ASK 2: STATISTICAL ANALYSIS AND TESTING – COMPLETE
REPORT TEMPLATE
A. RESEARCH QUESTION – PURPOSE
Research Question: Is there a statistically significant
difference/relationship between groups/variables? – e.g.,
Does mean revenue differ by region? Does conversion
rate differ by ad type? Is there correlation between
tenure and monthly charges?
Goal: Perform appropriate statistical test – t-test, ANOVA,
chi-square, correlation, z-test for proportions – Provide
Python code, assumptions checks, hypothesis
formulation, test statistic, p-value, effect size, confidence
interval, conclusion, business recommendation
Example Dataset: Use your assigned WGU D599 Task 2
dataset – e.g., Sales, Marketing, Churn, E-commerce –
Example here uses E-commerce – Compare Average
Order Value by Device – Mobile vs Desktop – And Test
Correlation Tenure vs MonthlyCharges – Replace with
your actual D599 task – Same workflow applies
Variables:
, • Numerical continuous – OrderValue, Tenure,
MonthlyCharges, Revenue
• Categorical – Device – Mobile/Desktop, Region –
East/West, AdType – A/B, Churn Yes/No
B. HYPOTHESIS TESTING FRAMEWORK – HIGH-YIELD
FOR D599
Steps:
1. Formulate Hypotheses: H0 null – No difference/no
effect – H1 alternative – Difference/effect – Two-
tailed vs one-tailed – Two-tailed – Difference – One-
tailed – Greater than – D599 usually two-tailed
unless specified
2. Choose Significance Level: α = 0.05 standard – 5%
chance Type I error – If p < α reject H0 – Statistically
significant
3. Choose Test: Based on variable types – Numerical vs
categorical groups
o 1 numerical + 2 groups – t-test – Independent t-
test – Mobile vs Desktop OrderValue – Paired t-
test – Before vs After same subjects
REPORT TEMPLATE
A. RESEARCH QUESTION – PURPOSE
Research Question: Is there a statistically significant
difference/relationship between groups/variables? – e.g.,
Does mean revenue differ by region? Does conversion
rate differ by ad type? Is there correlation between
tenure and monthly charges?
Goal: Perform appropriate statistical test – t-test, ANOVA,
chi-square, correlation, z-test for proportions – Provide
Python code, assumptions checks, hypothesis
formulation, test statistic, p-value, effect size, confidence
interval, conclusion, business recommendation
Example Dataset: Use your assigned WGU D599 Task 2
dataset – e.g., Sales, Marketing, Churn, E-commerce –
Example here uses E-commerce – Compare Average
Order Value by Device – Mobile vs Desktop – And Test
Correlation Tenure vs MonthlyCharges – Replace with
your actual D599 task – Same workflow applies
Variables:
, • Numerical continuous – OrderValue, Tenure,
MonthlyCharges, Revenue
• Categorical – Device – Mobile/Desktop, Region –
East/West, AdType – A/B, Churn Yes/No
B. HYPOTHESIS TESTING FRAMEWORK – HIGH-YIELD
FOR D599
Steps:
1. Formulate Hypotheses: H0 null – No difference/no
effect – H1 alternative – Difference/effect – Two-
tailed vs one-tailed – Two-tailed – Difference – One-
tailed – Greater than – D599 usually two-tailed
unless specified
2. Choose Significance Level: α = 0.05 standard – 5%
chance Type I error – If p < α reject H0 – Statistically
significant
3. Choose Test: Based on variable types – Numerical vs
categorical groups
o 1 numerical + 2 groups – t-test – Independent t-
test – Mobile vs Desktop OrderValue – Paired t-
test – Before vs After same subjects
