Powerpoint

1.

Descriptive statistics:

What is the average rating for all CBC movies? How about ABN movies and BBS movies?

Include a table that shows the average and the other descriptive statistics for the ratings of the three networks (one column for each network). Comment on which network is doing best and what you learn from the other key metrics in the table.

The average ratings for BBS, ABN and CBC are 12.7, 14.8 and 13.4 respectively. ABN’s rating is the highest and BBS’s rating is the lowest. The median, mode, maximum and minimum for ABN are all larger than those for BBS and CBC. All these facts indicate that ABN is the best.

For the spread of the data, ABN’s standard deviation is the smallest (2.27), meaning that the results are more consistent than the other two companies.

The Skewness values indicate that ABN’s rating is left skewed (having relatively few low values) and the ratings for BBS and CBC are right skewed (having relatively few large values).

The Kurtosis values indicate that CBC’s rating has a sharper peak (having more values in the center) than normal distribution and the ratings from the other two companies have a more rounded peak (having less values in the center) than normal distribution.

2.

Charting:

Create a line graph of the monthly average ratings for CBC for the year. Note that there are multiple ratings data for the months; you will need to calculate an average for each month and then plot the averages. After you create the graph, fit a linear trend line, displaying the formula and the r-squared. Explain to the executives if you can use this time series data to forecast the ratings of upcoming months. How accurate can you expect this forecast to be?

The chart is shown above with regression line and an R squared value of 0.339. The regression line is not significant (since the p-value is about 0.13) , meaning that it’s not accurate to use this equation to predict future ratings.

3.

Hypothesis testing :

Should the CBC hire stars for their movies? To answer this question, run a hypothesis test to see if there is a significant difference between the ratings of movies with stars vs. movies without stars. Use the data for CBC movies only. Use 95% confidence. Explain your answer donot simply say yes or no without referring to the relevant figures.

A two sample t-test is performed and the p-value is 0.58 for a two-tailed test. This p-value is large and so there is no evidence to show that there is a significant difference in the two programs. The data does not support the idea of hiring stars for CBC.

Unit 6

[GB513: Business Analytics]

4.

Regression:

CBC Management has several questions: Which has more impact on a movie’s rating: that it is fact based or that it has one star? How much does each of these factors change the ratings?

Do you expect a fact based movie that does not have any stars to get better ratings than a fiction movie with one star? Run a multiple regression where the dependent variable is ratings and the independent variables are star and fact. Use data from all networks, not just CBC

movies. How well does this regression analysis explain the ratings? Justify your answers

A multiple regression model is fitted. The F test p-value is 0.001695, meaning that the model is highly significant and it explains the ratings very well.

The equation is

Rating = 12.57 + 1.80 Fact + 1.26 Stars.

The p-value for the coefficients of Fact and Star are 0.0013 and 0.013 respectively. So both factors are highly significant.

The coefficient of 1.8 means that on average, a fact based movie will get a rating 1.8 higher than a fiction movie if the level of Star is fixed.

The coefficient of 1.26 means that on average, a movie with star will get a rating 1.26 higher than a movie without a star if the level of Fact is fixed.

For a fact based movie without a star, the estimated rating is

12.57 + 1.8 = 14.37.

For a fiction movie with a star, the estimated rating is

12.57 + 1.26 = 13.83.

Since 14.37 > 13.83, so on average, a fact based movie without a star will get a higher rating than a fiction movie without a star.

Rating 1 2 3 4 9 10 11 12 11.633333333333333 11.879999999999999 12.4 9.9833333333333325 12.433333333333332 12.624999999999998 13.233333333333334 12.475