TI-84 for AP Statistics: Where Every STAT Function Lives

Almost every “how do I calculate ___ on a TI-84” question from a stats class has the same answer, just pointing at a different one of three menus. This is a map of which menu to use for what, plus the handful of values — variance, IQR, mode — that don’t have a dedicated command at all.

The three menus

• STAT → CALC — regression and one-list summaries: 1-Var Stats, LinReg(ax+b).
• STAT → TESTS — hypothesis tests and confidence intervals: Z-Test, T-Test, χ²-Test, TInterval, and the rest.
• 2ND + VARS (the DISTR menu) — probability from a named distribution: normalcdf(, invNorm(, binompdf(, invT(, and others.

Mean, median, standard deviation — and what it won’t give you

STAT → CALC → 1: 1-Var Stats, point it at your list, run it. The first screen shows x̄ (mean), Σx, Σx², Sx (sample standard deviation), σx (population standard deviation), and n. Scroll down with the arrow key for a second screen: minX, Q1, Med, Q3, maxX.

Three things it does not hand you directly:

• Variance — square the standard deviation yourself: sample variance is Sx², population variance is σx².
• IQR — subtract the two values from the second screen: Q3 − Q1.
• Mode — there’s no built-in mode command at all. Sort the list first (SortA( from the LIST menu) and count the most frequent value by eye, or write a short program if you need it repeatedly.

The regression line and correlation

STAT → CALC → 4: LinReg(ax+b) runs the least-squares regression line (LSRL) on paired data and reports the slope, intercept, and — with diagnostics turned on — r and r². This is covered in more depth, including how to turn diagnostics on, in How to Find the Correlation Coefficient on a TI-84.

Probability from a distribution

Everything under 2ND + VARS answers “what’s the probability” for a specific distribution, split into continuous and discrete:

• normalcdf( / invNorm( — continuous, normal distribution. See How to Find a Z-Score on a TI-84 and How to Find a P-Value on a TI-84 for worked examples.
• binompdf(n,p,x) / binomcdf(n,p,x) — discrete, binomial. binompdf gives P(X = x) exactly; binomcdf gives the cumulative P(X ≤ x).
• poissonpdf( / geometpdf( — the same pdf/cdf split for Poisson and geometric distributions, if your course covers them.

Confidence intervals and critical values

STAT → TESTS → 7: ZInterval / 8: TInterval (and the 1-prop, 2-sample versions further down the same list) compute the interval directly from summary stats or a data list — you don’t need the critical value separately for these.

If you do need the bare critical value itself — a t* or z* to plug into your own formula — it comes from the DISTR menu, not TESTS: invT(area to the left, df) for t*, or invNorm(area to the left) for z*. For a 95% confidence interval, that’s the left tail plus half the remaining 5%, i.e. invT(.975, df). (Older, non-CE units on an outdated OS may not have invT — invNorm is the closest built-in substitute if so.)

Hypothesis tests and “the test statistic”

“How do I calculate the test statistic” doesn’t have one answer — it depends on which test your data calls for. Run the matching procedure from STAT → TESTS (Z-Test, T-Test, 2-SampTTest, 1-PropZTest, and so on), and the test statistic is part of what it outputs, labeled z= or t= on the results screen — it isn’t a separate calculation you do afterward.

Chi-square: two different tests, two different inputs

• χ²GOF-Test (goodness of fit) — compares observed counts against expected counts, entered as two separate Lists.
• χ²-Test (association/independence) — compares counts across two categorical variables, entered as a single Matrix (2nd + x⁻¹ → EDIT to build it first).

Mixing these up — entering a matrix where the goodness-of-fit test expects two lists, or vice versa — is the most common reason this one throws an error before it even runs.

Try it yourself

The TI-84 calculator on this site has both menus under Tools: Tools → Tests matches STAT TESTS (the Z/T tests, proportion tests, and χ² Goodness-of-Fit), and Tools → Distributions matches the DISTR menu with the same function names — normalpdf, normalcdf, invNorm, binompdf, and the rest. Full list in the docs.