WGU D606 EBN2 – DATA SCIENCE
CAPSTONE
TASK 1: DATA SCIENCE CAPSTONE TOPIC
APPROVAL AND RELEASE FORM
Western Governors University
Program: Master of Science in Data Analytics – Data Science Specialization
Course: D606 – Data Science Capstone
Assessment Code: EBN2
Assessment: Task 1 – Capstone Topic Approval and Release Form
Capstone Project Title:
Predicting Mobile Phone Customer Satisfaction Using Multiple Linear Regression: An
Analysis of Product Features, Pricing, and Brand
PART A: CAPSTONE TOPIC APPROVAL
FORM
1. Capstone Project Name
Predicting Mobile Phone Customer Satisfaction Using Multiple Linear Regression: An Analysis
of Product Features, Pricing, and Brand.
2. Project Topic
The proposed capstone project will examine the relationship between mobile phone product
characteristics and overall customer satisfaction using multiple linear regression.
The research will evaluate whether customer ratings for battery life, camera quality, device
performance, physical design, and display quality can predict overall mobile phone satisfaction.
Brand and price will also be investigated to determine whether they provide additional predictive
information.
The project will use the Global Mobile Reviews Dataset (2025 Edition), publicly available
through Kaggle, as the primary data source.
,The dataset described in the accompanying Task 2 report contains 50,000 mobile phone
customer reviews from eight countries, spanning October 2022 through October 2025.
The objective is to develop and evaluate an interpretable statistical model that can help mobile
phone manufacturers understand customer satisfaction patterns and identify potential areas for
product development and quality improvement.
The research will integrate data preparation, exploratory data analysis, statistical modeling,
feature selection, regression diagnostics, and model validation.
3. Human Subjects Research Declaration
Proposed classification:
This project is designed as secondary analysis of an existing dataset without direct interaction
with human participants.
The study will not recruit participants, conduct interviews, distribute questionnaires, or collect
new customer responses.
The analytical dataset will exclude customer names, review identifiers, review text, and other
fields unnecessary for the research question.
The project will primarily use numerical ratings, product prices, and categorical brand
information.
Although the research is intended to use de-identified information, the original dataset contains
customer-related fields. These will be reviewed and excluded from analytical outputs to protect
reviewer privacy.
A formal determination of whether the study is exempt from institutional review requirements
must be made according to WGU's applicable policies.
4. Research Question
Primary Research Question
Can a multiple linear regression model be constructed using mobile phone feature-specific
ratings, brand, and price to predict overall customer satisfaction with a test-set R-squared
greater than 0.70?
This question directly addresses the predictive modeling objective established for the capstone.
Supporting Research Questions
, 1. How strongly are battery life, camera quality, device performance, design, and display
ratings associated with overall mobile phone satisfaction?
2. Do mobile phone brand and price provide meaningful additional predictive value after
accounting for feature-specific ratings?
3. Which product characteristics are retained in a statistically useful multiple linear
regression model?
4. Can feature selection simplify the prediction model without materially reducing its
predictive performance?
5. How accurately can the resulting model predict customer ratings for previously unseen
observations?
The study will also examine relevant model assumptions and determine whether the fitted
regression is sufficiently reliable for its intended analytical purpose.
5. Research Hypothesis
The research will evaluate the predictive usefulness of a multiple linear regression model for
estimating overall mobile phone customer satisfaction.
Null Hypothesis (H₀)
A multiple linear regression model using product feature ratings, mobile phone brand, and price
will not achieve a test-set coefficient of determination greater than 0.70.
H₀: Test-set R² ≤ 0.70
Alternative Hypothesis (H₁)
A multiple linear regression model using product feature ratings, mobile phone brand, and price
will achieve a test-set coefficient of determination greater than 0.70.
H₁: Test-set R² > 0.70
The alternative hypothesis anticipates that customer evaluations of specific mobile phone
features will contain substantial information relevant to predicting overall satisfaction.
Battery performance, camera quality, display quality, device performance, and physical design
are expected to demonstrate positive relationships with overall customer ratings.
The influence of price and brand is less certain and will be assessed empirically.
Statistical Interpretation
The R² threshold of 0.70 represents a project-specific predictive performance benchmark.
CAPSTONE
TASK 1: DATA SCIENCE CAPSTONE TOPIC
APPROVAL AND RELEASE FORM
Western Governors University
Program: Master of Science in Data Analytics – Data Science Specialization
Course: D606 – Data Science Capstone
Assessment Code: EBN2
Assessment: Task 1 – Capstone Topic Approval and Release Form
Capstone Project Title:
Predicting Mobile Phone Customer Satisfaction Using Multiple Linear Regression: An
Analysis of Product Features, Pricing, and Brand
PART A: CAPSTONE TOPIC APPROVAL
FORM
1. Capstone Project Name
Predicting Mobile Phone Customer Satisfaction Using Multiple Linear Regression: An Analysis
of Product Features, Pricing, and Brand.
2. Project Topic
The proposed capstone project will examine the relationship between mobile phone product
characteristics and overall customer satisfaction using multiple linear regression.
The research will evaluate whether customer ratings for battery life, camera quality, device
performance, physical design, and display quality can predict overall mobile phone satisfaction.
Brand and price will also be investigated to determine whether they provide additional predictive
information.
The project will use the Global Mobile Reviews Dataset (2025 Edition), publicly available
through Kaggle, as the primary data source.
,The dataset described in the accompanying Task 2 report contains 50,000 mobile phone
customer reviews from eight countries, spanning October 2022 through October 2025.
The objective is to develop and evaluate an interpretable statistical model that can help mobile
phone manufacturers understand customer satisfaction patterns and identify potential areas for
product development and quality improvement.
The research will integrate data preparation, exploratory data analysis, statistical modeling,
feature selection, regression diagnostics, and model validation.
3. Human Subjects Research Declaration
Proposed classification:
This project is designed as secondary analysis of an existing dataset without direct interaction
with human participants.
The study will not recruit participants, conduct interviews, distribute questionnaires, or collect
new customer responses.
The analytical dataset will exclude customer names, review identifiers, review text, and other
fields unnecessary for the research question.
The project will primarily use numerical ratings, product prices, and categorical brand
information.
Although the research is intended to use de-identified information, the original dataset contains
customer-related fields. These will be reviewed and excluded from analytical outputs to protect
reviewer privacy.
A formal determination of whether the study is exempt from institutional review requirements
must be made according to WGU's applicable policies.
4. Research Question
Primary Research Question
Can a multiple linear regression model be constructed using mobile phone feature-specific
ratings, brand, and price to predict overall customer satisfaction with a test-set R-squared
greater than 0.70?
This question directly addresses the predictive modeling objective established for the capstone.
Supporting Research Questions
, 1. How strongly are battery life, camera quality, device performance, design, and display
ratings associated with overall mobile phone satisfaction?
2. Do mobile phone brand and price provide meaningful additional predictive value after
accounting for feature-specific ratings?
3. Which product characteristics are retained in a statistically useful multiple linear
regression model?
4. Can feature selection simplify the prediction model without materially reducing its
predictive performance?
5. How accurately can the resulting model predict customer ratings for previously unseen
observations?
The study will also examine relevant model assumptions and determine whether the fitted
regression is sufficiently reliable for its intended analytical purpose.
5. Research Hypothesis
The research will evaluate the predictive usefulness of a multiple linear regression model for
estimating overall mobile phone customer satisfaction.
Null Hypothesis (H₀)
A multiple linear regression model using product feature ratings, mobile phone brand, and price
will not achieve a test-set coefficient of determination greater than 0.70.
H₀: Test-set R² ≤ 0.70
Alternative Hypothesis (H₁)
A multiple linear regression model using product feature ratings, mobile phone brand, and price
will achieve a test-set coefficient of determination greater than 0.70.
H₁: Test-set R² > 0.70
The alternative hypothesis anticipates that customer evaluations of specific mobile phone
features will contain substantial information relevant to predicting overall satisfaction.
Battery performance, camera quality, display quality, device performance, and physical design
are expected to demonstrate positive relationships with overall customer ratings.
The influence of price and brand is less certain and will be assessed empirically.
Statistical Interpretation
The R² threshold of 0.70 represents a project-specific predictive performance benchmark.
