Statistic Test 2 ENG

  Patterns 

Scenario, Metric, Question

---

Z-test (Proportions / Rates)

Z-test:
Show size recommendation badge, Conversion rate, Does conversion rate increase?

New size prediction model, Return rate due to size, Does return rate decrease?

“Find your size” button, Click-through rate, Does CTR increase?

Longer size form, Completion rate, Does completion rate decrease?

Virtual fitting feature, Adoption rate, Does adoption rate increase?

New size guide page, Bounce rate, Does bounce rate decrease?

Helpful popup after purchase, Yes rate, Does Yes rate increase?

---

T-test (Continuous Means, 2 Groups)

T-test:
New size chart, Average order value, Does average order value increase?

New size guide layout, Average time on page, Does average time increase?

Outfit recommendation, Average items per order, Does average number of items increase?

Fit satisfaction survey, Average fit score, Does average fit score increase?

New return process, Average processing time, Does average processing time decrease?

Compare 2 size models, Average size error, Does average error decrease?

New PDP layout, Average revenue per session, Does average revenue increase?

---

Chi-square test (Categorical Distributions)

Chi-square test:
Return reasons, Too small / Just right / Too big, Is the distribution of reasons different?

Fit perception, Fit categories, Is the fit distribution different?

Size funnel steps, Step1 / Step2 / Step3 / Complete, Is the funnel distribution different?

Device type, Mobile / Desktop / Tablet, Are group and device independent?

Purchased size distribution, XS / S / M / L / XL, Is the size distribution different?

Feedback levels, Very unhelpful / … / Very helpful, Is the feedback distribution different?

---

ANOVA (More Than 2 Groups, Continuous Means)

ANOVA:
3 size recommendation UIs, Average conversion score, Is any group different?

4 ML size models, Average size error, Is any model better?

3 size guide layouts, Average time on page, Is any layout different?

3 outfit strategies, Average items per order, Is any strategy better?

3 size form versions, Average completion time, Is complexity different?

3 confidence displays, Average trust score, Is trust different?

---

One Ultra-Compact Version (Great for Fast Review)

You can memorize like this:

Z-test

• Size badge, conversion rate, does conversion increase?

• Size model, return rate, does return decrease?

• Button, CTR, does CTR increase?

T-test

• Size chart, avg order value, does mean increase?

• Size guide, avg time, does mean increase?

• Fit survey, avg score, does mean increase?

Chi-square

• Return reasons, category distribution, is distribution different?

• Fit categories, distribution, is distribution different?

• Funnel steps, distribution, is distribution different?

ANOVA

• 3 UIs, avg metric, is any group different?

• 4 models, avg error, is any model better?

• 3 layouts, avg time, is any layout different?