The Paper Towel Test With Three Variables
Four fourth-grade groups investigate the question 'Which brand of paper towel absorbs the most water?' Each group gets three brands, a graduated cylinder, a balance, and paper towels. Group C tears Brand A into a full sheet, Brand B into a half sheet, and Brand C into a quarter sheet. They pour 50 mL onto each, wait 10 seconds, 30 seconds, and 60 seconds respectively, then squeeze each towel into a beaker and measure the water. They find about 35 mL from A, 20 mL from B, and 9 mL from C. Group C concludes that Brand A is the most absorbent and recommends it to the class. Their data table lists brand, water poured, soak time, and water recovered, but not towel size. The teacher has one class period for revision before groups present.
1. Identify the flaw in Group C's investigation and explain why their conclusion is not supported.
Read the worked answer
Group C changed three things at once: towel size, soak time, and, depending on how the pour was done, the amount of water actually applied. Brand A was a full sheet, B a half sheet, and C a quarter sheet; A also soaked six times as long as C. Any difference in recovered water could be caused by size, time, or brand. The brands are confounded with the other variables, so the comparison is not a fair test of absorbency.
The measurement also needs scrutiny. Water recovered by squeezing depends on how hard and how long each towel is squeezed, and if the poured volume varied, the recovered amount partly reflects input rather than absorption. The missing size column in the data table is a clue that the group did not identify size as a variable. Their recommendation may turn out to be right, but this investigation cannot establish it.
2. Design a revised procedure the group could carry out or describe in the remaining period.
Read the worked answer
Hold everything constant except brand. Cut each brand to the same size square, for example 10 cm by 10 cm. Pour the same volume, 50 mL, from the same height, and let each towel soak for the same 30 seconds. Lift each with tongs and let it drip for a fixed 15 seconds, then squeeze using the same method, such as rolling and pressing with a specified mass for 10 seconds. Run three trials per brand in randomized order and average the results.
A stronger measure is mass difference: weigh each dry square, soak it, drip it for the fixed time, and weigh again; the gain in grams is the water absorbed. That avoids squeeze-force variability. Define absorbency operationally at the start, for example 'grams of water held after 30 seconds of soaking and 15 seconds of dripping.' Record all controlled variables in the table so another group could replicate the test.
3. What should the teacher do during the revision period, and why not simply correct the group?
Read the worked answer
Ask Group C to compare their procedure with another group's and list every difference they notice. Post a class chart of variables: size, volume, soak time, drip time, squeeze method, trials. Then ask, 'Which differences could explain your results besides the brand?' The group should identify size and time themselves. If time is short, have them rewrite the procedure and run one quick trial, then present the revised plan and note what they would still need to test.
Simply telling them the answer fixes one procedure but leaves the idea of a fair test untouched. Productive struggle with the design, followed by revision, is what builds the concept of a controlled experiment. The teacher should also publicly value what the group did well, especially the careful measurement and the clear data table, so the revision is about design, not about being wrong.