Critical Region Sketch Trainer: t-Test Rejection Regions
Sketching the curve and shading the region before you look anything up is the habit that stops most t-test mistakes: the wrong tail, the one-tailed column for a two-tailed test, or a missing minus sign.
Critical Region Sketch Trainer
Six rounds. Each round describes a one-sample t-test. Step 1: tap where the rejection region goes (left tail, right tail or both). Step 2: place t* on the curve with the slider, the ± buttons or a tap on the sketch. 1 point for the region and 1 for a t* within 0.05 of the t-table value.
0.00
Need the numbers? Open the t-table or the printable t* card in another tab. Red: your region. Green (after checking): the correct region.
Worked example
Question (illustrative numbers): a supplier claims a mean of 50. A sample of n = 15 gives x̄ = 52.1 and s = 4.2. Test H₀: μ = 50 against H₁: μ > 50 at α = 0.05.
Step 1. Sketch: “greater than” → the rejection region is the right tail, holding all of α = 0.05.
Step 2. Find t*: df = n − 1 = 14. One-tailed column α = 0.05, row 14 of the t-table: t* = 1.761. Reject H₀ if t > 1.761.
Step 3. Test statistic: t = (52.1 − 50) / (4.2 / √15) = 2.1 / 1.0844 = 1.936.
Answer: 1.936 > 1.761, so t is inside the region: reject H₀ at the 5% level (one-tailed p ≈ 0.037). If H₁ had been μ ≠ 50, t* would be ±2.145 and 1.936 would not be significant.
Region, column and sign at a glance
| H₁ | Region | Column | df 14, α 0.05 |
|---|---|---|---|
| μ < μ₀ | Left tail | One-tailed α | t < −1.761 |
| μ > μ₀ | Right tail | One-tailed α | t > 1.761 |
| μ ≠ μ₀ | Both tails | Two-tailed α | |t| > 2.145 |
FAQ
What is a critical region in a t-test?
It is the set of t values that lead you to reject H₀. Its edge is the critical value t*, read from the t-table at your α and df. If your test statistic falls inside the region, the result is significant at level α.
How do I know which tail the rejection region is in?
Look at H₁. “Less than” (μ < μ₀) puts the region in the left tail, “greater than” (μ > μ₀) in the right tail, and “not equal” (μ ≠ μ₀) in both tails, with α/2 in each.
Which t-table column do I use for a one-tailed test?
Use the one-tailed column for α. For α = 0.05 and df = 14 that is 1.761. A two-tailed test at α = 0.05 uses the two-tailed column (α/2 = 0.025 per tail), which gives 2.145 for df = 14.
Is the left-tailed critical value negative?
Yes. The t-distribution is symmetric, so the left-tailed t* is the negative of the right-tailed value: −1.761 for α = 0.05 and df = 14. Reject H₀ when t is below it.
How is scoring done in the game?
Each round is worth 2 points: 1 for the correct region (left, right or both tails) and 1 for placing t* within 0.05 of the table value. Six rounds give a maximum of 12.
What degrees of freedom does a one-sample t-test use?
df = n − 1. The game gives n and df each round. For paired tests use the number of pairs minus 1; for other designs see the degrees of freedom guide.
Related t-table pages
- T table (one-tailed and two-tailed): every t* used in the game
- Using the t-table for hypothesis testing and t-table critical values
- Confidence level → t* and the t-table calculator
- T-distribution simulator and the printable t* pocket card
- More practice: Guess the Critical t and DF Race
- For z-tests: the z-table on z-table.com · p-values at StatCalculators