Final Stats Project

Background:
    For quite a long time, I have been interested in why advertising is so crucial for big to medium fast-food companies and why there seems to be so much time, effort, and money pushed into such endeavors. This project provided me an opportunity to look more into the topic since I always needed more time to do so myself. The data I am using is the Fast-Food Marketing Campaign A\B Test, which essentially took three marketing methods and recorded data for various fast-food restaurants varying in size, location, and age.

Problem: 

    The problem or thing that is being addressed by me and the data is the question of these three marketing campaigns, which is most effective now. This could go further than I will be as you could also include the size of the restaurant's market, the establishment's location, the age, and add in the weeks and each week's financial gain. I decided to stick with just the different methods and campaign strategies.

·      

Hypothesis:

• The Null: All of the methods work to equal results in the form of sales. Meaning that there is not a statically significant difference in performance. 
  

• The Alternate: At least one of the methods gives a varying result in the form of sales meaning that there is a statically significant difference in performance.
  
  

Solution:

    My solution to this problem/query was to run a T-test on groupings of the campaign strategies. I grouped them like this: (1 with 2, 2 with 3, and 1 with 3), they examined the t-test, looking for things that would show significant evidence to reject the null. This evidence was displayed in Promotions 1 and 2 with P values of 1.03e-05, 2.97e-10, or 4.29e-10, respectively, depending on the test, as I also ran them through an ANOVA test. At the same time, the third promo seemed to underperform the other two options severely, with P values of 0.0552 or 0.1206. This still results in me rejecting the null hypothesis that all of the promotions would perform the same.

Summary:


    With the intent to satisfy my curiosities about the fast-food industry, I set out to find the reason and success behind marketing and advertising campaigns in more recent years. To do this, I found a data set that listed three different marketing campaigns tested on various fast-food restaurants varying in size, location, and age over four weeks. Having established a null stating that all the promotions perform roughly the same and an alternate hypothesis states the opposite. With this data, I ran ANOVA and T tests on the data, which I had separated by promotion types 1,2, or 3. Doing this led to the conclusion that promotions one and two experience more statistically significant success than those who went through the 3rd promotion, in which we found instead of rejecting the attributing to the rejection of the null hypothesis like the other two, it accepts it.