Exploratory Data Analysis
mezz
TELECOMMUNICATIONCUSTOMERSEXPLORATORYANALYSISASSIGNMENT1.docx
TELECOMMUNICATION CUSTOMERS EXPLORATORY ANALYSIS ASSIGNMENT
PART I:
Introduction
In this research analysis we have selected a telecommunication company dataset. The data summarizes characteristics or behaviours of different customers engaging in business with the telecommunication company.
The main aim of selecting this dataset was to help coming up with a suitable module that will help the company to evaluate metrics on whether to retain or dispense a customer.
The dataset contains 21 variables and 7043 valid observations. Rows represents different customers whereas columns represent every attribute of the customer.
customerID-Represents unique identities of the customers.
Gender-indicate whether the customer is a female or a male
SeniorCitizen-1 represent senior citizen whereas 0 represent non senior citizen
Partner-Yes means the customer has a partner, No-means he does not have a partner
Dependents-Yes means he has dependents while No means the customer do not have dependents
Tenure-Indicates the period the customer has done business with the company in months
PhoneService-Yes mean the customer has the service while No means otherwise.
MultipleLines-Yes means the customer has multiple line, No and no service means otherwise
InternetService-shows the provider of the internet service to the customer.
OnlineSecurity-Shows whether the customer has the online security or otherwise.
OnlineBackup- Shows whether the customer has the online back up or otherwise.
DeviceProtection- Shows whether the customer has the device protection or otherwise (No/no int. service).
TechSupport- Shows whether the customer has the technical support (Yes) or otherwise (No/no int. service).
StreamingTV- Shows whether the customer engage in TV streaming (Yes) or otherwise (No
StreamingMovie- Shows whether the customer engage in movie streaming (Yes) or otherwise (No
Contract-shows the customers’ subscribed contract type
PaperlessBilling- Shows whether the customer engage in paperless billing (Yes) or otherwise (No
Payment Method- shows the customers’ method of payment
MonthlyCharges-represent total monthly fees levied to the customers
TotalCharges-represents cumulative amount charged to the customer
Churn- Shows whether the customer is churned (Yes) or otherwise (NO
Summary of GendeR
As seen in the above chart the survey engaged 50% of every gender.
Monthly Charge
![]()
Payment Method
![]()
RESEARCH QUESTIONS
Questions that can be answered by this survey.
i) Is there a relationship between internet providers and monthly charge to the customers?
ii) Is there a relationship between online security and customers device protection?
iii) How does streaming of movies by the customers affect their total charges to the company?
Question 1
Is there a relationship between internet providers and monthly charge to the customers?
Internet as one of the pivot aspects of connectivity between individuals and businesses in the modern world, it has attracted various entities to engage in its business. This is a result of high revenue influx attributed with the business. Most internet providers across the world are improving all aspects relating to internet diversity to remain competitive in the market. This includes increasing the speed, enhancing high security firewalls especially from internet phishers, cost analysis to its subscribers, among others.
Globally, according to latest journal in US regarding internet service providers and their monthly charges to their subscribers, Google fiber emerged to be the highest monthly charging internet service provider amounting to at least $ 70.00 per month. Verizon, Grande communication and Xfinity followed the list with monthly charge of at least $ 39.99, $35.46, and $ 20.00, respectively.
The main importance to include this hypothetical question in our study was to evaluate and establish which among the service providers i.e. DSL and Fibre Optic had the highest monthly charges in comparison with the other.
Question 2
Is there a relationship between online security and customers device protection?
As internet continue to spread rapidly globally, online security has become a major threat to the users. Businesses has formulated easier method using devices such as mobile phones, debit cards, swift code transfers to ease their operations in terms of payments and receipts. This has posed a great concern on security aspect as hackers try to generate phishy links and duplicates to compromise security of the devices. With the alarming online business increasing rapidly, Customer transacting devices security is at risk.
This led us to try evaluating whether online security and customers device protection were in one way or the other associated.
Question 3
How does streaming of movies by the customers affect their total charges to the company?
Subscribers of internet engage in various activities ranging from commercial based to social-based. These activities include watching and playing games, emailing, cyber services, video streaming, among others. These activities accrue to different monthly charges.
In our case study we choose to evaluate among the group which used the internet for Movies streaming and the group which never used the internet for movie streaming, which group of customers had the higher monthly charges as compared to the other.
VARIABLES
Is there a relationship between internet providers and monthly charge to the customers?
The variables that will be used to answer this hypothesis will be
-Internet providers and Monthly Charge.
Is there a relationship between online security and customers device protection?
The variables that will be used to answer this hypothesis will be
Online Security and Device protection
How does streaming of movies by the customers affect their total charges to the company?
The variables that will be used to answer this hypothesis will be
Streaming Movies and total charges.
SAMPLE OUTCOMES
i) There exists a relationship between internet provider and their monthly charge to their customers
ii) Online security and device protection are not associated
iii) Movie streaming increases total charges a customer owes to the company.
DATA PRESENTATION AND ANALYSIS
i) To identify the columns used in this study
telecomcustomers$InternetServicetelecomcustomers$OnlineSecuritytelecomcustomers$DeviceProtectiontelecomcustomers$StreamingMoviestelecomcustomers$MonthlyChargestelecomcustomers$TotalCharges
ii) R-script to extract the subset of the data needed in this study
> telecomcustomers%>%+ select(InternetService,OnlineSecurity,DeviceProtection,Stre amingMovies,MonthlyCharges,TotalCharges)InternetService OnlineSecurity DeviceProtection StreamingMovies MonthlyCharges1 DSL No No No 29.852 DSL Yes Yes No 56.953 DSL Yes No No 53.854 DSL Yes Yes No 42.305 Fiber optic No No No 70.706 Fiber optic No Yes Yes 99.657 Fiber optic No No No 89.108 DSL Yes No No 29.759 Fiber optic No Yes Yes 104.8010 DSL Yes No No 56.1511 DSL Yes No No 49.9512 No No internet service No internet service No internet service 18.9513 Fiber optic No Yes Yes 100.3514 Fiber optic No Yes Yes 103.7015 Fiber optic Yes Yes Yes 105.5016 Fiber optic Yes Yes Yes 113.2517 No No internet service No internet service No internet service 20.6518 Fiber optic Yes Yes Yes 106.7019 DSL No Yes No 55.2020 Fiber optic No Yes Yes 90.0521 DSL No Yes Yes 39.6522 No No internet service No internet service No internet service 19.8023 No No internet service No internet service No internet service 20.1524 DSL No No No 59.90TotalCharges1 29.852 1889.503 108.154 1840.755 151.656 820.507 1949.408 301.909 3046.0510 3487.9511 587.4512 326.8013 5681.1014 5036.3015 2686.0516 7895.1517 1022.9518 7382.2519 528.3520 1862.9021 39.6522 202.2523 20.1524 3505.10
iii) R-script to convert selected categorical variables into numerical variables in this study
mydata=mutate(telecomcustomers,InternetSer=as.numeric(InternetService),OnlineSec=as.numeric(OnlineSecurity),DeviceProtec=as.numeric(DeviceProtection),StrMovies=as.numeric(StreamingMovies),Mcharges=as.numeric(MonthlyCharges),TCharges=as.numeric(TotalCharges))> mydata
R-script to select our subset data for analysis
> newdata=mydata[c("InternetSer","OnlineSec","DeviceProtec","Mcharges","TCharges")]> newdata
|
head(newdata) InternetSer OnlineSec DeviceProtec Mcharges TCharges 1 1 1 1 29.85 29.85 2 1 3 3 56.95 1889.50 3 1 3 1 53.85 108.15 4 1 3 3 42.30 1840.75 5 2 1 1 70.70 151.65 6 2 1 3 99.65 820.50
|
|
|
|
|
|
Is there a relationship between internet providers and monthly charge to the customers?
ggplot(newdata, aes(x = Mcharges)) +
+ geom_histogram(fill = "cornflowerblue",
+ color = "white") +
+ facet_wrap(~InternetSer, ncol = 1) +
+ labs(title = "Monthly Charges histograms by Internet service provider")
Is there a relationship between online security and customers device protection?
ggplot(newdata,
+ aes(x = OnlineSec,
+ fill = DeviceProtec)) +
+ geom_bar(position = position_dodge(preserve = "single"))
How does streaming of movies by the customers affect their total charges to the company?
ggplot(mydata, aes(x = TCharges)) +
+ geom_histogram(fill = "cornflowerblue",
+ color = "white") +
+ facet_wrap(~StrMovies, ncol = 1) +
+ labs(title = "Total Charges histograms by Streaming Movies")
![]()
REPORTING
The main aim of this study was to establish a module to be used by the telecommunication company to retain its customers based on various attributes.
The main research questions were:
a) To establish whether internet service providers were associated in any way with Monthly charges levied to the customers.
b) To analyse whether a customer possessing a protected device had a guaranteed advantage of Online security
c) The final aim was to evaluate how usage of the internet in movie streaming affected total monthly charges levied to the customers by the telecommunication company.
i) Is there a relationship between internet providers and monthly charge to the customers?
From the chart above titled Monthly Charges histograms by Internet service provider, 1 represented Digital subscriber loop (DSL) internet service providers, 2 represented Fiber optic service providers while indicated neither of the two internet providers.
As seen from the chart the two variables are associated. Digital subscriber loop (DSL) has a larger number of customers as compared to Fibre optic service providers. This is because Fibre optic service providers had a higher Monthly Charges of at most $ 300.00 as compared to DSL whose approximate monthly charges were $ 270.00.
Choosing Fibre optic as internet service provider would subject the customer to a higher Monthly charge budget by the telecommunication company.
ii) Is there a relationship between online security and customers device protection?
From the figure above on this hypothetical question, 1 showed that the device was internet protected and showed online security, 2 represented No online security whereas 3 was a control factor.
We can see that customers whose devices were protected showed a higher count as the same illustrated higher online security. This would mean that the two variables were highly associated.
Choosing a device with protection would guarantee a customer online security safety.
iii) How does streaming of movies by the customers affect their total charges to the company?
From the chart above titled Total Charges histograms by Streaming Movies, 1 represented subscribers who used the internet to stream movies, 2 represented subscribers who did not use the internet in streaming movies, 3 represent the control group who never had internet services.
The chart indicates that group 1 who used the internet for movie streaming had a higher total charge owing to the telecommunication company as compare to group 2 who never used the internet for movie streaming.
This clearly shows that the two variables are associated. Movies streaming would tally to higher total charges compared to no movie streaming.
