EC5203 Assignment - Final

8/14/2019 EC5203 Assignment - Final

http://slidepdf.com/reader/full/ec5203-assignment-final 1/21

EC5203 Assignment

Opening Case and Case Study 2

Name: Shannon Frick

Student ID: 11159183

Subject Name: Statistics for Business

Subject Code: EC5203

Subject Coordinator: Shane Zhang

Due Date: 16th June, 2009



Table of Contents

Table of Contents .................................................................................................................2Opening Case .......................................................................................................................3

Strong Predictor Variables of Gross Box Office Takings ...............................................4Other Potentially Relevant Variables ...............................................................................5

Results ..............................................................................................................................5

Developed Mathematical Model .......................................................................................................................................... 7

Strength of the Model ......................................................................................................7

The Need for Multiple Regression ...................................................................................8Case Study 2 ........................................................................................................................9

1.Multiple Regression Model: Admissions ..........................................................................9

Introduction ....................................................................................................................10Results ............................................................................................................................10Coefficient of Multiple Determination: r2 and Adjusted r2 ...........................................12

Linear Equation for Multiple Regression ......................................................................12

Significance of Dependent Variables to the Multiple Linear Regression ......................13Conclusion: Implications of Analysis ............................................................................14

2. Other Relevant Variables ...............................................................................................15

3. Multiple Regression Model: Movies Seen .....................................................................16Introduction ....................................................................................................................17

Results

........................................................................................................................................ 17

Coefficient of Multiple Determination: r2 and Adjusted r2 ...........................................19Linear Equation for Multiple Regression ......................................................................19

Significance of Dependent Variables to the Multiple Linear Regression ......................20

Conclusion: Marketing Implications of Analysis ..........................................................21

2



Opening Case

The following data is provided by the Motion Picture Distributors Association of Australia, and relates to the gross box office takings of the yearly top movies in Australia from 1986 to 2004. There is also information relating to total cinemaattendance in these years, the total number of screens (the most popular filmsare often shown in two or more screens simultaneously in some of the larger multi-screen cinemas) and total number of films shown in the year (this could act as a measure of competition for the leading movies).

3



1. Dependent and Independent Variables

Several variables are presented that may be related to the yearly gross box

office takings of the top Australian film. Which variables are stronger predictorsof gross box office takings? Might other variables not mentioned here be related to gross box office takings?

Strong Predictor Variables of Gross Box Office Takings

From observation of the graph above, Total Australian Admissions and Screenings, there are several assumptions that can be made. First, it is evidentthat the variable “Year” is a label for each year, and is most likely not particularly

useful for making future predictions regarding gross box office takings. Secondly,the title of the movie shown in the Number 1 film column in this table can also beassumed to be irrelevant for future predictions involving gross box office takings.With these two columns excluded for statistical purposes, the remaining variablesthat will be looked at for this part will be “Number of Admissions (millions),”“Number of Films Screened,” “Number of Screens,” and “Box Office ($ million).”

The dependent variable in this section is “Box Office (millions),” since thisis the variable around which relationships are being questioned. This deems“Number of Admissions,” “Number of Films Screened,” and “Number of Screens”as dependent variables for this case.

The independent variables’ strength in predicting the gross Box Officetakings for a given year can be found when performing a multiple linear regression. For this assignment, an Excel add-in tool for the multiple linear regression calculations called PHStat has been used. The PHStat tool greatlyreduces the time and effort required to perform a multiple linear regressionanalysis, along with providing ANOVA and residual error calculations. The outputfor the multiple linear regression modeling is given below in the section titled“Mathematical Model to Predict Gross Box Office Takings.”

While the multiple linear regression analysis in subsequent paragraphs

below will show how weak the relationship is between the dependent andindependent variables, it should be noted that the strongest relationship is thatbetween “Number of Admissions” and “Box Office (millions).” This seems tomake sense, since the more viewers are admitted to a movie, the more paymenti.e. gross box office takings the cinemas will receive.

4



Other Potentially Relevant Variables

Apart from the given independent variables of “Number of Admissions,”“Number of Films Screened,” and “Number of Screens,” there could be a number of other unmentioned variables that might affect the Gross Box Office Takings.

Some of these variables are included in Part 2 of this assignment, such as:

o “Theatres”o “Top Price of Cinema Tickets”o “Capacity”o “Age”o “No. of Emailed Discount Cinema Tickets Received Last Year”o “Income (‘000s)”

Other perhaps important variables that have been omitted from the data,which are also mentioned in Part 2 of the assignment, may be:

o Screening Times During the Day o Screening Times During the Year o Location of Cinema

o Gender, Nationality, Religion or Political View o Genre of the Film (Horror/Comedy/etc) o Amount ($) Spent on Advertising for the Film

2. Mathematical Model to Predict Gross Box Office

Takings

Is it possible to develop a mathematical model to predict gross box office takingsusing the data given? If so, how strong is the model? With three independent variables, will we need to develop three different simple regression models and compare their results?

Results

Regression Analysis

Regression Statistics

Multiple R 0.644693511R Square 0.415629723Adjusted R Square 0.298755668Standard Error 8.884734142Observations 19

5



ANOVA

df SS MS F Significance

F

Regression 3.00000 842.16776 280.72259 3.55622 0.04006Residual 15.00000 1184.07751 78.93850

Total 18.00000 2026.24527

CoefficientsStandard

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

Intercept -5.71881 25.95778 -0.22031 0.82860 -61.04651 49.60889Number of Admissions(millions) -0.10880 0.36397 -0.29894 0.76909 -0.88459 0.66698Number of filmsscreened 0.07348 0.10052 0.73095 0.47608 -0.14078 0.28773Number of Screens 0.01799 0.01692 1.06357 0.30435 -0.01807 0.05405

RESIDUAL OUTPUT

Observation Predicted Box Office ($ million) Residuals

1 20.14328 20.656722 21.56637 -10.166373 23.67032 1.229684 25.31063 -9.520635 23.43136 2.668646 22.73710 -3.517107 22.27410 -3.54410

8 24.38503 7.354979 23.88529 0.76471

10 25.72425 -7.7942511 29.39766 -0.1076612 32.54003 -9.8400313 33.99416 13.3358414 34.89610 3.8239015 36.40136 -5.6813616 35.59706 -3.5470617 36.85815 -3.0481518 38.51647 -1.3764719 42.04129 8.30871

6



Developed Mathematical Model

Using multiple linear regression analysis, it is possible to develop a

mathematical model to predict gross box office takings using the data tableabove. The results of a multiple linear regression using PHStat with thementioned dependent and independent variables are given below under theheading “Results.”

From the given results (above under the heading Results), we can form amultiple linear regression model based on the coefficients from the results, whichcan be used to create a mathematical formula that could predict potential futurevalues for gross box office takings.

The form of the multiple linear regression equation will be:

Ŷ = b0 + b1X1+ b2X2 + b3X3

where:

Ŷ = Box Office ($ millions)b0 = Intercept (i.e. when Y = 0)X1 = Number of Admissions (millions)b1 = rate of increase of Box Office ($ million) given Number of Admissions

(millions)X2 = Number of Films Screened

b2 = rate of increase of Box Office ($ million) given Number of FilmsScreenedX3 = Number of ScreensB3 = rate of increase of Box Office ($ million) given Number of Screens

Therefore, the final equation for predicting gross box office takings is:

Box Office ($ million) = -5.7188 - 0.1088(Number of Admissions(millions)) +0.0735 (Number of Films Screened) + 0.01799(Number of Screens)

Strength of the Model

The r 2 value (“r square”) of roughly 0.41563 indicates that approximately41.56% of the variation in the dependent variable “Box Office ($ million)” can beexplained by the independent variables used, “Number of Admissions,” “Number of Films Screened,” and “Number of Screens.” This is a weak linear relationship

7



that suggests there may not be such a strong relationship between thedependent and independent variables.

The adjusted r 2 value of 0.29875 suggests that approximately 29.88% of “Box Office ($ million)” can be explained by all the aforementioned dependent

variables, while also taking into account the sample size and number of independent variables. This further substantiates the notion that there is a veryweak linear relationship between the dependent and independent variables.

The Need for Multiple Regression

As shown above, we have not needed three different simple regressionmodels to compare their results, since multiple linear regression has done the jobquite adequately for the three independent variables. In fact, it is preferable to

perform multiple linear regression analysis instead of performing severalinstances of single linear regression analysis due to the fact that multipleregression takes into account relationships between independent variables andthe dependent variable at the same time, whereas three separate single linear regression models would not take into account the effect that each relationshipwould have on the other when calculating relationships with the dependent andindependent variables.

8



Case Study 2

1. Multiple Regression Model: Admissions

The following data corresponds to the annual number of admissions (inmillions) to the movies in Australia over the years 1984 to 2004. It alsocontains the yearly totals of the numbers of screens, theatres and filmsscreened, as well as the top price paid for a cinema ticket. Using these data,develop a multiple regression model to study how well the number of annual cinema admissions can be explained by the other variables. Which variablesseem to be more promising predictors? What implications for film distributor executives might be evident from this analysis?

9



Introduction

For this part of the assignment, an Excel add-in tool for the multiple linear

regression calculations called PHStat was used. This tool greatly reduces thetime and effort required to perform a multiple linear regression analysis. Theoutput for the multiple linear regression modeling is shown below.

In order to accurately perform the regression analysis, it must be confirmedthat the regression modeling is accurate. In doing so, the following assumptionswill be made:

“Year” is not a variable, but rather, a label for each observation. The dependant variable in this analysis is “No. of admissions (millions).” The independent variables in this analysis are “Year,” “No. of Screens,”

“Theatres,” “No. of Films Screened,” “Top Price of Cinema Tickets,” and“Capacity (000 seats).”

Residuals in the model are normally distributed, have constant variance,and are independent.

We will use a confidence level for regression coefficients of 95%, such thatα = 0.05.

Results

The results for the multiple linear regression analysis are given below:



ANOVAdf SS MS F Significance F

Regression 5 10862.0855 2172.417099 114.2685399 2.1558E-11Residual 15 285.1725988 19.01150659Total 20 11147.2581

CoefficientsStandard

Error t Stat P-valueLower 95%

Upper 95%

Intercept 55.1391 32.2236 1.7111 0.1076 -13.5439 123.8222No of screens 0.1199 0.0226 5.3155 0.0001 0.0718 0.1680Theatres 0.0725 0.0349 2.0750 0.0556 -0.0020 0.1469No of filmsscreened 0.0474 0.0491 0.9650 0.3498 -0.0573 0.1522Top price of cinema ticket ($) -2.5388 2.0464 -1.2406 0.2338 -6.9007 1.8230

10



Capacity (000seats) -0.4447 0.1080 -4.1164 0.0009 -0.6750 -0.2144

Observation Predicted No. of admissions (millions) Residuals

1 29.7170 -0.81702 31.7007 -2.0007

3 32.9106 2.58944 25.0980 5.70205 40.6546 -3.25466 43.9614 -4.96147 46.9636 -3.96368 47.4028 -0.50289 50.7525 -3.5525

10 56.3712 -0.871211 61.3113 6.788712 65.7125 4.187513 68.3302 5.569814 76.1044 -0.1044

15 82.7255 -2.725516 88.7593 -0.759317 89.2581 -7.058118 88.6222 3.877819 89.5627 2.937320 90.2856 -0.485621 92.0960 -0.5960

Confidence Interval Estimate and Prediction Interval

t Statistic 2.13145Predicted Y (YHat) 55.13912

For Average Predicted Y (YHat)

Interval Half Width 68.68304

Confidence Interval Lower Limit -13.5439

Confidence Interval Upper Limit 123.8222

For Individual Response Y


Prediction Interval Lower Limit -14.1698

Prediction Interval Upper Limit 124.4481

The given Regression Analysis above presents several key statistics thatare useful for further analysis. These include regression statistics such as r 2,ANOVA (analysis of variance) statistics, and confidence interval estimates.These are discussed below in more detail.

11



Coefficient of Multiple Determination: r 2 and Adjusted r 2

The r 2 value (“r square”) of roughly 0.97442 indicates that approximately

97.44% of the variation in the dependent variable “Number of Admissions” can beexplained by the independent variables used, “No. of Screens,” “Theatres,” “No.of Films Screened,” “Top Price of Cinema Tickets,” and “Capacity (000 seats).”

The adjusted r 2 value of 0.965890255 suggests that approximately 96.59%of “Number of Admissions” can be explained by all the aforementioneddependent variables, while also taking into account the sample size and number of independent variables.

Linear Equation for Multiple Regression

From the given linear regression analysis above, we can form the multipleregression equation for the Number of Admissions in the audience demand for cinema films:

The form of the equation will be:

Ŷ = b0 + b1X1+ b2X2 + b3X3 + b4X4+ b5X5+ b6X6

where:

Ŷ = Number of admissions (millions)b0 = Intercept (i.e. when Y = 0)X1 = Number of Screensb1 = rate of increase of Number of Admissions given No. of ScreensX2 = Theatresb2 = rate of incrase of Number of Admissions given TheatresX3 = Number of Films ScreenedB3 = rate of increase of Number of Admissions given Number of

Films ScreenedX4 = Top Price for Cinema Tickets

b4 = rate of increase of Number of Admissions given Top Price for Cinema TicketsX5 = Capacity (‘000s)b5 = rate of increase of Number of Admissions given Capacity.

Therefore, the final equation is as below:

12



Number of admissions (millions) = 5.1391 + 0.1199(Number of Screens) +0.0725 (Theatres) + 0.0474(Number of Films Screened) – 2.5388(Top Price for Cinema Tickets ($)) – 0.4447(Capacity (000s))

The equation above shows by how much the Number of Admissionsincreases by the following amount with an increment of 1 in each of thedependent variables:

o 119,900 admissions for each unit increase in “Number of Screens.”o 80,000 admissions for each unit increase in “Theatres.”o 25,500 admissions for each unit increase in “No. of Films Screened.”o - 7,334,100 admissions for unit increase in “Top Price of Cinema Tickets.”o - 283 admissions for each unit increase in “Capacity.”

Significance of Dependent Variables to the Multiple Linear Regression

The t-Test can be used to determine if there is a linear relationshipbetween the dependent variables considered with Y. In order to test for significance, the null hypothesis shall be formed, which states that there is norelationship among independent variables in determining the dependent variable,“Number of Admissions.” In other words:

H0: b1 = b2 = b3 = b4 = b5 = 0 (no linear relationship)H1: at least one b value above ≠ 0 (at least one independent variable

affects Ŷ)

With df = 5 (degrees of freedom) and α = 0.05, the critical value of t(α/2, 5) =2.5706. This shows that the confidence interval is (-2.5706, 2.5706), andtherefore any t-test statistic values for the dependent variables falling outside of this region will be rejected, proving that the dependent variable affects Ŷ. Each t-value for each dependent variable can then be examined to see if they affect Ŷusing this approach:

13



Dependent Variable t-test Statistic Within (-2.5706,2.5706)?

Reject H0?

X1: No. of Screens 5.3155 No Yes

X2: Theatres 2.0750 Yes NoX3: No. of Films Screened 0.9650 Yes NoX4: Top Price of CinemaTickets -1.2406

Yes No

X5: Capacity (000 seats) -4.1164 No Yes

From the above results, it can be said that one can be 95% confident thatthe independent variables “No. of Screens” and “Capacity” affect the dependentvariable “Number of Admissions.” The other independent variables, “Theatres,”“No. of Films Screened,” and “Top Price of Cinema Tickets” cannot be proven assuch to affect the dependent variable.

Conclusion: Implications of Analysis

The evident implications for film distributor executives that arise from thisanalysis are the “Number of Screens” and “Capacity” of theatres; these aresignificant variables when the intention of increasing the number of admissions(and therefore profit) is involved. If executives wish to increase the number of admissions in their theatres and increase profits, then they may first wish toincrease the number of screens and capacity of their theatres, based on theresults provided.

It should be noted, however, that even though it could not be proven withcertainty that an increase in “Theatres” and “No. of Films Screened” gave adefinite rise in admissions, it could be seen that there is a positive relationship onaverage between these variables and the number of admissions. It would be of interest to film distributor executives to also take into account this finding.Increasing the number of theatres and number of films being screened might alsohave a positive effect on admissions, albeit on a smaller scale. Of course, the

14



decision of which dependent variable to leverage would no doubt depend onbusiness factors such as the company’s budget, among other things.

2. Other Relevant Variables

Think about some other variables, not included above, that may also beimportant predictors in determining the annual numbers of movie-goers.

The given variables in the linear regression analysis may not be the mostexhaustive list of relevant variables in the equation. Other variables that mayhave been important predictors in determining the annual numbers of movie-goers include the following:

o Screening Times During the Day may be relevant because the audience mayhave a particular time during the day that viewing movies is most favourablei.e. at night, rather than in the afternoons when many people may be working.

o Screening Times During the Year is also a potentially strong indicator in thisanalysis, due the fact that people may be unable to attend, or unwilling toattend, cinemas during certain parts of the year.

o Location of Cinema could be important for the purpose of convenience for viewers to attend. The audience members would not doubt be more willing to

travel less distance to attend a movie rather than spend a long time travellingto the cinema to watch a particular movie.

o Certain demographic information (other than age, which is covered below)such as Gender, Nationality, Religion or Political View of the audience mayalso be prevalent factors in distinguishing movie-going trends. Someone whois devoutly religious, for example, may be unwilling to watch moviesaltogether. Similarly, women may be less willing to watch movies than men,as another example.

o The Genre of the Film (Horror/Comedy/etc) could also be another telling

factor in the search for more relationships among variables. Horror flicks, for example, may be more popular than say Romance or Western movies.

o Lastly, the Amount ($) Spent on Advertising for the Film may play a big part inthe reason for why admissions may increase. If one film is given a lot of advertising, then it would be expected that it’s presence would reach morepeople, and more people would therefore know about it and/or attend thecinema to view it.

15



3. Multiple Regression Model: Movies Seen

Some film distributors offer discount cinema tickets via email. Suppose that a

random sample of 25 movie-goers is undertaken, and suppose that thenumber of movies the person has seen in the last year, their age and income,and the number of discount cinema tickets they have received via email in thelast year, is recorded. Use the data to develop a multiple regression model to

predict the number of times an individual goes to the movies per year fromtheir age, income and number of discount cinema tickets received. Which

particular independent variables seem to have more promise in predicting thenumber of times a person goes to the movies? What marketing implicationsmight be evident from this analysis?

16



Introduction

In order to accurately perform the regression analysis and makepredictions for this model, it must be confirmed that the regression modeling isaccurate. In doing so, the following assumptions will be made:

The dependant variable in this analysis is “No. of Movies Seen at CinemaLast Year.”

The independent variables in this analysis are “Age,” “No. of EmailedDiscount Cinema Tickets Received Last Year,” and “Income (‘000s).”

Residuals in the model are normally distributed, have constant variance,and are independent.

We will use a confidence level for regression coefficients of 95%, such thatα = 0.05.

Results

The results for the multiple linear regression analysis are given below:



ANOVAdf SS MS F Significance F

Regression 3.0000 171.2060 57.0687 3.9063 0.0231Residual 21.0000 306.7940 14.6092Total 24.0000 478.0000

Coefficients

Standard Error t Stat

P-value

Lower 95%

Upper 95%

Intercept 5.4654 3.57281.529

70.141

0 -1.9646 12.8954

Age -0.0521 0.0974-0.534

90.598

4 -0.2545 0.1504No of discount tickets

received 1.1775 0.3645

3.230

5

0.004

0 0.4195 1.9355

Income ($000) 0.0383 0.05930.645

80.525

4 -0.0851 0.1617

17



RESIDUAL OUTPUTObservatio

n Predicted No of movies seen Residuals

1 6.1075 -2.10752 11.1321 0.8679

3 9.9192 0.08084 10.3155 -2.31555 14.3334 -3.33346 12.3506 -0.35067 7.5439 0.45618 10.3509 -4.35099 12.9997 3.0003

10 6.2117 3.788311 10.8029 7.197112 8.7083 3.291713 9.4561 -8.456114 7.2517 4.7483

15 12.3359 2.664116 5.8659 -2.865917 13.5710 -3.571018 10.0921 -2.092119 11.3404 3.659620 16.9863 2.013721 8.9333 3.066722 12.3617 1.638323 10.1295 -0.129524 9.5435 -3.543525 11.3570 -3.3570

Confidence and Prediction Estimate Intervals

t Statistic 2.079614Predicted Y (YHat) 5.46538

For Average Predicted Y (YHat)


Confidence Interval Lower Limit -1.96462

Confidence Interval Upper Limit 12.89538

For Individual Response Y


Prediction Interval Lower Limit -5.41519

Prediction Interval Upper Limit 16.34595

As mentioned previously, the given regression analysis above presentsseveral key statistics that are useful for further analysis. These includeregression statistics such as r 2, ANOVA (analysis of variance) statistics, andconfidence interval estimates. These are discussed below in more detail.

18



Coefficient of Multiple Determination: r 2 and Adjusted r 2

The r 2 value (“r square”) of roughly 0.3582 indicates that approximately35.82% of the variation in the dependent variable “No. of Movies Seen at CinemaLast Year” can be explained by the independent variables used, “Age,” “No. of

Emailed Discount Cinema Tickets Received Last Year,” and “Income (‘000s).”

The adjusted r 2 value of 0.2665 suggests that approximately 26.65% of “No. of Movies Seen at Cinema Last Year” can be explained by all theaforementioned dependent variables, while also taking into account the samplesize and number of independent variables.

Linear Equation for Multiple Regression

From the given linear regression analysis above, we can form the multipleregression equation for the Number of Admissions in the audience demand for cinema films:

The form of the multiple linear regression equation will be:

Ŷ = b0 + b1X1+ b2X2 + b3X3

where:

Ŷ = Number of Movies Seen at Cinema Last Year

b0 = Intercept (i.e. when Y = 0)X1 = Ageb1 = rate of increase of No. of Movies Seen at Cinema Last Year given

AgeX2 = No. of Emailed Discount Cinema Tickets Received Last Year b2 = rate of increase of No. of Movies Seen at Cinema Last Year given No.

of Emailed Discount Cinema Tickets Received Last Year X3 = Income (‘000s)b3 = rate of increase of No. of Movies Seen at Cinema Last Year given

Income (‘000s)

Therefore, the final equation is as below:

Number of Movies Seen Last Year = 5.4654 - 0.0521(Age) + 1.1775(No. of Emailed Discount Cinema Tickets Received Last Year) + 0.0383(Income (‘000s))

19



The equation above shows by how much the Number of Movies Seen LastYear increases by the following amount with an increment of 1 in each of thedependent variables:

o - 0.0521 movies seen last year for each unit increase in “Age”

o 1.1775 movies seen last year for each unit increase in “No. of EmailedDiscount Cinema Tickets Received Last Year.”o 0.0383 movies seen last year for each unit increase in “Income (‘000s).”

Significance of Dependent Variables to the Multiple Linear Regression

The t-Test can be used to determine if there is a linear relationshipbetween the dependent variables considered with Ŷ. In order to test for

significance, the null hypothesis shall be formed, which states that there is norelationship among independent variables in determining the dependent variable,“Number of Movies Seen Last Year.” In other words:

H0: b1 = b2 = b3 = b4 = b5 = 0 (no linear relationship)H1: at least one b value above ≠ 0 (at least one independent variable

affects Ŷ)

With df = 3 (degrees of freedom) and α = 0.05, the critical value of t(α/2, 3) =3.1824. This shows that the confidence interval is (-3.1824, 3.1824), andtherefore any t-test statistic values for the dependent variables falling outside of

this region will be rejected, proving that the dependent variable affects Ŷ. Each t-value for each dependent variable can then be examined to see if they affect Ŷusing this approach:

20



Dependent Variable t-test Statistic Within(-3.1824, 3.1824)?

Reject H0?

X1: Age -0.5349 Yes NoX2: No. of Emailed DiscountCinema Tickets Received Last

Year 3.2305

No Yes

X3: Income (‘000s) 0.6458 Yes No

From the above results, it can be said that there is 95% confidence thatthe independent variable “No. of Emailed Discount Cinema Tickets ReceivedLast Year” affects the dependent variable “Number of Movies Seen Last Year.”The other independent variables, “Age” and “Income (‘000s),” can not be provenas such to affect the dependent variables.

Conclusion: Marketing Implications of Analysis

The evident implications for marketing professionals given the aboveanalysis is that the number of emailed discount cinema tickets received last year had a direct impact on the number of movies a person saw last year. If marketers wish to increase the number of movies that people see in future (andtherefore increase profits for the companies they work for), then it would be intheir best interests to either continue emailing discount cinema tickets tocustomers, or perhaps even increase the amount of such emailed tickets, basedon the results provided.

While less obvious to prove given the results of the t-Test, it should also

be taken into account the other factors of the given multiple linear equation -“Age” and “Income.” It could be assumed that the older one gets (i.e. the higher the “Age” value), the lower the value of movies being seen in a year becomes.Conversely, the more a person is paid (i.e. the higher the “Income (‘000s)” value),the higher the value of movies being seen may become, based on our results.Both of these assumptions, while not proven statistically, certainly make sense.Marketing representatives could use this knowledge to target both younger andwealthier populations to try and increase the number of movies being seen in anysubsequent year as an initiative to boost company profits.

EC5203 Assignment - Final

Documents