• Statistical method for decision making using data
• Widely used in UI/UX through A/B testing
• Helps determine if design changes actually work

• Version A: Old checkout button
• Version B: New checkout button
• Metric: Click-through rate (CTR) ## Hypotheses

\[ H_0: p_A = p_B \]

\[ H_1: p_A \neq p_B \]

\[ z = \frac{\hat{p}_A - \hat{p}_B}{\sqrt{p(1-p)(1/n_A + 1/n_B)}} \]

Click-Through Rate Plot

library(ggplot2)

ctr <- aggregate(clicks ~ version, data, mean)

ggplot(ctr, aes(x = version, y = clicks, fill = version)) +
  geom_col() +
  labs(
    title = "Click-Through Rate by Design Version",
    x = "Design Version",
    y = "Click-Through Rate"
  ) +
  theme_minimal()

Distribution of Clicks

## [1] 0.2306148

• p-value > 0.05
• We fail to reject the null hypothesis
• The observed difference may be due to chance

• Small p-value → strong evidence against H₀
• In UI/UX, this means the design change likely matters
• Statistical significance ≠ good user experience

-Hypothesis testing supports UX decisions

-A/B testing prevents guesswork

-Statistics improves product outcomes