtract my weight from it to get the purse weight.  To take out most of the affect my weight had on the data I was weighed every day before we went out to measure to account for any change in weight I had had.  I feel that this should have taken out most of the measurement bias out, but there are definitely some unavoidable affects to measuring this way instead of measuring the purse alone.  Lastly, some of the data could have been altered by variation in the type of purse the girls chose to carry that day.  Some of the girls we measured made comments like my purse usually isn’t this big or I left my big purse at home today.  This variability could have changed our data depending on what size purse the girl decided to bring on the day we measured them.  Hopefully, using a fairly large sample size minimized this affect, but it still could be a factor.
    Looking back on the study, it went fairly well.  To improve it in the future, I would use a more precise way of measuring the purses to minimize measurement bias.  I would also measure all the girls at the same time to take out the factor of whether they had had lunch yet out of the equation.  Other then these small changes, I wouldn’t change the study much except maybe to extend the population of interest and increase the sample size, which we weren’t able to do with our limited resources.