
fit a normal distribution using a data set’s mean and standard deviation, estimate population percentages with the empirical rule and technology, and recognize when a normal model is not appropriate.
Instruction
Many measurements pile up in a symmetric, bell-shaped pattern: most values near the center, fewer and fewer farther out. The normal distribution is the mathematical model of that shape. It is described completely by two numbers, its mean μ (the center) and its standard deviation σ (the typical distance from the mean).
The empirical rule. For any normal distribution, about 68% of values lie within 1 standard deviation of the mean, about 95% within 2, and about 99.7% within 3. Suppose, as an illustrative example, the battery life of a phone model is approximately normal with mean 40 hours and standard deviation 5 hours. Then about 68% of batteries last between 35 and 45 hours, and about 95% between 30 and 50 hours. Because the curve is symmetric, about 2.5% last less than 30 hours.
z-scores. To compare a value with the distribution, measure its distance from the mean in standard deviations: z = (x − μ)/σ. A 52-hour battery has z = (52 − 40)/5 = 2.4, well above typical.
Areas with technology. The proportion of values in an interval is the area under the curve over that interval. Spreadsheets and calculators compute it. In a spreadsheet, NORM.DIST(x, mean, sd, TRUE) gives the area to the left of x:
- P(X < 45) = NORM.DIST(45, 40, 5, TRUE) ≈ 0.8413, about 84%.
- P(35 < X < 50) = NORM.DIST(50, 40, 5, TRUE) − NORM.DIST(35, 40, 5, TRUE) ≈ 0.8186.
- P(X > 52) = 1 − NORM.DIST(52, 40, 5, TRUE) ≈ 0.0082, less than 1%.
- Working backward, the value with 90% below it is NORM.INV(0.9, 40, 5) ≈ 46.4 hours: a warranty for 46.4 hours would be claimed by about 10% of buyers.
Always sketch the curve and shade the region first. The sketch catches the most common error, which is reporting the area on the wrong side.
Fitting a normal model to data (A2.DS.B.8 / A2.DS.B.9). Given a data set, compute its mean and standard deviation and use them as μ and σ. Then check whether the model is reasonable: make a histogram and see whether it is roughly symmetric and single-peaked; compare the actual share of data within 1 and 2 standard deviations with 68% and 95%. If the check passes, you can use the model to estimate percentages for the population the data represent (A2.DS.A.1).
When not to use it. Household incomes are strongly right-skewed: a long tail of very high incomes pulls the mean above the median, and a normal model would predict negative incomes for some households. Wait times, home prices and the number of social media followers are skewed too. A data set with two peaks, such as heights of a group mixing children and adults, is not normal either. Applying the empirical rule to such data gives confident-looking, wrong percentages. Choosing whether to use the tool is as important as using it correctly (MP.5).
📖 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.
Battery life is normal with mean 40 h and SD 5 h (illustrative). About what percent of batteries last between 30 and 50 hours?
Using technology, P(X < 45) ≈ and the 90th percentile is about hours.
Because a data set has a mean and a standard deviation, the empirical rule can always be used to estimate its percentages.
Work with two data sets. Data Set A (illustrative), hand span in cm for 30 students: 17.1, 18.4, 19.0, 19.6, 20.2, 20.5, 20.8, 21.0, 21.3, 21.5, 21.6, 21.8, 22.0, 22.1, 22.3, 22.4, 22.6, 22.8, 23.0, 23.2, 23.4, 23.6, 23.9, 24.2, 24.5, 24.8, 25.2, 25.7, 26.4, 27.3. Data Set B (illustrative), annual household income in thousands of dollars for 30 households: 28, 31, 34, 36, 38, 41, 43, 45, 47, 49, 52, 54, 57, 59, 62, 65, 68, 72, 76, 81, 86, 92, 99, 108, 119, 133, 152, 181, 228, 340. For each set, make a histogram, compute the mean and SD, and compare the actual percentages within 1 and 2 SD with 68% and 95%. Then decide for each whether a normal model is appropriate and justify the decision. Check your work: Set A has mean 22.41 cm and SD 2.33 cm, with 70.0% of values within 1 SD and 93.3% within 2 SD; Set B has mean $85.87 thousand and SD $66.90 thousand, with 90.0% within 1 SD and 93.3% within 2 SD, and mean − 2 SD is −$47.94 thousand, an income no household can have. Choose one route for the pair you analyze; all three earn full credit and ask the same mathematics. Route 1: use Sets A and B exactly as printed, with no internet and nothing to collect. Route 2: replace Set A with 30 hand spans you measure yourself (same hand, fingers spread, thumb tip to little-finger tip, to the nearest 0.1 cm), keeping Set B. Route 3: replace Set B with Set C, the median household income of 30 Missouri counties (the first thirty in alphabetical order) from the Census Bureau’s American Community Survey 2019-2023 5-year estimates, table B19013, retrieved 2026-11-05; the thirty values, with their margins of error and the mean, SD and percentages to check your work against, are in this lesson’s reading on county incomes, keeping Set A. Those are real figures, so cite the table and the retrieval date.
How confident are you that you can fit a normal model, estimate percentages with technology and judge when the model is inappropriate?
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 distribution is approximately normal with mean 70 and standard deviation 8. Find the z-score of 76.4 and the proportion of values below it.
Check your answer
Answer: z = 0.80; proportion below ≈ 0.7881z = (value − mean) / SD. - A distribution is approximately normal with mean 64 and standard deviation 2.8. Find the z-score of 65.2 and the proportion of values below it.
Check your answer
Answer: z = 0.43; proportion below ≈ 0.6659z = (value − mean) / SD. - A distribution is approximately normal with mean 3.1 and standard deviation 0.4. Find the z-score of 3.9 and the proportion of values below it.
Check your answer
Answer: z = 2.00; proportion below ≈ 0.9772z = (value − mean) / SD. - A distribution is approximately normal with mean 100 and standard deviation 15. Find the z-score of 119.0 and the proportion of values below it.
Check your answer
Answer: z = 1.27; proportion below ≈ 0.8974z = (value − mean) / SD. - X is normal with mean 500 and standard deviation 100. Find P(406 < X < 566).
Check your answer
Answer: 0.5718Standardize both ends, then subtract the cumulative areas. - X is normal with mean 170 and standard deviation 9. Find P(168 < X < 183).
Check your answer
Answer: 0.5136Standardize both ends, then subtract the cumulative areas. - X is normal with mean 24 and standard deviation 3. Find P(20 < X < 25).
Check your answer
Answer: 0.5393Standardize both ends, then subtract the cumulative areas. - X is normal with mean 24 and standard deviation 3. Find P(19 < X < 30).
Check your answer
Answer: 0.9295Standardize both ends, then subtract the cumulative areas. - Battery life is normal with mean 40 h and SD 5 h (illustrative). Use the empirical rule to estimate the percent lasting 35-45 h and less than 30 h.
Check your answer
Answer: About 68%; about 2.5%
Application
Use the skill in context. Show your reasoning.
- For the batteries above, find P(X < 45), P(X > 52) and the 90th percentile with technology.
Check your answer
Answer: 0.8413; 0.0082; 46.4 h - A warranty covers batteries that fail before 32 hours. What percent of batteries will be claimed?
Check your answer
Answer: About 5.5% - County household incomes have mean $95,000 and median $78,000. Is a normal model appropriate? Explain.
Check your answer
Answer: No: the mean well above the median signals right skew; a normal model would misstate percentages.
Challenge
Stretch problems. Expect to think before you write.
- Above what battery life are the top 2.5% of batteries? Compare the technology answer with the empirical-rule estimate.
Check your answer
Answer: 49.80 h by technology; about 50 h by the empirical rule (2 SD)
Review
Keep earlier skills sharp.
- Convert 11π/6 radians to degrees.
Check your answer
Answer: 330°Multiply by 180/π. - Evaluate: log10 √10
Check your answer
Answer: 1/2Ask: 10 to what power gives √10?
Normal Areas Without a Calculator, and One Real Data Set to Test the Model On
NOAA NCEI GHCN-Daily, station USW00013994 (St. Louis Lambert International Airport, MO), daily maximum temperature, July 2025; NOAA NCEI U.S. Climate Normals 1991-2020, monthly, same station (July normal maximum 89.6 F); standard normal table computed for this course; retrieved November 5, 2026 · Public domain (works of the United States government), with an SLS original computed table
Why this reading. The lab already gives you two data sets to judge, one roughly normal and one badly skewed, with their means and standard deviations to check against. What it does not give you is (a) any way to find an area under the curve without NORM.DIST, and (b) a data set that came from an instrument rather than from an author. Both are here. Do not re-run Set A and Set B from this reading; they are in the lesson, and their answers are stated there.
Table 1. Standard normal areas (SLS original, computed for this course)
Φ(z) is the area to the left of z under the standard normal curve. For a value x from a normal distribution with mean μ and standard deviation σ, compute z = (x − μ)/σ and look it up.
| z | −3.0 | −2.5 | −2.0 | −1.5 | −1.0 | −0.75 | −0.5 | −0.25 | 0 |
|---|---|---|---|---|---|---|---|---|---|
| Φ(z) | 0.0013 | 0.0062 | 0.0228 | 0.0668 | 0.1587 | 0.2266 | 0.3085 | 0.4013 | 0.5000 |
| z | 0.25 | 0.5 | 0.75 | 1.0 | 1.28 | 1.5 | 1.645 | 1.96 | 2.0 |
| Φ(z) | 0.5987 | 0.6915 | 0.7734 | 0.8413 | 0.8997 | 0.9332 | 0.9500 | 0.9750 | 0.9772 |
| z | 2.4 | 2.5 | 3.0 | Symmetry: Φ(−z) = 1 − Φ(z), so one half of the table would do. | |||||
| Φ(z) | 0.9918 | 0.9938 | 0.9987 | ||||||
Three uses, on the lesson’s own battery example (μ = 40 h, σ = 5 h, illustrative). P(X < 45): z = 1.0, so Φ(1.0) = 0.8413, matching the spreadsheet value the lesson quotes. P(35 < X < 50): Φ(2.0) − Φ(−1.0) = 0.9772 − 0.1587 = 0.8185. P(X > 52): z = 2.4, so 1 − 0.9918 = 0.0082. Working backward for the 90th percentile: find z with Φ(z) = 0.90, which the table puts at z = 1.28, so x = 40 + 1.28(5) = 46.4 hours. Every number the lesson gets from technology comes out of this table too, which means you can check the technology.
Notice the 68–95–99.7 rule falling out of the table rather than being asserted: Φ(1) − Φ(−1) = 0.8413 − 0.1587 = 0.6826; Φ(2) − Φ(−2) = 0.9772 − 0.0228 = 0.9544; Φ(3) − Φ(−3) = 0.9987 − 0.0013 = 0.9974.
Table 2. Daily maximum temperature, St. Louis Lambert International Airport, July 2025 (°F)
Thirty-one real observations from NOAA station USW00013994.
| Day | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Max °F | 86 | 89 | 92 | 93 | 92 | 93 | 89 | 91 | 88 | 92 | 94 | 86 | 84 | 85 | 89 | 95 |
| Day | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | |
| Max °F | 87 | 88 | 88 | 96 | 87 | 93 | 97 | 96 | 95 | 94 | 94 | 95 | 97 | 90 | 79 |
Run the fit, and then run the check
Mean 90.77°F, standard deviation 4.33°F (sample, n − 1), median 92°F, range 79 to 97. The mean sits a little more than a degree below the median, so the pile leans slightly: a few cool days pull the mean down, most of all July 31 at 79°F. Fit a normal model with μ = 90.77 and σ = 4.33 and test it the way A2.DS.B.8 / A2.DS.B.9 asks:
| Interval | Bounds (°F) | Days inside | Actual percent | Normal model predicts |
|---|---|---|---|---|
| μ ± 1σ | 86.45 to 95.10 | 22 of 31 | 71.0% | 68% |
| μ ± 2σ | 82.12 to 99.43 | 30 of 31 | 96.8% | 95% |
| μ ± 3σ | 77.80 to 103.75 | 31 of 31 | 100% | 99.7% |
Close on all three. So the model passes the numeric check; the one day outside 2σ is July 31, at z = (79 − 90.77)/4.33 = −2.72. Now look at the data in order rather than as a pile. Days 23–29 are 97, 96, 95, 94, 94, 95, 97 — a heat wave, seven consecutive days at 94°F or above, all in the upper part of the distribution. Days 12–14 are 86, 84, 85 — three consecutive days in the lower tail. The high values are not scattered through the month; they arrive together.
This is the honest limit of the empirical-rule check. A normal model treats observations as independent draws. July daily maxima are not independent — today’s weather predicts tomorrow’s. The histogram cannot see that, and neither can the percent-within-one-SD test, because both throw away the order. So “a normal model fits this data” is defensible for a question like what fraction of July days exceed 95°F? — the model gives 1 − Φ((95 − 90.77)/4.33) = 1 − Φ(0.98) = about 16%, against an actual 4 days of 31, or 12.9%, which is already a gap worth reporting — and it is simply false for a question like what is the chance of seven days in a row at 94°F or more?, where assuming independence gives an answer far too small.
One month is one month. Thirty-one observations from one July at one airport estimate a mean with a standard error of 4.33/√31 = 0.78°F, so “St. Louis’s July mean maximum is 90.8°F” is really “90.8 ± about 1.6°F for July 2025.” Compare it with NOAA’s 1991–2020 normal for July at the same station, 89.6°F: the difference of 1.2°F is inside that interval, so this July is not evidence of anything unusual. Saying which of your numbers is a statistic about a month and which is an estimate of a parameter is the distinction this unit opened with.
Sources: NOAA NCEI Access Data Service, daily summaries, USW00013994, TMAX, July 2025 (the query used here) (ncei.noaa.gov); NOAA NCEI Access Data Service, monthly normals 1991-2020, USW00013994, MLY-TMAX-NORMAL (ncei.noaa.gov); NOAA NCEI, Daily Summaries (GHCN-Daily) data search (ncei.noaa.gov)
One Hundred Fifteen Missouri Counties: Median Household Income, With the Census Bureau’s Margins
U.S. Census Bureau, American Community Survey 2019-2023 5-year estimates, table B19013 (median household income in the past 12 months, 2023 inflation-adjusted dollars), Missouri counties and St. Louis city, from the ACS Summary File (table-based format); retrieved November 5, 2026 · Public domain (works of the United States government)
Why this reading. Route 3 of the lab replaces Data Set B with real county incomes. Here they are, so nothing has to be looked up: the median household income of every one of Missouri’s 115 county-level areas from the Census Bureau’s American Community Survey, with the Bureau’s own margin of error for each, and the numbers the lab asks you to check (mean, SD, percent within 1 and 2 SD) already worked for a fixed set of 30 counties. Use Set C below in place of Set B; the mathematics is the same.
What the numbers are
Table B19013, Median Household Income in the Past 12 Months (in 2023 Inflation-Adjusted Dollars), from the U.S. Census Bureau’s American Community Survey 2019-2023 5-year estimates, for Missouri’s 114 counties and the independent city of St. Louis, which the Bureau counts as a county equivalent (summary level 050). The estimate is the county’s median: half its households earn more, half less. The margin of error is at the 90% confidence level, as the Bureau publishes it; the median of a county with few households is a small-sample estimate, and its margin is large (Bollinger County: ±$11,309). For all of Missouri the same table gives $68,920 ± $368. Retrieved November 5, 2026 from the Bureau’s ACS Summary File, table-based format (5YRData, file acsdt5y2023-b19013.dat, county names from Geos20235YR.txt).
Set C: the first thirty counties in alphabetical order
The thirty are simply the first thirty of the 115 in alphabetical order, Adair through Dallas, so nobody chose them for how they look. A median household income is a dollar amount; the lab’s Set B is in thousands of dollars, so divide by 1,000 (Adair: $56,583 is 56.583 thousand).
| # | County | Median household income ($) | Margin of error (90%) |
|---|---|---|---|
| 1 | Adair County | 56,583 | ±5,235 |
| 2 | Andrew County | 74,007 | ±5,338 |
| 3 | Atchison County | 59,260 | ±5,314 |
| 4 | Audrain County | 56,232 | ±4,379 |
| 5 | Barry County | 56,611 | ±4,204 |
| 6 | Barton County | 49,503 | ±5,066 |
| 7 | Bates County | 57,914 | ±4,158 |
| 8 | Benton County | 52,200 | ±4,080 |
| 9 | Bollinger County | 57,286 | ±11,309 |
| 10 | Boone County | 69,913 | ±2,041 |
| 11 | Buchanan County | 59,423 | ±2,840 |
| 12 | Butler County | 49,213 | ±3,349 |
| 13 | Caldwell County | 62,520 | ±5,082 |
| 14 | Callaway County | 72,645 | ±2,446 |
| 15 | Camden County | 62,927 | ±3,560 |
| 16 | Cape Girardeau County | 68,912 | ±2,419 |
| 17 | Carroll County | 61,712 | ±2,791 |
| 18 | Carter County | 45,083 | ±7,157 |
| 19 | Cass County | 87,413 | ±3,813 |
| 20 | Cedar County | 45,577 | ±2,949 |
| 21 | Chariton County | 63,307 | ±6,622 |
| 22 | Christian County | 81,245 | ±2,391 |
| 23 | Clark County | 51,458 | ±4,979 |
| 24 | Clay County | 86,150 | ±1,755 |
| 25 | Clinton County | 70,625 | ±3,666 |
| 26 | Cole County | 73,273 | ±3,685 |
| 27 | Cooper County | 67,548 | ±4,278 |
| 28 | Crawford County | 56,345 | ±2,695 |
| 29 | Dade County | 49,899 | ±3,538 |
| 30 | Dallas County | 53,177 | ±3,923 |
The lab’s check, worked for Set C
- Mean. The thirty medians sum to $1,857,961, so the mean of Set C is 1,857,961/30 = $61,932.03, or 61.932 thousand dollars. (This is the mean of thirty county medians, not Missouri’s median household income, which is the $68,920 above.)
- Standard deviation (sample SD, dividing by n − 1 = 29, as in the lesson’s Sets A and B): $11,274.10, or 11.274 thousand dollars.
- Within 1 SD of the mean, that is from $50,658 to $73,206: 20 of 30 counties, 66.7%. The 10 outside are Andrew ($74,007), Barton ($49,503), Butler ($49,213), Carter ($45,083), Cass ($87,413), Cedar ($45,577), Christian ($81,245), Clay ($86,150), Cole ($73,273), Dade ($49,899).
- Within 2 SD, from $39,384 to $84,480: 28 of 30, 93.3%. The two counties outside are Cass ($87,413, 2.26 standard deviations above the mean) and Clay ($86,150, 2.15 above), both suburban Kansas City counties, and both on the high side.
- mean − 2 SD is $39,384: positive here, so unlike Set B this set does not push a normal model below zero income. Ask instead whether the shape looks like a bell: in $10,000 bins the counts are 5 in the $40,000s, 11 in the $50,000s, 7 in the $60,000s, 4 in the $70,000s and 3 in the $80,000s — a peak low down and a tail stretching to the right.
Compare with the empirical rule (68% and 95%) and with the lesson’s Sets A and B, and decide whether a normal model is appropriate for county medians. The median county income of Set C is $59,341.50, below its mean of $61,932, which is the signature of a right skew — milder here than in Set B, and worth saying so.
The other eighty-five
The rest of the table, so that you can use all 115 (mean of the 115 medians $59,503.10, SD $11,477.52; 85 of 115 within 1 SD and 109 within 2 SD) or choose your own thirty by a rule you state. The highest is St. Charles County ($102,912) and the lowest Hickory County ($35,084).
| # | County | Median household income ($) | Margin of error (90%) |
|---|---|---|---|
| 31 | Daviess County | 60,665 | ±4,257 |
| 32 | DeKalb County | 69,093 | ±7,380 |
| 33 | Dent County | 54,306 | ±3,496 |
| 34 | Douglas County | 49,828 | ±5,387 |
| 35 | Dunklin County | 47,368 | ±3,139 |
| 36 | Franklin County | 71,973 | ±2,698 |
| 37 | Gasconade County | 65,096 | ±4,300 |
| 38 | Gentry County | 57,458 | ±3,989 |
| 39 | Greene County | 57,488 | ±1,316 |
| 40 | Grundy County | 53,839 | ±5,582 |
| 41 | Harrison County | 53,364 | ±3,151 |
| 42 | Henry County | 56,621 | ±3,379 |
| 43 | Hickory County | 35,084 | ±4,310 |
| 44 | Holt County | 58,516 | ±2,169 |
| 45 | Howard County | 62,628 | ±4,959 |
| 46 | Howell County | 49,814 | ±3,263 |
| 47 | Iron County | 51,161 | ±8,529 |
| 48 | Jackson County | 67,178 | ±1,127 |
| 49 | Jasper County | 57,525 | ±2,931 |
| 50 | Jefferson County | 80,522 | ±2,296 |
| 51 | Johnson County | 67,123 | ±2,893 |
| 52 | Knox County | 53,333 | ±8,971 |
| 53 | Laclede County | 50,825 | ±2,533 |
| 54 | Lafayette County | 79,091 | ±5,403 |
| 55 | Lawrence County | 55,849 | ±5,275 |
| 56 | Lewis County | 52,340 | ±5,529 |
| 57 | Lincoln County | 85,276 | ±2,307 |
| 58 | Linn County | 61,975 | ±4,188 |
| 59 | Livingston County | 60,148 | ±5,113 |
| 60 | Macon County | 58,318 | ±4,046 |
| 61 | Madison County | 58,435 | ±6,862 |
| 62 | Maries County | 61,359 | ±9,505 |
| 63 | Marion County | 63,908 | ±3,770 |
| 64 | McDonald County | 48,145 | ±3,838 |
| 65 | Mercer County | 60,357 | ±10,789 |
| 66 | Miller County | 55,172 | ±5,736 |
| 67 | Mississippi County | 46,258 | ±4,629 |
| 68 | Moniteau County | 65,161 | ±4,266 |
| 69 | Monroe County | 48,299 | ±4,302 |
| 70 | Montgomery County | 61,042 | ±6,219 |
| 71 | Morgan County | 49,663 | ±4,821 |
| 72 | New Madrid County | 49,237 | ±4,291 |
| 73 | Newton County | 64,583 | ±3,380 |
| 74 | Nodaway County | 55,970 | ±2,917 |
| 75 | Oregon County | 44,259 | ±8,105 |
| 76 | Osage County | 75,355 | ±4,098 |
| 77 | Ozark County | 42,329 | ±5,408 |
| 78 | Pemiscot County | 40,748 | ±4,460 |
| 79 | Perry County | 63,356 | ±3,606 |
| 80 | Pettis County | 60,232 | ±2,670 |
| 81 | Phelps County | 55,427 | ±3,157 |
| 82 | Pike County | 57,572 | ±6,127 |
| 83 | Platte County | 95,748 | ±3,176 |
| 84 | Polk County | 56,686 | ±3,583 |
| 85 | Pulaski County | 66,435 | ±3,925 |
| 86 | Putnam County | 56,875 | ±9,667 |
| 87 | Ralls County | 62,054 | ±5,169 |
| 88 | Randolph County | 55,310 | ±5,921 |
| 89 | Ray County | 72,152 | ±5,449 |
| 90 | Reynolds County | 44,357 | ±5,562 |
| 91 | Ripley County | 43,898 | ±4,901 |
| 92 | Saline County | 56,566 | ±3,401 |
| 93 | Schuyler County | 56,023 | ±2,785 |
| 94 | Scotland County | 67,568 | ±4,430 |
| 95 | Scott County | 58,847 | ±4,433 |
| 96 | Shannon County | 59,000 | ±8,604 |
| 97 | Shelby County | 52,127 | ±4,896 |
| 98 | St. Charles County | 102,912 | ±1,674 |
| 99 | St. Clair County | 46,731 | ±4,770 |
| 100 | St. Francois County | 54,813 | ±3,857 |
| 101 | St. Louis city | 55,279 | ±1,530 |
| 102 | St. Louis County | 81,340 | ±1,143 |
| 103 | Ste. Genevieve County | 61,310 | ±5,784 |
| 104 | Stoddard County | 54,067 | ±4,113 |
| 105 | Stone County | 61,323 | ±2,272 |
| 106 | Sullivan County | 55,500 | ±4,677 |
| 107 | Taney County | 58,258 | ±2,800 |
| 108 | Texas County | 48,055 | ±3,638 |
| 109 | Vernon County | 54,519 | ±5,230 |
| 110 | Warren County | 77,989 | ±4,738 |
| 111 | Washington County | 51,886 | ±4,268 |
| 112 | Wayne County | 43,393 | ±4,123 |
| 113 | Webster County | 69,731 | ±2,598 |
| 114 | Worth County | 46,167 | ±4,590 |
| 115 | Wright County | 47,304 | ±3,000 |
What to do
Make a histogram of Set C in thousands of dollars, compute the mean and SD yourself and check them against the values above, count the percentages within 1 and 2 SD, and write the decision and its justification the lab asks for. Then say what the margins of error add: a county’s median is an estimate, and for the smallest counties it is very uncertain, so two counties whose medians differ by less than their margins may not differ at all (Adair, $56,583 ± $5,235, and Barry, $56,611 ± $4,204, are an example).
Sources: U.S. Census Bureau, ACS 2023 Summary File, table-based format, 5-year data, table B19013 (www2.census.gov); U.S. Census Bureau, ACS 2023 5-year geography file (Geos20235YR.txt) (www2.census.gov)