Water for Schools
Menu

Classroom activity Grade 7 Science

Redfern Reservoir Sampling: A Representative-Sample Simulation

A grade 7 science activity where students draw bead samples from a model reservoir to see why water monitoring needs random, representative samples, not one biased scoop.

Objectives

  1. Students will explain why a water-quality sample must be representative to support a valid conclusion about a whole reservoir.
  2. Students will draw multiple random samples and describe how their estimates of a contaminated fraction vary.

Vocabulary

Key words: sample, population, representative, random sample, proportion

Teacher overview

This is a small-group simulation for groups of three or four students. Each bag of 100 beads is a model of a fictional water body, Redfern Reservoir, where some sampling points have a contaminant over a safe benchmark, shown by red beads, and the rest are fine, shown by blue beads. Groups cannot see inside the bag, so they must estimate the contaminated fraction the way a real monitoring team does: by taking samples. The activity shows directly why a random, representative sample supports a fair conclusion while a biased scoop from one spot does not.

Activity Directions

  1. Setup (7 min): Before class, fill each group's bag with 100 beads, 20 red and 80 blue, and mix them. Do not tell students how many red beads are inside. Hand each group one bag, one cup, and a recording handout. Explain that red beads mark points where a contaminant is over the benchmark, and the group's job is to estimate what fraction of the reservoir is over the benchmark.
  2. Random sampling rounds (12 min): Without looking, a group member shakes the bag, draws exactly 10 beads into the cup, and the group records how many are red. They return the 10 beads, mix, and repeat until they have five samples. For each sample, they write the red count out of 10 as a proportion, such as 3 out of 10 equals 0.3.
  3. Biased sample (6 min): Now have groups pour the beads onto a tray and push most of the red beads into one corner to model a contaminant concentrated near the inlet, then scoop 10 beads only from that corner. Record the red count. Explain that this models testing water from only one spot near the inlet instead of across the whole reservoir.
  4. Class pooling (8 min): Collect every group's random-sample red counts on the board. Add all the red beads drawn and divide by the total beads drawn across the class to get a pooled estimate of the contaminated fraction.
  5. Reveal and debrief (7 min): Tell students the true fraction is 20 red out of 100, or 0.20. Compare the pooled random estimate, which should land near 0.20, with the single biased scoop. Use the debrief questions on the handout to connect the result to real monitoring.
Redfern Reservoir Sampling: A Representative-Sample Simulation — Grade 7 Science
Name: ____________________________ Date: ______________

Student Handout

Record your group's samples and answer the debrief questions.

  • Random sample red counts (out of 10): Sample 1 ___, Sample 2 ___, Sample 3 ___, Sample 4 ___, Sample 5 ___.
  • Biased scoop red count (out of 10): ___.

Debrief questions:

  1. Look at your five random estimates. How much did they vary from lowest to highest?
  2. Was your biased scoop's estimate higher or lower than most of your random estimates? Give one reason a scoop from one corner could be misleading.
  3. Why does pooling many samples across the class usually give a more reliable estimate of the true fraction than a single sample does?
  4. A water utility must decide where to collect samples across a reservoir. Using this activity, explain why testing only one spot could give the wrong picture of the whole reservoir.

From Water for Schools (waterforschools.org). This material is in the public domain (CC0) and may be copied, edited, and shared freely.

Answer Key
  1. Answer: Estimates will vary from sample to sample. Accept any honest range the group recorded; most groups will see their five proportions spread by a few tenths, for example from 0.1 to 0.4, around the true 0.20.
  2. Answer: The biased scoop is usually too high because the red beads were bunched in one corner instead of spread through the reservoir, so that corner holds far more red beads than the reservoir as a whole. Accept any reason that names the single location or the uneven spread of the red beads.
  3. Answer: Pooling combines many draws into a much larger sample, so the random ups and downs of small samples cancel out and the pooled estimate usually lands closer to the true 0.20 fraction than a typical single sample.
  4. Answer: One spot may not represent the whole reservoir; a contaminant could be concentrated near an inlet or spread unevenly, so testing only there could miss a problem elsewhere or overstate one. A representative plan samples many points across the reservoir.

Differentiation

Support

Pre-draw the recording table on the board and complete one random sample together as a class before groups work on their own. Let students report red counts as simple fractions out of 10 rather than decimals.

Extension

Have groups draw five more random samples and add them to their own pooled estimate, then write a sentence comparing their ten-sample estimate to their five-sample estimate and to the true fraction of 0.20.

ELL supports

Pre-teach "sample," "random," "biased," and "representative" with a picture of one scoop from a corner next to many scoops from across a lake. Provide sentence frames for the debrief: "My random estimates went from ___ to ___," and "Testing one spot is misleading because ___."