
construct and interpret a t-interval for a population mean, explain why t replaces z when σ is unknown, and determine the sample size needed for a desired margin of error.
Instruction
From z to t. If we knew σ, an interval for μ would be x̄ ± z*σ/√n. We almost never know σ, so we use the sample SD s, and the standard error SE = s/√n. Replacing σ with s adds uncertainty, especially for small samples. In 1908 William Gosset, a statistician at the Guinness brewery who published as “Student”, worked out the exact distribution that results. The t distribution is symmetric and bell-shaped like the standard normal, but with heavier tails. There is one for each degrees of freedom value, df = n − 1, and as df grows it approaches the normal.
| df | 9 | 14 | 29 | 99 | ∞ (z) |
|---|---|---|---|---|---|
| t* for 95% | 2.262 | 2.145 | 2.045 | 1.984 | 1.960 |
The one-sample t-interval is
x̄ ± t*·s/√n, df = n − 1
Conditions. Random sample or random assignment. 10% when sampling without replacement. Normal/large sample: the population is approximately normal, or n ≥ 30 and the sample shows no strong skew or outliers. For small samples, graph the data and look for skew and outliers.
Worked example. The population is the daily high temperature (TMAX) at St. Louis Lambert International Airport for every day from 2000 to 2024 (NOAA GHCN-Daily, 9,132 days). A random sample of 30 days, listed in date order, gave (°F): 37, 88, 71, 94, 75, 82, 30, 75, 82, 30, 78, 74, 59, 28, 56, 84, 86, 34, 70, 77, 28, 71, 62, 47, 24, 82, 37, 39, 90, 46.
- Parameter: μ = the mean daily high at St. Louis over 2000-2024.
- Conditions: random sample; 30 < 10% of 9,132; n = 30 and a dot plot of the sample shows a wide spread with no outliers (the values run from 24 to 94°F).
- Compute: x̄ = 61.20°F, s = 22.596°F, SE = 22.596/√30 ≈ 4.125. With df = 29, t* = 2.045. Margin = 2.045 × 4.125 ≈ 8.44. Interval: 52.76°F to 69.64°F.
- Interpret: We are 95% confident that the mean daily high temperature at St. Louis over 2000-2024 was between 52.8°F and 69.6°F.
Check against the population: μ = 67.59°F, inside the interval, though near its upper end. This sample happened to run cool, by 6.4 degrees. That is sampling variability, and the interval’s width accounts for it. The interval is wide because daily highs in St. Louis swing from the 20s in winter to the 90s in summer (σ = 20.1°F), so thirty days pin the mean down only roughly.
Spreadsheet. =AVERAGE, =STDEV.S, and =T.INV.2T(0.05, 29) for t*, or =CONFIDENCE.T(0.05, s, n) for the margin directly. GeoGebra Probability’s Student t option shows the curve and its critical values.
Sample size for a mean. Solve E = z*σ/√n for n, using z* (not t*, because n is unknown) and a planning value for σ from earlier data:
n = (z*·σ/E)2
To estimate St. Louis’s mean daily high within ±1°F with 95% confidence, using σ ≈ 20.1°F: n = (1.96 × 20.1/1)2 ≈ 1,545.0, so 1,546 days.
Robustness. t-procedures are fairly robust: they work well when the population is not normal, as long as there are no outliers and n is not tiny. They are not resistant: one extreme value moves both x̄ and s. With a small, skewed sample, report that the conditions are doubtful rather than pretending they are met.
📖 Read for this lesson
The reading opens over the lesson; close it to come back to exactly where you were. The checks below draw on it.
Formative check
Work through these before moving on. They are not graded — they tell you whether the standard below has landed yet.
Why does the t-interval use t* = 2.045 (df = 29) instead of z* = 1.96?
x̄ = 61.20, s = 22.60, n = 30: SE ≈ and the 95% interval is 61.20 ± .
Simulation & tools
Compare t and normal critical values
In GeoGebra Probability, choose Student (the t distribution).
- Set degrees of freedom to 29 and use the two-sided interval to find the value that leaves 95% in the middle. Record it and compare with 2.045.
- Repeat for df = 4, 9 and 99. Record t* for each.
- Overlay the normal curve (the checkbox or a second tab with Normal μ = 0, σ = 1). Where do the curves differ most?
- Explain in two sentences why a sample of 5 needs a much larger t* than a sample of 100.
Without GeoGebra: use =T.INV.2T(0.05, df) in a spreadsheet for the same df values.
Open GeoGebra ↗ · GeoGebra · free for non-commercial use
Draw your own random sample of 30 days of TMAX from any Missouri station’s NOAA daily file (the Kansas City reading shows a finished version). Build a 95% t-interval by hand, confirm it with CONFIDENCE.T or =T.INV.2T, interpret it, and compute the station’s true mean over the whole file. Did your interval capture it? Then find the sample size needed to estimate that mean within ±0.5°F.
How confident are you that you can build and interpret a t-interval for a mean, with conditions checked, and plan a sample size?
Practice
Work these on paper or in your notebook, then open Check your answer. Aim for all of Fluency and Application; try at least one Challenge.
Printable version: this unit’s practice workbook (PDF)
Fluency
Build speed and accuracy with the core skill.
- A random sample of 15 has mean 66.9 and standard deviation 10.2. Construct a 95% confidence interval for the population mean.
Check your answer
Answer: t* ≈ 2.145 (df = 14); margin ≈ 5.65; interval ≈ (61.25, 72.55). We are 95% confident that the population mean is between 61.25 and 72.55.Use t, not z, because the population standard deviation is unknown. Conditions: a random sample, less than 10% of the population, and a roughly symmetric sample without strong outliers (or n ≥ 30). - A random sample of 12 has mean 52.7 and standard deviation 14.9. Construct a 90% confidence interval for the population mean.
Check your answer
Answer: t* ≈ 1.796 (df = 11); margin ≈ 7.72; interval ≈ (44.98, 60.42). We are 90% confident that the population mean is between 44.98 and 60.42.Use t, not z, because the population standard deviation is unknown. Conditions: a random sample, less than 10% of the population, and a roughly symmetric sample without strong outliers (or n ≥ 30). - A random sample of 25 has mean 57.2 and standard deviation 8.9. Construct a 95% confidence interval for the population mean.
Check your answer
Answer: t* ≈ 2.064 (df = 24); margin ≈ 3.67; interval ≈ (53.53, 60.87). We are 95% confident that the population mean is between 53.53 and 60.87.Use t, not z, because the population standard deviation is unknown. Conditions: a random sample, less than 10% of the population, and a roughly symmetric sample without strong outliers (or n ≥ 30). - A random sample of 12 has mean 72.6 and standard deviation 13.0. Construct a 90% confidence interval for the population mean.
Check your answer
Answer: t* ≈ 1.796 (df = 11); margin ≈ 6.74; interval ≈ (65.86, 79.34). We are 90% confident that the population mean is between 65.86 and 79.34.Use t, not z, because the population standard deviation is unknown. Conditions: a random sample, less than 10% of the population, and a roughly symmetric sample without strong outliers (or n ≥ 30). - A random sample of 20 has mean 67.6 and standard deviation 7.4. Construct a 90% confidence interval for the population mean.
Check your answer
Answer: t* ≈ 1.729 (df = 19); margin ≈ 2.86; interval ≈ (64.74, 70.46). We are 90% confident that the population mean is between 64.74 and 70.46.Use t, not z, because the population standard deviation is unknown. Conditions: a random sample, less than 10% of the population, and a roughly symmetric sample without strong outliers (or n ≥ 30). - A random sample of 30 has mean 59.2 and standard deviation 8.5. Construct a 90% confidence interval for the population mean.
Check your answer
Answer: t* ≈ 1.699 (df = 29); margin ≈ 2.64; interval ≈ (56.56, 61.84). We are 90% confident that the population mean is between 56.56 and 61.84.Use t, not z, because the population standard deviation is unknown. Conditions: a random sample, less than 10% of the population, and a roughly symmetric sample without strong outliers (or n ≥ 30). - Find t* for 95% confidence with n = 10 and with n = 30.
Check your answer
Answer: df = 9: 2.262; df = 29: 2.045. - Why use t* rather than z* when σ is unknown?
Check your answer
Answer: Estimating σ with s adds uncertainty; the t distribution’s heavier tails widen the interval to account for it.
Application
Use the skill in context. Show your reasoning.
- A random sample of 25 days from Springfield’s 2000-2024 daily highs (NOAA) has x̄ = 67.60°F and s = 25.18°F. Construct and interpret a 95% t-interval for the mean daily high.
Check your answer
Answer: df = 24, t* ≈ 2.064, SE = 25.18/5 ≈ 5.037; interval ≈ (57.20, 78.00)°F. We are 95% confident that the mean daily high at Springfield over 2000-2024 is in this range.The population mean is 67.92°F; the interval captured it. - Springfield daily highs have σ ≈ 18.83°F. How many randomly chosen days would estimate the mean daily high within ±0.5°F with 95% confidence?
Check your answer
Answer: n = (1.96 × 18.83/0.5)² ≈ 5446.2, so 5,447 days.Use z* when planning; round up. - A random sample of 8 daily rainfall totals at Springfield contains six zeros and two large values. Is a t-interval for the mean appropriate?
Check your answer
Answer: No. The sample is tiny and strongly skewed with outliers (rainfall is mostly zeros), so the normal/large-sample condition fails; the interval would be unreliable.
Challenge
Stretch problems. Expect to think before you write.
- For a sample of 5 with s = 4, compare the 95% margins of error computed with t* and with z*. By what percent does z* understate the margin?
Check your answer
Answer: t* (df 4) ≈ 2.776: margin ≈ 4.97; z*: margin ≈ 3.51. z* understates it by about 29%.
Review
Keep earlier skills sharp.
- (Unit 1) Find the SD of x̄ for samples of 49 from a population with σ = 14.
Check your answer
Answer: 2.σ/√n. - (RLP Statistics A) Name the sampling method: a random sample of days within each month, combined.
Check your answer
Answer: Stratified random sampling (strata = months).
A Second Station, a Second Sample: Thirty Kansas City Days Against the Whole Record
NOAA NCEI GHCN-Daily, station USW00003947 (Kansas City International Airport), element TMAX, 2000-2024 (N = 9,132 days); 30-day sample drawn for this course with Python’s random module, seed 2026; retrieved 2026-11-10 · Public domain (works of the United States government), with an SLS original sample and computation
Why this reading. The lesson’s open question asks you to draw a 30-day random sample from a NOAA daily file, build a t-interval, and then compute the station’s true mean “over the whole file”. That last step needs the whole file. This reading does the whole exercise on a second station, Kansas City, and prints what the exercise needs: the sample with its dates, and the population parameters computed from all 9,132 days.
First, the lesson’s St. Louis sample, checked
The 30 values in the lesson sum to 1,836, so x̄ = 1,836/30 = 61.20, and the sum of squared deviations is 14,806.8, so s = √(14,806.8/29) = 22.596. Both agree with the lesson. SE = 22.596/√30 = 4.125; with t* = 2.045 (df = 29), the margin is 8.44 and the interval is 52.76 to 69.64°F.
The Kansas City population
NOAA GHCN-Daily, station USW00003947 (Kansas City International Airport), element TMAX, January 1, 2000 to December 31, 2024: N = 9,132 days with a recorded high, none missing. Retrieved 2026-11-10. Parameters: μ = 65.78°F, σ = 20.39°F, minimum -3, maximum 107, 976 days at or above 90°F and 75 at or above 100°F.
| TMAX (°F) | 0–9 | 10–19 | 20–29 | 30–39 | 40–49 | 50–59 | 60–69 | 70–79 | 80–89 | 90–99 | 100–109 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Days | 17 | 110 | 311 | 738 | 973 | 1,222 | 1,315 | 1,478 | 1,990 | 901 | 75 |
This is not a normal population. Kansas City has cold winters and hot summers, so the daily highs are spread over a very wide range, with the most days in the 80s and a long tail toward the cold. Keep that in mind: the t-interval’s condition is about the shape of the sample and the size of n, and the claim that t is robust is about to be tested on a population it was not designed for.
The sample
Thirty days drawn without replacement with random.Random(2026).sample(range(9132), 30) from the days in date order, then sorted by date for printing.
| # | Date | TMAX | # | Date | TMAX | # | Date | TMAX |
|---|---|---|---|---|---|---|---|---|
| 1 | 2000-02-11 | 30 | 11 | 2010-10-06 | 80 | 21 | 2018-11-13 | 33 |
| 2 | 2000-07-07 | 91 | 12 | 2011-04-13 | 78 | 22 | 2019-10-14 | 72 |
| 3 | 2003-06-04 | 73 | 13 | 2012-11-19 | 65 | 23 | 2020-03-03 | 62 |
| 4 | 2003-08-15 | 94 | 14 | 2014-02-04 | 25 | 24 | 2021-12-28 | 57 |
| 5 | 2004-05-25 | 71 | 15 | 2014-05-01 | 54 | 25 | 2022-01-07 | 26 |
| 6 | 2004-08-08 | 81 | 16 | 2015-04-09 | 77 | 26 | 2022-07-17 | 87 |
| 7 | 2004-12-19 | 23 | 17 | 2015-08-06 | 82 | 27 | 2022-12-16 | 33 |
| 8 | 2005-05-05 | 72 | 18 | 2015-12-31 | 39 | 28 | 2023-01-14 | 50 |
| 9 | 2009-06-06 | 86 | 19 | 2016-11-18 | 68 | 29 | 2024-07-28 | 85 |
| 10 | 2010-01-06 | 20 | 20 | 2017-10-26 | 70 | 30 | 2024-12-18 | 42 |
A dot plot of the 30 values runs from 20 to 94 with no gap large enough to call an outlier; by decade, 4 in the 20s, 4 in the 30s, 1 in the 40s, 3 in the 50s, 3 in the 60s, 7 in the 70s, 6 in the 80s, 2 in the 90s. Spread across the whole range, as a year-round sample from a four-season climate should be, with no single extreme value.
The interval, by hand
Parameter: μ = the mean daily high at Kansas City International Airport over 2000–2024.
Conditions: random sample (seeded generator, stated); 30 < 10% of 9,132; n = 30 with no outliers in the dot plot.
Compute: sum 1,826, x̄ = 60.87; sum of squared deviations 15,451.47, s = √(15,451.47/29) = 23.083; SE = 23.083/√30 = 4.214; t* = 2.045; margin = 8.62. Interval: 52.25 to 69.49°F. Spreadsheet: =CONFIDENCE.T(0.05, 23.083, 30) = 8.62.
Interpret: we are 95% confident that the mean daily high at Kansas City over 2000–2024 was between 52.2 and 69.5°F.
Check against the truth: μ = 65.78, inside the interval. This sample happened to run cool, by 4.9 degrees; the interval’s width absorbed it.
Compare the two stations: the St. Louis sample gave a margin of 8.44 with s = 22.6; the Kansas City sample gives 8.62 with s = 23.1. Same n, same t*, nearly the same width, because both cities swing through a very wide range of temperatures over the year (population SDs 20.1 and 20.4). The margin is proportional to s.
Sample size
To estimate Kansas City’s mean high within ±0.5°F at 95% confidence, use the population’s σ as the planning value: n = (1.96 × 20.39/0.5)2 = 6,390.9, so 6,391 days, about 70% of the whole record. Within ±1°F: 1,597.7, so 1,598. The lesson’s St. Louis figure for ±1°F was 1,546. When the data swing this widely over a year, precision is expensive; you buy it with sample size, and with nothing else.
Sources: NOAA NCEI, Daily Summaries (GHCN-Daily) data search (ncei.noaa.gov); NOAA NCEI Access Data Service (the query used here) (ncei.noaa.gov); Python documentation, the random module (docs.python.org)
The Tables: t Critical Values, t Tail Areas, Chi-Square and z
Critical values and tail areas of the t, chi-square and standard normal distributions, computed for this course and rounded; any published table will agree · SLS original
Why this reading. Several lessons tell you to compute a t-interval or a P-value “by hand” and then confirm with T.INV.2T or CHISQ.DIST.RT. The lessons print five t* values and two chi-square critical values between them. This page is the rest: enough of each table to do every hand computation in Units 2 to 4, and to bracket a P-value without a device. All values were computed for this course and rounded to three decimals; a printed table in any textbook will agree.
Table 1. Critical values t* of Student’s t
The column heading gives the confidence level for a two-sided interval and, in parentheses, the area in one tail. For a one-sided test at α = 0.05, use the 0.05-tail column (90%); for a two-sided test at α = 0.05, the 0.025-tail column (95%). If your df is not listed, use the next smaller df, which is conservative.
| df | 80% (tail 0.10) | 90% (tail 0.05) | 95% (tail 0.025) | 98% (tail 0.01) | 99% (tail 0.005) | 99.9% (tail 0.0005) |
|---|---|---|---|---|---|---|
| 1 | 3.078 | 6.314 | 12.706 | 31.821 | 63.657 | 636.619 |
| 2 | 1.886 | 2.920 | 4.303 | 6.965 | 9.925 | 31.599 |
| 3 | 1.638 | 2.353 | 3.182 | 4.541 | 5.841 | 12.924 |
| 4 | 1.533 | 2.132 | 2.776 | 3.747 | 4.604 | 8.610 |
| 5 | 1.476 | 2.015 | 2.571 | 3.365 | 4.032 | 6.869 |
| 6 | 1.440 | 1.943 | 2.447 | 3.143 | 3.707 | 5.959 |
| 7 | 1.415 | 1.895 | 2.365 | 2.998 | 3.499 | 5.408 |
| 8 | 1.397 | 1.860 | 2.306 | 2.896 | 3.355 | 5.041 |
| 9 | 1.383 | 1.833 | 2.262 | 2.821 | 3.250 | 4.781 |
| 10 | 1.372 | 1.812 | 2.228 | 2.764 | 3.169 | 4.587 |
| 12 | 1.356 | 1.782 | 2.179 | 2.681 | 3.055 | 4.318 |
| 14 | 1.345 | 1.761 | 2.145 | 2.624 | 2.977 | 4.140 |
| 16 | 1.337 | 1.746 | 2.120 | 2.583 | 2.921 | 4.015 |
| 18 | 1.330 | 1.734 | 2.101 | 2.552 | 2.878 | 3.922 |
| 19 | 1.328 | 1.729 | 2.093 | 2.539 | 2.861 | 3.883 |
| 20 | 1.325 | 1.725 | 2.086 | 2.528 | 2.845 | 3.850 |
| 24 | 1.318 | 1.711 | 2.064 | 2.492 | 2.797 | 3.745 |
| 25 | 1.316 | 1.708 | 2.060 | 2.485 | 2.787 | 3.725 |
| 29 | 1.311 | 1.699 | 2.045 | 2.462 | 2.756 | 3.659 |
| 30 | 1.310 | 1.697 | 2.042 | 2.457 | 2.750 | 3.646 |
| 40 | 1.303 | 1.684 | 2.021 | 2.423 | 2.704 | 3.551 |
| 50 | 1.299 | 1.676 | 2.009 | 2.403 | 2.678 | 3.496 |
| 53 | 1.298 | 1.674 | 2.006 | 2.399 | 2.672 | 3.484 |
| 56 | 1.297 | 1.673 | 2.003 | 2.395 | 2.667 | 3.473 |
| 60 | 1.296 | 1.671 | 2.000 | 2.390 | 2.660 | 3.460 |
| 73 | 1.293 | 1.666 | 1.993 | 2.379 | 2.645 | 3.429 |
| 80 | 1.292 | 1.664 | 1.990 | 2.374 | 2.639 | 3.416 |
| 100 | 1.290 | 1.660 | 1.984 | 2.364 | 2.626 | 3.390 |
| 199 | 1.286 | 1.653 | 1.972 | 2.345 | 2.601 | 3.340 |
| 1,000 (≈ z) | 1.282 | 1.646 | 1.962 | 2.330 | 2.581 | 3.300 |
Table 2. Upper-tail areas P(Tdf > t), for bracketing a P-value
For a two-sided P-value, double the area. Example: t = 2.53 with df = 29 sits between 2.5 (area 0.0092) and 3.0 (area 0.0027), so the two-sided P is between 0.005 and 0.018; the spreadsheet gives 0.017. The df rows are the ones the course’s examples use: 6 (CDC slope, 8 years), 9 and 18 (music experiment), 17 (CDC paired changes), 29 (samples of 30), 53 and 56 (Welch df for the era comparisons), 73 (75-year regressions).
| df t | 1.0 | 1.5 | 2.0 | 2.5 | 3.0 | 3.5 | 4.0 | 5.0 |
|---|---|---|---|---|---|---|---|---|
| 6 | 0.1780 | 0.0921 | 0.0462 | 0.0233 | 0.0120 | 0.0064 | 0.0036 | 0.0012 |
| 9 | 0.1717 | 0.0839 | 0.0383 | 0.0169 | 0.0075 | 0.0034 | 0.0016 | 0.0004 |
| 17 | 0.1657 | 0.0760 | 0.0309 | 0.0115 | 0.0040 | 0.0014 | 0.0005 | 0.0001 |
| 18 | 0.1653 | 0.0755 | 0.0304 | 0.0112 | 0.0038 | 0.0013 | 0.0004 | 0.0000 |
| 29 | 0.1628 | 0.0722 | 0.0275 | 0.0092 | 0.0027 | 0.0008 | 0.0002 | 0.0000 |
| 53 | 0.1609 | 0.0698 | 0.0253 | 0.0078 | 0.0021 | 0.0005 | 0.0001 | 0.0000 |
| 56 | 0.1608 | 0.0696 | 0.0252 | 0.0077 | 0.0020 | 0.0005 | 0.0001 | 0.0000 |
| 73 | 0.1603 | 0.0690 | 0.0246 | 0.0073 | 0.0018 | 0.0004 | 0.0001 | 0.0000 |
| 1000 | 0.1588 | 0.0670 | 0.0229 | 0.0063 | 0.0014 | 0.0002 | 0.0000 | 0.0000 |
Table 3. Chi-square critical values (upper tail) and the mean of each distribution
The P-value of a chi-square test is always the upper-tail area. A statistic near the df is unremarkable; a statistic above the 0.05 column is significant at α = 0.05.
| df | Tail 0.10 | Tail 0.05 | Tail 0.01 | Tail 0.001 | Mean |
|---|---|---|---|---|---|
| 1 | 2.71 | 3.84 | 6.63 | 10.83 | 1 |
| 2 | 4.61 | 5.99 | 9.21 | 13.82 | 2 |
| 3 | 6.25 | 7.81 | 11.34 | 16.27 | 3 |
| 4 | 7.78 | 9.49 | 13.28 | 18.47 | 4 |
| 5 | 9.24 | 11.07 | 15.09 | 20.52 | 5 |
| 6 | 10.64 | 12.59 | 16.81 | 22.46 | 6 |
| 7 | 12.02 | 14.07 | 18.48 | 24.32 | 7 |
| 8 | 13.36 | 15.51 | 20.09 | 26.12 | 8 |
| 9 | 14.68 | 16.92 | 21.67 | 27.88 | 9 |
| 10 | 15.99 | 18.31 | 23.21 | 29.59 | 10 |
| 11 | 17.28 | 19.68 | 24.72 | 31.26 | 11 |
| 12 | 18.55 | 21.03 | 26.22 | 32.91 | 12 |
Table 4. Standard normal upper tails for tests
| z | 1.00 | 1.28 | 1.50 | 1.65 | 1.96 | 2.00 | 2.24 | 2.50 | 2.58 | 3.00 | 3.50 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Upper tail | 0.1587 | 0.1003 | 0.0668 | 0.0500 | 0.0250 | 0.0228 | 0.0125 | 0.0062 | 0.0050 | 0.0013 | 0.0002 |
Two-sided P = 2 × upper tail. For z = 2.00, 2 × 0.0228 = 0.0455. The full table of Φ(z) is in the Lesson 1.2 reading.
How to bracket, and how to report
Find your df row. Move right until the table value exceeds your statistic; your tail area is between that column’s tail and the previous one’s. Report the bracket (“0.01 < P < 0.02”), then the exact value from technology. If your statistic is beyond the last column, report “P < 0.001” and stop; the exact size of a very small P-value carries no extra meaning, as Lesson 3.2 says.
Sources: Python documentation, the statistics module (NormalDist) (docs.python.org)