Regression Analysis Section Three Exercises
The following are application exercises from section three involving Simple Linear Regression problem #41 and #53 and Multiple Regression problem #20 and #48. Number #41 involves a portion of the computer output for a regression analysis relating to y = maintenance expense (dollars per month) to x = usage or hours per week of a particular brand of computer terminal. The regression equation is as follows:
(chart)
The following are answers to problem #41 found within the class text. First, the book wants to know the estimated regression equation. The estimated regression equation is Y = 6.1092 + 0.8951 X. Next, they want to use a t-test to determine whether monthly maintenance expense is related to usage at the 95% level of significance. The answer is: t = 6.53 and t = 6.01. There is significance relationship because the t-values are greater than the t .05 at 1.869. The last question is regarding the estimated equation used to predict monthly maintenance expense for any terminal that is used 25 hours per week. Multiplying 25 hours by four weeks gives us 100 hours per month. If I were to plug that into the equation, I would get the following results. 100X (.08951) + 6.1092 = $95.62. Therefore, the predicted monthly maintenance cost would be about $95.62 a month.
In question #53, Neilsen Media Research collects data that shows the number of households tuned into shows that carry a particular advisement. This data is beneficial because advertisers can find out how many consumers they are reaching with their promotional plan. The following data show the number of household exposures in the millions and the number of times the ad was aired for November 24-30, 1997.
(chart)
