Think of the Wilcoxon signed-rank test as the non-parametric cousin of the t-test! It’s kind of like choosing between gelato and ice cream. Both are delicious but made differently!

• Your measurement level should be at least ordinal. That's like knowing which student ranks first, second, third in a race, but not knowing their exact running times.
• You're using a repeated measures or matched pairs design. Imagine testing if the same person runs faster with running shoes A or B.

• Find the difference between paired values.
• Arrange them by their absolute values (like sorting candies by size but ignoring the color). 🍬
• Assign ranks and toss out zero differences. Note: Treat tied values like twins! They share a rank.
• Return the original signs to the ranks.
• Compute the sums of positive (W+) and negative ranks (W–).

• Suppose we want to see if people remember better in quiet vs. noisy conditions. We test the same group in both conditions.
• After calculations, W+ suggests that people recall better in quiet conditions. But, is this difference big enough to matter? 🤔