find and fix common errors in sampling and in finding and comparing measures of center and spread, including choosing mean or median when there is an outlier.
Phase 3: Make Sense
Do Now. Sample data, minutes to get to school for 9 students: 8, 10, 12, 12, 15, 15, 18, 20, 70. One student lives far away. Find the mean and the median. Which one better describes a typical trip?
What does the 70 do to the mean and the median?
The reading opens over the lesson; close it to come back to exactly where you were.
Detectives check the work. Today we collect the mistakes people make most often with surveys and data, and we fix them. Each one has a test you can run on your own work.
Mistake 1: a median from unordered data. Sample data: 15, 8, 70, 12, 20. Someone picks the middle number as written, 70. Fix: order first: 8, 12, 15, 20, 70, so the median is 15.
Mistake 2: signed differences for the MAD. Our reading shows that differences from the mean always add to zero. Sample data: 15, 22, 27, 30, 31, mean 25: differences −10, −3, 2, 5, 6 add to 0. Someone concludes “MAD = 0, no spread.” Fix: use distances 10, 3, 2, 5, 6; sum 26; MAD = 26 ÷ 5 = 5.2 days.
Worked example 1: the outlier trip. Sample data from the Do Now. Step 1: sum = 180, mean = 20 minutes. Step 2: the 5th of 9 ordered values is 15, so median = 15. Step 3: without the 70, the other 8 trips have mean 110 ÷ 8 = 13.75. Step 4: one value moved the mean by more than 6 minutes, while the median stayed near the middle. Step 5: for this skewed data, report the median, 15 minutes, with the IQR.
Worked example 2: a sample that cannot be scaled. Sample situation: Kiara surveys the 30 students in her robotics club; 24 want a science booth at the carnival. She reports “24/30 = 80%, so 384 of our 480 students want it.” Step 1: the arithmetic is right: 0.8 × 480 = 384. Step 2: the sample is a convenience sample of club members, who like science more than most. Step 3: fix: draw a random sample from the whole school before scaling.
Mistake 3: comparing centers only. “Group B’s mean is 3 points higher, so group B is better.” Fix: check the spread. If the MAD is 6, the gap is only half a MAD.
Mistake 4: treating an estimate as exact. Fix: give a range, as in lesson 2.10.
For each item, find the error or choose the fix. Explain briefly.
Be the data detective.
Sample data: 4, 23, 28, 31, 131 (the CDC ages with an outlier). Find the mean and the median. Which would you report, and why?
Sample data: 15, 8, 70, 12, 20. Someone says the median is 70. What did they forget?
Choose one mistake from today. Write how you will check our School Survey Report so it does not happen there.
We know the traps. Next we use box plots to compare two groups at a glance.
Exit ticket: when should our report use the median instead of the mean? Give an example.
Pacing: Do Now 5 min, reading 7, mini lesson 10, guided practice 13, independent practice and exit ticket 5.
Accuracy: the definition of central location, choice by shape and use, the median example 4, 23, 28, 31, 32 (median 28, still 28 with 131), the mean example 27, 31, 15, 30, 22 (125, 25.0), the centering property, the 131 outlier (45.0) and the median for skewed data are from the CDC lesson page opened 2026-11-20. Note: the CDC page replaces a different value with 131 in each example (32 for the median example, 31 for the mean example); the lesson keeps those separate. Checked: trip data sum 180, mean 20, median 15, without 70 mean 13.75; distances 10, 3, 2, 5, 6 sum 26, MAD 5.2; 0.8 × 480 = 384; 3, 4, 4, 5, 5, 6, 48 sum 75, mean 10.7, median 5; 4, 23, 28, 31, 131 sum 217, mean 43.4, median 28.
Likely misconceptions: each error above is a real student error we re-derived: picking the written middle of unordered data gives 70; adding signed differences gives 0; scaling the club sample gives 384.
Supports: an error-analysis frame: what they did, why it is wrong, the fix, the right answer. Stretch: invent a data set where the mean and median are equal, then add one outlier and describe what changes.
Answer guide: CDC ages with 131: mean 43.4, median 28; report the median because the outlier pulls the mean above four of the five values. Reply to the classmate: 3 ÷ 6 = 0.5 MAD, so the groups overlap a lot and we cannot say group B is better.
Centers for Disease Control and Prevention, Principles of Epidemiology, Lesson 2, Section 6: Measures of Central Location · U.S. government work, public domain
The CDC’s lesson says a measure of central location gives one value that sums up a whole set of data. Which measure to use depends on the shape of the data and how the number will be used.
The median is the middle value of data put in order. The lesson’s example uses five ages: 4, 23, 28, 31 and 32. The median is 28, with two values below it and two above. If the 32 were 131 instead, the median would still be 28. The median is not much affected by one or two extreme values, called outliers.
The mean is found by adding the values and dividing by how many there are. For incubation periods of 27, 31, 15, 30 and 22 days, the sum is 125 and the mean is 25.0 days. The lesson shows a balance property: if you subtract the mean from each value, the differences add to zero.
Because the mean uses every value, one extreme value can change it a lot. If the 31 were 131, the mean would jump to 45.0 days, larger than every value except the outlier. The lesson says the median is a good choice for skewed data, and the mean works best for data that are balanced.
Source: Centers for Disease Control and Prevention, Principles of Epidemiology, Lesson 2, Section 6: Measures of Central Location (archive.cdc.gov)