Chapter 9 — Hypothesis Test (Commute Lab) - ICE

Your Progress

Green dots light up as you complete each step.

Step 1 — Paste your sample (minutes)

Reminder: We test the class mean against a target \( \mu_0 \) (default 30). Keep dataset consistent across t, p, and CI.

Step 2 — Summaries (x̄, s, SE, df)

Excel mirror — summaries (with a simple cell map)
Cell map: Put your sample in column A (one value per cell). Then set:
B1 = n, B2 = μ₀, B3 = df, B4 = α; C2 = x̄, C3 = s, C4 = SE.
B1 (n): =COUNTA(A:A)
C2 (x̄): =AVERAGE(A:A)
C3 (s): =STDEV.S(A:A)
C4 (SE): =C3/SQRT(B1)
B3 (df): =B1-1
MIN: =MIN(A:A)   •   MAX: =MAX(A:A)

Step 3 — Hypotheses & tail

H₀: μ = 30; H₁: μ ≠ 30 (two-sided)

Step 4 — Test statistic t

Excel mirror — t-statistic
Using the cells above: =(C2-B2)/(C3/SQRT(B1))
One-liner from the data column (μ₀ in B2): =(AVERAGE(A:A)-B2)/(STDEV.S(A:A)/SQRT(COUNTA(A:A)))

Step 5 — p-value & decision

Excel mirrors — p-value (copy)
Using t in D2 (or wherever you put it) and df in B3:
Right: =T.DIST.RT(D2,B3)   Left: =T.DIST(D2,B3,TRUE)   Two: =T.DIST.2T(ABS(D2),B3)
One-liners (direct from column A; μ₀ in B2):
Right: =T.DIST.RT((AVERAGE(A:A)-B2)/(STDEV.S(A:A)/SQRT(COUNTA(A:A))), COUNTA(A:A)-1)
Left: =T.DIST((AVERAGE(A:A)-B2)/(STDEV.S(A:A)/SQRT(COUNTA(A:A))), COUNTA(A:A)-1, TRUE)
Two: =T.DIST.2T(ABS((AVERAGE(A:A)-B2)/(STDEV.S(A:A)/SQRT(COUNTA(A:A)))), COUNTA(A:A)-1)
Decision templates (compare to α in B4):
Right: =IF(T.DIST.RT(D2,B3)<=B4,"Reject H0","Fail to reject H0")   Left: =IF(T.DIST(D2,B3,TRUE)<=B4,"Reject H0","Fail to reject H0")   Two: =IF(T.DIST.2T(ABS(D2),B3)<=B4,"Reject H0","Fail to reject H0")

Step 6 — t* and CI / bound

Excel mirrors — t* & CI/bounds (α in B4, df in B3, x̄ in C2, SE in C4)
Two-sided t*: =T.INV.2T(B4,B3)   One-sided t* (both cases below use this): =T.INV(1-B4,B3)
Two-sided CI lower: =C2 - T.INV.2T(B4,B3)*C4   upper: =C2 + T.INV.2T(B4,B3)*C4
Right-tail (H₁: μ > μ₀) — lower bound: =C2 - T.INV(1-B4,B3)*C4   (upper is +∞)
Left-tail (H₁: μ < μ₀) — upper bound: =C2 + T.INV(1-B4,B3)*C4   (lower is −∞)

Step 7 — Auto summary

Hypothesis → t → p (cheat-sheet)

A) Hypotheses (H₀ includes “=”).
  • Longer than 30 (right tail) ⇒ H₁: μ > 30 (right); H₀: μ ≤ 30 (includes “=”).
  • Less than 30 (left tail) ⇒ H₁: μ < 30 (left); H₀: μ ≥ 30 (includes “=”).
  • Different from 30 (two tails) ⇒ H₁: μ ≠ 30 (two); H₀: μ = 30.
B) Test statistic t = (x̄ − μ₀)/(s/√n). Excel: =(C2-B2)/(C3/SQRT(B1))
C) p-value Right: =T.DIST.RT(t,df) • Left: =T.DIST(t,df,TRUE) • Two: =T.DIST.2T(ABS(t),df)

\(p=1-F(t)\) (right), \(p=F(t)\) (left), \(p=2\min\{F(t),1-F(t)\}\) (two).

Right-tailed

p = shaded area to the right of t
p = —

Left-tailed

p = shaded area to the left of t
p = —

Two-sided

p = both tails beyond |t|
p = —

Examples — left, right, and two tails for common t (df = 22)

Exact t-CDF. Rows: t ∈ {1.5, 2, 3, 3.8}. Columns: Right, Left, Two.

Video — Hypothesis Testing (Introduction)

Open on YouTube Link: https://www.youtube.com/watch?v=DlwOTOydeyk