TIME OF DAY VS. STORE BUSINESS

                

 Background Research | The Study | Discussion | Conclusion

  


STATEMENT OF THE PROBLEM

  

For over a year and a half I worked at a grocery store as a cart attendant.  I would normally work from five o’clock p.m. to nine o’clock p.m.  It seemed that there were significantly more customers coming at the beginning of that shift than at the end.  This course has given me the tools I need to prove or disprove what I have thought this whole time:  more people shop at grocery stores towards the middle of the day than early in the morning or late at night.  To study this I will go back to the store I worked at, since that is where I first observed this effect.  Though it was open 24 hours, the store never scheduled cart attendants before seven o’clock a.m. or after eleven o’clock p.m. because the store expected little business during this time.  Thus I took the middle hour for this interval as one of my samples.  I chose a time period late at night for the other sample.

ABSTRACT

  

This study was designed to determine whether there was a difference in the number of people who entered a grocery store between 2:30 p.m. and 3:30 p.m. was greater than the number of people to enter a grocery store between 8:30 p.m. and 9:30 p.m.  My study took place from March to May of 2007.

            The background research I conducted yielded virtually no information.  I checked local grocery stores and the Internet for any studies similar to or related to my own.  I found absolutely nothing, meaning my study is unique.

            My plan for collecting data was to choose a particular store and then to count the number of people who entered the store on 30 randomly selected days for each interval.  The store I chose was the Giant Eagle in Rocky River.  I had worked at the store for over a year and it was during my employment that I first noticed this effect.  I planned to collect at least 30 pieces of data.  Assigning each possible day during the data collection phase a number and then using a random number generator to choose 32 days for each of the intervals accomplished this.  Of the 32 days drawn for the 2:30 p.m. to 3:30 p.m. time interval there was one day I could not conduct the study because of the AP Statistics exam.  I was able to conduct my study on all of the days for the 8:30 p.m. and 9:30 p.m. time interval.

            Right away I realized that I was probably going to reject my null hypothesis.  The lowest amount for the 2:30 p.m. and 3:30 p.m. time interval was greater than the highest amount of the 8:30 p.m. and 9:30 p.m. time interval.  When I performed my two-sample t-test I received a T-value of 18.83, which gave me a p-value of zero.  Therefore I could reject the null hypothesis at any significance level.

           However, like most studies, mine had many weaknesses.  First of all, I could do nothing to account for any change in weather that might affect my data.  Also, it should be noted that my study only counted the people who entered the store and no the people who made purchases.  Finally, the population to which my study can be extrapolated is relatively small.