The Cafeteria Waste Audit
A team wants to measure how much food students throw away at lunch. They plan to weigh cafeteria waste for five school days. Constraints: one kitchen scale (accuracy ±10 grams), two team members available during a 25-minute lunch period, three lunch waves, and a custodian who empties bins right after each wave. Their first plan is to weigh the entire contents of each trash bin once at the end of the day. A teammate worries that this will not distinguish food waste from packaging, and that the custodian's schedule will interfere. They also debate whether to tell students they are being observed, since awareness might change behavior.
1. Identify the main threats to the validity of the team's original plan (weighing each bin once at day's end) and propose a revised procedure.
Read the worked answer
The original plan has at least four validity threats. (1) Construct validity: total bin weight includes packaging, trays, and non-food trash, so it does not measure food waste. (2) Timing: the custodian empties bins after each wave, so an end-of-day weighing may capture only the last wave or a mixture. (3) Aggregation: combining three waves loses the ability to compare waves or link waste to menu items. (4) Observer capacity: two people cannot sort and weigh three waves in 25-minute windows without a plan. A revised procedure: coordinate with the custodian to delay emptying or to provide labeled bins per wave; pre-sort food waste from packaging at the point of disposal using clearly marked bins; weigh each wave's food-only bin immediately after that wave; record weight, wave, date, and menu; and tare the scale before each weighing. With ±10 g accuracy on a kitchen scale, small differences will be noise, so the team should report totals per wave rather than claiming precision below about 50 g.
2. Should the team tell students they are being observed? Analyze the trade-off.
Read the worked answer
Telling students may create a reactivity effect: people modify behavior when they know they are watched, so waste might drop simply because of awareness, not because of any real change. Not telling them avoids reactivity but raises an ethical concern: observing people without their knowledge, even in a cafeteria, can be inappropriate, and the team may need permission from administrators. A middle path is usually best: inform students generally that a waste study is happening (satisfying transparency and consent norms) without singling out individuals or announcing exact weighing times. The team should then acknowledge reactivity as a limitation in its write-up, because even general awareness can shift behavior. If the goal is to compare waste before and after an intervention (like a signage campaign), the team should use the same level of announcement in both phases so the comparison is fair. The key judgment is that reactivity is a bias to be disclosed and controlled, not ignored.
3. The team's scale reads ±10 g and daily food waste totals are around 4 kg. A teammate wants to report waste "to the nearest gram." Evaluate this.
Read the worked answer
Reporting to the nearest gram overstates precision. If the scale is accurate to about ±10 g, then a reading of 4,000 g could plausibly be anywhere from roughly 3,990 to 4,010 g from instrument error alone, before counting sorting mistakes, spillage, and bin residue. Reporting "4,000 g" or "4.0 kg" is defensible; reporting "4,003 g" implies a level of certainty the instrument cannot support. The teammate may be confusing the display's resolution (how many digits it shows) with accuracy (how close the reading is to the true value). The team should also consider that sorting errors may dwarf the scale error: a single misplaced item could add tens of grams. Best practice: report weights rounded to a sensible unit (e.g., nearest 10 g or 0.1 kg), state the instrument's accuracy and the sorting protocol, and avoid implying that differences smaller than the measurement uncertainty are meaningful.