Project on the screen the 2008 Chinese melamine milk scandal headline and a photo of contaminated powdered formula. Tell students: 'Melamine was added to milk to fake high protein readings on a nitrogen-based test. Six infants died, 54,000 were hospitalized. The scandal was uncovered because one lab tech ran a CONTROL sample — pure unadulterated milk — and noticed her result didn't match what the protein test said it should be.' Ask: 'What single experimental design choice cracked this case open?' Cold-call 2–3 students. Steer them to the word 'control.' Then say: 'Today you're going to design experiments where, if you skip the control or confuse your variables, you would be the one who missed it.' Transition by writing the objective on the board.

1. 8m

   Science is iterative, not a 7-step list

   Draw on the board two diagrams side-by-side: LEFT — a vertical 7-step flowchart labeled 'Question → Research → Hypothesis → Experiment → Data → Conclusion → Report' with a big red X through it. RIGHT — a messy cyclical web with arrows going every direction between 'Observation,' 'Question,' 'Hypothesis,' 'Test,' 'Revise,' 'New Question.' Tell students: the textbook 7-step version is a fiction taught to elementary kids. Real science loops back. When Marie Curie isolated radium, she revised her hypothesis dozens of times. Emphasize: a hypothesis that is NOT supported is still a successful experiment — it eliminated a wrong idea. Ask: 'If your data disprove your hypothesis, did your experiment fail?' Expected pushback: yes. Correct them: no — disconfirming data is valid data.

2. 9m

   Variables: independent, dependent, controlled

   Project on the screen the variables-table template: three columns (Independent / Dependent / Controlled) with rows 'What it is' and 'How it's measured.' Fill it in live with this example: 'Does the concentration of HCl affect how fast Mg ribbon dissolves?' Independent = HCl concentration (0.5 M, 1.0 M, 2.0 M); Dependent = time to fully dissolve (seconds, stopwatch); Controlled = mass of Mg ribbon (0.10 g), volume of HCl (20.0 mL), temperature (22 °C), surface area of ribbon. Then show a side-by-side diagram of two beakers — one labeled 'control: 20 mL distilled water + 0.10 g Mg' and one 'experimental: 20 mL 1.0 M HCl + 0.10 g Mg' — and circle the single varied factor in red. Check for understanding: ask a student to identify the dependent variable if you swapped the question to 'Does temperature affect Mg dissolution rate?'

3. 8m

   Observation vs inference, qualitative vs quantitative

   Draw on the board a T-chart. Left column 'Observation,' right column 'Inference.' Drop a piece of zinc into dilute HCl in a small test tube at the demo bench. As bubbles form, narrate: 'Observation: small bubbles forming on the zinc surface, gas collecting at the top, tube feels warm.' Then say: 'Inference: hydrogen gas is being released and the reaction is exothermic.' Make students articulate the difference — observations are what you record, inferences are interpretations you'll defend later. Then make a second T-chart: Qualitative ('bubbles, warm, gray solid disappearing') vs Quantitative ('0.15 g Zn, 5.0 mL of 1.0 M HCl, ΔT = +4.2 °C in 90 s'). Stress: honors lab notebooks need BOTH columns populated.

4. 5m

   Sources of error — be specific

   Write on the board: 'BANNED PHRASE: human error.' Tell students this phrase will lose them points all year. A real source of error is specific and traceable: 'cross-contamination from reusing the spatula between the NaHCO₃ and the CaCO₃ samples,' or 'parallax when reading the meniscus of the graduated cylinder,' or 'the balance was not tared between trials.' Ask students to flip to the back of their lab notebook and write three categories: instrumental, procedural, environmental — they'll fill these in during today's lab.

1. 30m

   Protocol Autopsy: Critique a Flawed Experiment

   Hand each pair the following flawed protocol: 'Jamie wants to know if caffeine affects plant growth. He puts a bean seed in a cup of soil on his windowsill and waters it with 50 mL of coffee every morning for 2 weeks. He measures the height with a ruler on day 14 and gets 8 cm. He concludes that caffeine helps plants grow.' Project on the screen the blank variables-table template (Independent / Dependent / Controlled with rows for 'What it is' and 'How it's measured'). Give pairs 12 minutes to: (1) fill in the variables table for Jamie's experiment as stated, (2) list every design flaw they can find, (3) rewrite the protocol to fix the flaws. Walk around and check: are they catching n=1, no control group, confounding variables (light, soil, water volume), no replicate trials, no baseline height, single measurement point? Expected flaws to surface: no control plant watered with plain water; n=1 so no way to assess variability; coffee contains more than caffeine (sugars, acids, oils); no measurement of starting height; only one measurement at day 14. After 12 minutes, regroup. Cold-call pairs to share one flaw each on the board until exhausted. Then give 10 minutes for pairs to write a fixed protocol with at least: a control group, n≥5, isolated variable (use pure caffeine solution, not coffee), measurements every 2 days, controlled light/soil/water volume. Last 3 minutes: two pairs present their revised protocol; class votes on which is more rigorous and why.

   Materials

   • Printed protocol handout (one per pair)
   • Red pens
   • Variables table template (blank)

   Example outputs

   • Variables table: IV = type of liquid given (coffee vs water); DV = plant height on day 14 (cm, ruler); Controlled = soil type, sunlight, water volume, seed variety, starting height. Flaws: n=1, no control, coffee ≠ pure caffeine (confounded with sugar/acid/oils), no starting height recorded, only one time point.
   • Revised protocol: 10 identical bean seeds, 5 watered with 50 mL of 0.05% caffeine solution daily, 5 watered with 50 mL distilled water daily (control); same potting soil, same windowsill, height measured with caliper every 2 days for 14 days; record starting height on day 0; report mean ± standard deviation.
2. 42m

   Mystery White Powder Identification LabLab

   Tell students: each group gets 5 unknown white powders labeled A–E. They are some combination of NaCl, sucrose, NaHCO₃, cornstarch, and CaCO₃. Available tests: solubility in water, reaction with 1.0 M HCl (look for bubbling = CO₂), reaction with iodine (turns blue-black with starch), pH of the dissolved powder, conductivity of the solution. Project on the screen the side-by-side control vs experimental diagram from the intro and remind them: every test needs a control (a blank spot with just water + reagent). PHASE 1 (8 min — design): each group must submit a written test plan to the teacher before getting reagents. Plan must include: hypothesis for each unknown's identity, which tests they'll run, what controlled variables they'll hold constant (volume of powder, drops of reagent), and their control. Walk around and check plans — reject any without a control or with vague test sequences. PHASE 2 (22 min — execute): groups run tests on spot plates, record BOTH qualitative observations (fizzing, color change, cloudy) AND quantitative data (pH value, drops of HCl until bubbling stops). Lab notebook must have a data table. Stop the class at minute 20 to ask 'Has anyone gotten a result they didn't expect? What's your next step?' — reinforces iterative science. PHASE 3 (12 min — conclude): groups write a conclusion identifying each powder, citing the SPECIFIC evidence (e.g., 'Powder C is NaHCO₃ because it fizzed with HCl AND had a pH of 9'), and list at least 3 specific sources of error (instrumental, procedural, environmental). Spot-check: did they distinguish CaCO₃ from NaHCO₃? Both fizz with HCl; the discriminator is solubility (NaHCO₃ dissolves, CaCO₃ doesn't) and pH (NaHCO₃ ~9, CaCO₃ neutral until acid added).

   Materials

   • 5 labeled unknowns A–E (NaCl, sucrose, NaHCO₃, cornstarch, CaCO₃) — 2 g each per group
   • Spot plates
   • Distilled water in wash bottle
   • 1.0 M HCl in dropper bottles
   • Iodine solution (Lugol's) in dropper bottles
   • pH paper (range 1–14)
   • Conductivity testers (or 9V battery + LED setup)
   • Spatulas (one per powder)
   • Lab notebooks
   • Variables table template
   • Goggles, aprons, nitrile gloves

   Example outputs

   • Powder A is sucrose: dissolved in water (clear solution), no reaction with HCl, no color change with iodine, pH ≈ 7, did NOT conduct electricity. Sources of error: spatula was not rinsed between A and B (possible contamination); pH paper only resolves to ±0.5 units; one drop of HCl varied in volume.
   • Powder C is NaHCO₃: dissolved in water, vigorous fizzing with HCl (CO₂ gas), no color change with iodine, pH ≈ 9, solution conducted electricity. Distinguished from CaCO₃ because CaCO₃ remained cloudy in water (did not fully dissolve) and its pH was closer to 7 before adding acid.

1. A student tests whether the amount of baking soda affects how high a vinegar-baking soda volcano erupts. She uses 5 g, 10 g, and 15 g of baking soda, each with 50 mL of 5% vinegar in identical 250 mL flasks at room temperature, and measures eruption height with a meter stick. Identify the independent variable, the dependent variable, and TWO controlled variables.

   short answerIndependent variable: mass of baking soda (5, 10, 15 g). Dependent variable: eruption height (cm, measured with meter stick). Controlled variables (any two): volume of vinegar (50 mL), concentration of vinegar (5%), flask size/shape (250 mL identical flasks), temperature (room temp).

2. A classmate writes this conclusion: 'My hypothesis was that hotter water dissolves sugar faster. My data showed the opposite. The experiment failed.' What is wrong with this conclusion, and what should they have written instead?

   short answerThe experiment did not fail — disconfirming data is a valid scientific outcome. A correct conclusion: 'The data did not support my hypothesis; sugar actually dissolved more slowly in hotter water under my conditions. This suggests either my hypothesis was incorrect or a confounding variable (e.g., evaporation, stirring rate) influenced the result. Next step: re-examine the procedure and retest.'

3. Which of the following is the BEST example of a specific source of error in a titration lab? A) Human error B) The experiment didn't work right C) The burette was read at eye level but the meniscus was estimated to ±0.05 mL, contributing uncertainty to the endpoint volume D) We didn't have enough time

   multiple choiceC — it names the specific instrument, the specific measurement, and quantifies the uncertainty. 'Human error,' 'didn't work,' and 'not enough time' are vague and untraceable.

4. A student claims their experiment is more valid than another group's because they took 200 data points instead of 20. Under what conditions is this claim FALSE? Give one specific example.

   short answerThe claim is false when the experimental design itself is flawed — more data points don't fix a confounded design. Example: if the student measured plant growth 200 times but never had a control group, or varied two factors at once (e.g., changed both light AND water), the extra n provides no additional validity. Design quality > sample size when the design is broken.

• ☐5 labeled unknown white powders A–E per group: NaCl, sucrose, NaHCO₃, cornstarch, CaCO₃ (~2 g each)
• ☐Spot plates (1 per group)
• ☐Distilled water in wash bottles
• ☐1.0 M HCl in dropper bottles
• ☐Iodine (Lugol's) solution in dropper bottles
• ☐pH paper, range 1–14
• ☐Conductivity testers or 9V battery + LED probe setup
• ☐Spatulas (5 per group, one per powder, to avoid cross-contamination)
• ☐Goggles, lab aprons, nitrile gloves for all students
• ☐Lab notebooks
• ☐Printed flawed-protocol handout (Jamie's caffeine experiment)
• ☐Printed variables-table template
• ☐Red pens
• ☐Stopwatch or wall clock visible to all groups