Scribd Logo
Upload

Welcome to Scribd!

  • 19 views

    Advertising Sales ($) Average # of Viewers (Millions) Length of Program (Minutes) Average Viewer Age (Years)

    Uploaded by

    Kiran Jawaid
    The document describes building multiple linear regression models to predict advertising sales based on various metrics. A regression model with two predictors (average viewers and length of program) was not useful as only average viewers significantly impacted sales. A third predictor (viewer age) also did not improve the model. Ultimately, a regression model with just average viewers as a predictor was found to best predict advertising sales. This single variable regression equation can be used to make the two example predictions required.

    Copyright:

    © All Rights Reserved
    Download as XLSX, PDF, TXT or read online from Scribd

    Advertising Sales ($) Average # of Viewers (Millions) Length of Program (Minutes) Average Viewer Age (Years)

    Uploaded by

    Kiran Jawaid
    19 views9 pages
    The document describes building multiple linear regression models to predict advertising sales based on various metrics. A regression model with two predictors (average viewers and length of program) was not useful as only average viewers significantly impacted sales. A third predictor (viewer age) also did not improve the model. Ultimately, a regression model with just average viewers as a predictor was found to best predict advertising sales. This single variable regression equation can be used to make the two example predictions required.

    Original Description:

    Excel Format for Decision Analysis

    Original Title

    excel_template_4th_assignment

    Copyright

    © © All Rights Reserved

    Available Formats

    XLSX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document describes building multiple linear regression models to predict advertising sales based on various metrics. A regression model with two predictors (average viewers and length of program) was not useful as only average viewers significantly impacted sales. A third predictor (viewer age) also did not improve the model. Ultimately, a regression model with just average viewers as a predictor was found to best predict advertising sales. This single variable regression equation can be used to make the two example predictions required.

    Copyright:

    © All Rights Reserved
    Download as XLSX, PDF, TXT or read online from Scribd
    Download as xlsx, pdf, or txt
    19 views9 pages

    Advertising Sales ($) Average # of Viewers (Millions) Length of Program (Minutes) Average Viewer Age (Years)

    Uploaded by

    Kiran Jawaid
    The document describes building multiple linear regression models to predict advertising sales based on various metrics. A regression model with two predictors (average viewers and length of program) was not useful as only average viewers significantly impacted sales. A third predictor (viewer age) also did not improve the model. Ultimately, a regression model with just average viewers as a predictor was found to best predict advertising sales. This single variable regression equation can be used to make the two example predictions required.

    Copyright:

    © All Rights Reserved
    Download as XLSX, PDF, TXT or read online from Scribd
    Download as xlsx, pdf, or txt
    of 9

    Average # of Length of

    Advertising Average Viewer


    Viewers Program
    Sales ($) Age (years)
    (Millions) (Minutes)

    28,200 11.1 30 31 <--As usual, the template contains different numbers f


    25,300 10.4 30 24 We must build two multiple regression models: the fir
    31,100 18.9 60 31
    28,900 14.6 60 37 Building the multiple regression model with 2 indepe
    20,000 13.5 60 21 Go to Data tab>Data Analysis>Regression>Select the a
    15,000 4.5 30 14 variables (average # of viewers and length of program
    27,300 16.1 60 25 You will obtain a new sheet (Model A). Now, each ind
    23,200 2.7 30 18 For example, the slope for Average # of viewers is 744
    19,700 5.3 30 18 but the slope for Length of Program is not significant (
    22,800 11.2 120 44 The conclusion is that this regression model is not use
    18,500 6.1 120 18
    29,000 14.9 60 29 Building the multiple regression model with 3 indepe
    30,100 17.8 120 30 Repeat the steps that are indicated above and see if a
    24,900 7.5 120 57 Using my numbers, I obtained that the only significant
    22,700 9.1 30 44 Therefore, this model can not be used to obtain accur

    Ok… So, what should we do if both models are not ap


    Well, since in both cases the only significant variable w
    Details about this linear regression model, also called
    can be viewed by accesing Reduced Model sheet.
    Using this final model, we are able to make both requ
    plate contains different numbers from the ones you will have in your dataset.
    multiple regression models: the first one has two independent variables and the second has three independent variables.

    e regression model with 2 independent variables


    Analysis>Regression>Select the area for the dependent variable (advertising sales) and the area for the independent
    of viewers and length of program)>OK.
    w sheet (Model A). Now, each independent variable has a different slope, but are these slopes significant for our model?
    pe for Average # of viewers is 744.02 and it is significant (explain why? Hint: look at your p-value),
    gth of Program is not significant (again explain this using the p-value).
    at this regression model is not useful for predictions since the only significant variable is Average # of Viewers.

    e regression model with 3 independent variables


    t are indicated above and see if all the three variables are significant (Model B sheet).
    obtained that the only significant variable is Average # of viewers again.
    el can not be used to obtain accurate predictions.

    d we do if both models are not appropriate for our scenario?


    ases the only significant variable was Average # of viewers, we should build a linear regression model like we did last week.
    ear regression model, also called reduced model since the number of independent variables was reduced,
    cesing Reduced Model sheet.
    el, we are able to make both required predictions and we are done with the assignment.
    nt variables.

    our model?

    did last week.


    SUMMARY OUTPUT

    Regression Statistics
    Multiple R 0.77585941109415
    R Square 0.60195782578336
    Adjusted R Squ 0.53561746341391
    Standard Error 3228.20678872946
    Observations 15

    ANOVA
    df SS MS F Significance F
    Regression 2 189121504.484 94560752 9.07377959 0.003977174
    Residual 12 125055828.85 10421319
    Total 14 314177333.333

    Coefficients Standard Error t Stat P-value Lower 95% Upper 95%


    Intercept 16864.5034406637 2315.54411523 7.283171 9.6968E-06 11819.36621 21909.6407
    X Variable 1 744.026083995749 176.648628881 4.211898 0.00120613 359.141785 1128.91038
    X Variable 2 -8.4006474583954 23.7267211962 -0.354059 0.72943828 -60.096732 43.2954371
    Lower 95.0% Upper 95.0%
    11819.36621 21909.64067
    359.141785 1128.910383
    -60.096732 43.29543709
    SUMMARY OUTPUT

    Regression Statistics
    Multiple R 0.822766908
    R Square 0.676945385
    Adjusted R Sq 0.588839581
    Standard Error 3037.589316
    Observations 15

    ANOVA
    df SS MS F Significance F
    Regression 3 212680896 70893632 7.6833234 0.004809135
    Residual 11 101496437.37 9226949
    Total 14 314177333.33

    Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95.0%
    Intercept 14726.22723 2556.9393755 5.759318 0.00012662 9098.441608 20354.01285 9098.441608
    X Variable 1 702.9688961 168.19218103 4.179558 0.0015377 332.7804016 1073.157391 332.7804016
    X Variable 2 -23.98865406 24.363953732 -0.984596 0.34598388 -77.6133547 29.63604655 -77.6133547
    X Variable 3 121.9040616 76.289575446 1.597912 0.13836829 -46.0081618 289.8162851 -46.0081618
    Upper 95.0%
    20354.012849
    1073.1573906
    29.636046547
    289.81628507
    SUMMARY OUTPUT

    Regression Statistics
    Multiple R 0.7731750754
    R Square 0.5977996973
    Adjusted R S 0.5668612124
    Standard Erro3117.7188231
    Observations 15

    ANOVA
    df SS MS F Significance F
    Regression 1 187815114.8 1.88E+08 19.32220342 0.000723364
    Residual 13 126362218.6 9720171
    Total 14 314177333.3

    Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95.0%
    Intercept 16484.875009 1982.094766 8.316895 1.45777E-06 12202.8196 20766.93042 12202.819601
    X Variable 1 729.5471892 165.9682687 4.395703 0.000723364 370.9945436 1088.099835 370.9945436

    The regression equation is: y=16484.875+729.547*x and it can be used to make both predictions.
    First prediction: y=16484.875+729.547*3.2 18819.42601
    Second prediction: y=16484.875+729.547*14.1 26771.49038
    Upper 95.0%
    20766.930416
    1088.09983481

    You might also like