Great news! XYZ Advertising has called you in for an interview to join its advertising team. During the interview process, you are expected to solve a linear regression problem. Consider the following data in the table below that show the cost for a Super Bowl ad from the years 1990-2006 and the average number of homes and viewers of the game.
Based on the data in the attached document, please supply the following:
1) Provide graphs of Cost versus Year, Average Number of Homes versus Year, and Average Number of Viewers versus Year.
2) Which pair from the first part has the best linear graph for predicting? Explain.
3) Create prediction equations for all three graphs, and explain what these equations mean.
4) Complete a 5-year prediction for advertising costs. Next, complete a 20-year prediction. Explain the validity in these predictions.
5) In what year would you expect the average number of viewers to reach 100 million? In what year would you expect the average number of homes to reach 50 million? Explain the validity of these predictions.
Objective: Use the Cartesian coordinate system to find the distance between two points; illustrate linear functions and their relationship to business, economics, social science, etc., and find the intersection of two lines.
A file for is attached for reference.
See the attached Excel spreadsheet which contains the graphs. First, what you do is you must type out the numbers from the table into the spreadsheet manually. This is a pain, but it must be done.
Next, you construct simple "Line" graphs by highlighting the columns you want in the graph. After you construct the 3 graphs, then you adjust the y-axis so that the data points fill the graph nicely and you can see the data easily.
Clearly, when you look at the graphs, you can see that the most "linear" of the graphs is the first one, the relationship between year and cost. There's a really easily distinguishable pattern in that graph. The other two look kind of "messy," don't they? Therefore, they aren't as good when it comes to prediction.
In order to create prediction equations, right click on the data points in each graph. A drop down menu appears. Choose "add trendline." If they are intended to be linear (which I assume they are supposed to be), then choose "linear" as the Trend/Regression type. ...
This response provides aid to solving a linear regression model. Prediction equations are focused on.