The presentations are complete and everything went as planned. My data was actually contradictory of what I believed it would be. When I first set out for my project, I expected to see that the New Orleans restaurants would have more business, but this was not the case. I saw that the two areas had about an even amount of patrons in the local restaurants. Seeing this, I began to assess the possible reasons for this. The one I came up with is that while the French Quarter is the center of tourism in our city (which is why I assumed they would have more business), it does not necessarily mean that the tourists are expected to dine at these establishments. Also, those visiting our city are not likely to stay in a hotel near the French Quarter, for the high costs, therefore may stay in hotels located near Metairie and Kenner, in turn causing them to dine close for a quick meal. It was indeed interesting to see such results from my experiment. I carried out a t-test (A statistical test to determine whether the difference between two sample means is statistically significant) with my data and came out with the following results:
t= .4884
p= .6312
mean 1=66.1 (mean on French Quarter data)
mean 2=58.6 (mean on Kenner and Metairie data)
standard deviation 1= 34.3287 (std. dev. of French Quarter)
standard deviation 2= 34.3485 (std. dev. of Kenner and Metairie data)
Final.
Data Update.
This past weekend, I completed my data collection. I was able to squash my worries of variability in my data as well as keep some sort of pattern in my data, and eliminate as many confounding variables as possible; I collected all of my data from up and down Veterans Boulevard. By doing so, I eliminated the amount of distance between restaurants that would have been present if I were to collect the data from different parts of the city. This also allowed me to add in a similarity between the two data collections (the New Orleans data was collected up and down Decatur). I used the same tactics to collect this set of data, as before--estimating if absolutely necessary and asking the hostess if possible. As you can see, there are much lower numbers at some restaurants in this area, and I concluded that there are two possible reasons for this-- that either a)the restaurant may have been having a slow night or b)that in the location of the restaurant is not as popular. Now that I have completed my data collection, I need to decide what programs would be best suited to analyze my data. I am not very familiar with MiniTab, so I am going to attempt to do my data analysis using Microsoft Excel. Aside from data analysis, I need to begin work on my final presentation. My plan is to attempt to make an interactive PowerPoint using the Promethean board (a board used in class, that when paired with a laptop and the proper technology, is interactive and touch screen). Whether or not this is plausible will be deciphered in the beginning of next week. I am completely open to suggestions on data analysis as well as how I should carry out my presentation. So, please, give me some feedback! Anything is welcome.
Food for Thought.
I began to think about the second half of my data and how it differs slightly from the criteria of my first. By this I mean that the fact that all of the restaurants from which I collected from in the French Quarter were located almost right next door to each other, and in Metairie and Kenner, this is not the case. The restaurants in these areas are located blocks away, sometimes with three or four streets in between. I am starting to wonder if this will have a profound effect on my data--possibly adding in some sort of bias into my results. Does anyone have any suggestions or comments on this issue? In the end I guess it might just be true when people say, "Location is everything."
Data Collection Update.
At this point I have collected my data from the restaurants in the French Quarter. Between the hours of 6 and 8 on Saturday April 4, I was able to collect data from 10 restaurants: Landry's, Hard Rock Cafe, Jax Brewery, Masperos, Tujague's, Envie, Margaritaville, Frank's, The Corner, and Bubba Gump. While some restaurants were easier to enter, others had lines out of the door. For those that I could actually step foot into, I asked the hostess for help. I explained my project to him/her, and in turn asked if they could give me the number of people per table and did simple addition to find the total. For the ones that had lines out of the door, or in which I was unable to communicate with a worker, I guesstimated. I counted by twos or fives, depending on how crowded the establishment seemed to be. By looking at my data, you can tell which ones I was able to get an exact number and which were estimated to the best of my ability. Although I did have to somewhat guess at a few, I think the numbers were reasonable enough to be close to the actual number, therefore while it will have some effect on the graphing of my data, it will not be as sever as it could have been. It will in fact contribute some error to my final project, one that I wish could have been avoided.
Now that I am near to completing my data collection, I will begin to use Excel and/or Minitab data analysis programs) to analyze and graph my data.
Data Progress.
Data collection--stress central. With a project such as mine, the data collection process is quite tedious and requires much thought. After much consideration, I decided not to pre-make a list of restaurants, but rather to visit the French Quarter, Kenner, and Metairie and collect data from as many restaurants as humanly possible in the time allotted. My plan is to enter each restaurant and estimate to the best of my ability the amount of patrons in the establishment. If this seems to be difficult, I will attempt to ask the hostess the number of people at each table and add up the total. The next task, which is almost more difficult than the first, is deciding what to do with the data after I have collected it. In the beginning, I had decided to only produce a scatter plot to display my data, but then realized that this would not work because that is used to show correlation--something that is not necessary with this activity. I then decided to make a box plot of each set of data to show the five number summary of each, just to give an overview of the data. I also decided to make an o-give (describes the probability distribution of real valued random variables) and histograms (graphical display of tabulated frequencies) for each data set. By performing each of these displays, I am giving my audience a chance to see the data in different settings--one in which they see the pattern of number of customers in each restaurant, and the probability of each number. By viewing the box plots of the data, it can be shown which setting receives more business, which can be helpful in many ways (refer to previous blog entry entitled "How is this useful?") I would really like helpful feedback on my process, both pre and post data collection. So please, tell me what you think!!
Changes!
This week I had to make some changes to my project. I found that the direction in which my project was heading was not going to work. It was way more feasible to compare and contrast the business levels between restaurants in the French Quarter and those restaurants located in surrounding cities such as Kenner and Metairie. Also, I decided not to limit myself to a certain number of restaurants-- I decided to visit as many as possible in a given time. By doing this, I am giving myself a wider range of data. It will also help me show a better correlation between the business levels. I am actually more excited about this project layout because I think people will care more about the comparison between the business levels in different areas as opposed to the comparison between the different cuisines.
How Is This Useful?
The data that I will be collecting may not seem important at first glance, but if you really think about it, it can be quite helpful. One of the most important reasons for this is the economic situation. In tough times as these, entrepreneurs are very skeptical when it comes to opening a business. By carrying out my project, I can possibly display the areas that are best for business. For example, if my data shows that there is a significant increase in the amount of business from Kenner to the French Quarter, then an aspiring restaurant owner would be more likely inclined to opening a restaurant in the Quarter. The same works if more business is shown in the smaller cities. Also, it may not be scientifically or mathematically proven that restaurants with more business are actually better in quality, but some may see it this way. Therefore, whichever area of New Orleans my data shows receives more business, that displays that I choose may supply these establishments with even more business. The data can build up the hype around each restaurant and make people that are natives to the areas, and even tourists even more interested in visiting them.
My data can also be useful to the restaurants themselves. I will provide my results to all restaurants involved, showing the differences in the amount of business between the different establishments and different areas of New Orleans. By seeing that restaurants in other areas may receive more business than they do will, of course, not make them want to move their location, but will most likely take this into consideration. They can reevaluate their establishment and make the proper changes to increase business.
While my data may not be as vital as some, it is vital in the business aspect. It can help people who are struggling entrepreneurs or already struggling business owners.