choose a random sample with numbered cards or random digits, and explain why skipping repeats and out-of-range numbers keeps the method fair.
Phase 2: Investigate
Do Now. Sample school: there are 480 students, numbered 001 to 480. You want a random sample of 5. Write two ways you could let chance pick the 5 numbers.
A random digit list gives the three-digit number 902. What should you do with it?
The reading opens over the lesson; close it to come back to exactly where you were.
Step one is a list. To choose at random, you first need a list of everyone in the population, like NAEP’s list of schools. Statisticians call it a sampling frame. If someone is missing from the list, they have no chance to be chosen, and the sample can be biased before it even starts.
Step two is a number. Give every member a number with the same number of digits: 001, 002, … 480. Then let chance pick numbers. Tools include numbered cards drawn from a bag, a spinner, number cubes, a random digit list, or a calculator’s random number key.
Rules for random digits. Read the digits in groups of the right size (three digits for 001-480). Skip any number that is not on the list, like 000 or 902. Skip a number that has already been chosen, so no one is counted twice. Stop when you have enough.
Worked example 1: reading a digit list. Sample digits: 317 902 045 317 488 126 733 260. Aaliyah needs 4 students from 001-480. Step 1: 317 is on the list, so take it. Step 2: 902 is too big, skip. Step 3: 045 is on the list, take it. Step 4: 317 again is a repeat, skip. Step 5: 488 is too big, skip. Step 6: 126 is on the list, take it. Step 7: 733 is too big, skip. Step 8: 260, take it. Her sample is students 317, 045, 126 and 260.
Worked example 2: a smaller frame. Sample class: a homeroom has 24 students, numbered 01 to 24. Diego needs 3 using two-digit groups: 73 08 51 19 08 22. Step 1: 73 skip. Step 2: 08 take. Step 3: 51 skip. Step 4: 19 take. Step 5: 08 repeat, skip. Step 6: 22 take. Sample: 08, 19, 22. Each of the 24 students had the same chance, because every two-digit number from 01 to 24 was equally likely to come up.
Why the rules are fair. Skipping is not cheating; it keeps every student’s chance the same. Changing 902 into 480, or picking a friend when a number “looks wrong,” would let a person make the choice.
Today’s hands-on. Each team gets a bag of 30 numbered cards and a printed random digit strip. You will choose samples both ways and compare the work.
Hands-on. Your bag holds cards numbered 01-30. (1) Draw 5 cards without replacement and record them. (2) Then use this sample digit strip, two digits at a time, for another sample of 5: 48 07 93 22 07 15 61 30 00 11.
Practice the method on paper.
Sample digits: 512 064 377 064 291 805 448 130. Choose 4 students from a list numbered 001-480. Show which numbers you take and which you skip, and why.
Sample school: 480 students, random sample of 40. What percent of the school is in the sample? (Round to the nearest tenth of a percent.)
Explain why the Census Bureau and NAEP both start with a complete list before they choose anyone.
We can draw a random sample by hand. Next time we use a random sample to estimate something about the whole school.
Exit ticket: list the three skip rules for random digits in your own words.
Pacing: Do Now 5 min, reading 7, mini lesson 10, card and digit activity with guided practice 13, independent practice and exit ticket 5.
Materials: per team, 30 cards numbered 01-30 in a bag and a printed copy of the digit strip. The digit strings here are sample strings written for the lesson, not from a published table.
Accuracy: the Common Core of Data as the public-school frame, the Private School Survey for private schools, checking data up to three years old with states, random selection of students within schools in up to two subjects, and weighting are from the NAEP page opened 2026-11-20. Checked: strip 48 07 93 22 07 15 61 30 00 11 gives 07, 22, 15, 30, 11; 480 ÷ 60 = 8; 40/480 = 8.33%; independent digits 512 064 377 064 291 805 448 130 give 064, 377, 291, 448 (skip 512, the repeat 064 and 805).
Likely misconceptions: reading digits one at a time for a 3-digit list; keeping a repeat; thinking two different random samples mean one is wrong.
Supports: pre-split digit strips with boxes for each group. Stretch: how would you use one number cube to choose 1 of 6 teams fairly? How about 1 of 5? (Roll again on a 6.)
Answer guide: different random samples are both fair; they differ by chance, which we study next. Both NAEP and the Census need a list so that every member can be given a chance.
National Center for Education Statistics, NAEP Assessment Process: Select the Participants · U.S. government work, public domain
Before NAEP can choose anyone, it needs a complete list. For public schools, it uses a list called the Common Core of Data, a yearly list of the public schools in each state. For private schools it uses a separate survey of private schools. A list like this is called a sampling frame.
Because school lists can be up to three years old, NAEP checks its school data with each state’s department of education. It then puts schools in an ordered list, sorted by kind of place and other traits, and draws a separate sample from each group.
Next comes the student step. Within each chosen school and grade, students are chosen at random to take part, in up to two subjects. Chance, not a person, makes the choice, and every student in the grade has a chance.
Because the students who take part are only part of the whole group, NAEP adjusts, or weights, the results so they describe the full population fairly. Some groups, such as very small schools, are planned to be included so they are not missed.
Source: National Center for Education Statistics, NAEP Assessment Process: Select the Participants (nces.ed.gov)