Confounded Variables in an Eighth-Grade Insulation Investigation
Ms. Chen's eighth-grade science class is investigating thermal energy transfer. Each of six groups designed a cup to keep 200 mL of warm water hot for ten minutes, choosing among foam, plastic, and metal cups, with or without lids. Four groups changed both the cup material and the starting water temperature; two groups also changed the water volume and added a lid. During data sharing, group 3 reports that its foam cup started at 60 degrees Celsius and lost 8 degrees, while group 5 reports that its plastic cup started at 75 degrees Celsius and lost 15 degrees. Several students conclude that foam is the best insulator. Ms. Chen has one class period left before lab reports are due and does not want to simply tell students the answer. Safety is not an issue, and all groups have usable data, but the comparisons are confounded.
1. What should Ms. Chen do in the remaining period? Justify your decision in terms of both inquiry learning and the accuracy of the scientific claim students are making.
Read the worked answer
She should not let the confounded comparison stand as the basis for a causal claim, and she should not replace the investigation with a lecture. The remaining period should become a data audit. Each group lists every variable it changed and every variable it held constant, then marks which comparisons are valid. Group 3 versus group 5 is invalid for insulation because starting temperature differs; group 3's smaller loss could be due to the cooler starting water rather than the foam.
This preserves the inquiry goal because students discover the problem in their own data rather than being told. It also teaches the difference between a result and a warranted conclusion, which is the core idea behind controlled experiments. Ms. Chen can ask each group one question: 'What single comparison would let you claim that material caused the difference?' Groups then propose a fair test. The lab report can be revised to state what the data do and do not support, which is more scientifically honest than a false conclusion. If time permits, one demonstration with identical starting temperatures can be run, but the audit is the priority.
2. Design the fair-test protocol students should use if they revise the investigation. Specify the variables to control and how temperature should be measured.
Read the worked answer
Students should compare cup materials only, holding everything else constant: same volume (200 mL), same starting temperature (for example, 70 degrees Celsius, checked with the same thermometer), same lid condition, same cup size and shape, same room location, same elapsed time (10 minutes), and the same measurement schedule. Each cup should be measured every 2 minutes with the same thermometer inserted to the same depth, and the lid should be replaced quickly to avoid heat loss during measurement.
The protocol should also specify the comparison: the difference between starting and ending temperature for each material, with at least two trials per material if time allows, and a class mean. Students should write a 'fair test' checklist before running the trial and have another group verify it. This matters because the original design confounded material, starting temperature, volume, and lid use. Holding starting temperature constant also removes a subtle problem: a hotter cup loses heat faster, so comparing a 75-degree plastic cup with a 60-degree foam cup stacks the result against plastic before the material effect is even considered.
3. How should Ms. Chen grade the lab reports so that students are accountable for understanding variables without being penalized for an initial design flaw?
Read the worked answer
She should separate the design from the reasoning and allow revision. The rubric can have two strands. The design strand rewards identifying confounds and proposing a controlled comparison. The explanation strand rewards matching the claim to the evidence: a student who writes 'foam is best because it lost only 8 degrees' without noting the different starting temperatures should score low on evidence-claim alignment even if the arithmetic is correct. A student who writes 'our data cannot isolate material because starting temperatures differed; our revised test held temperature constant and found ___' should score high.
A short independent item can check the concept without punishing the group's original choices: give two hypothetical groups, one that changed only the cup material and one that changed material and starting temperature, and ask which comparison supports the claim and why. That item measures control of variables directly. Allowing one revision is also fair, because the point of the investigation is to learn how to make a warranted comparison, not to perform a flawless procedure on the first attempt. The final grade should reflect the revised reasoning.