What a Fizz Can Certify: Response Ranges, Colour Steps and Lime Misclassification Rates for Vinegar, Baking-Soda and Plant-Juice Soil Tests

Three kitchen tests are widely sold and shared as ways to judge soil acidity and lime need: dropping vinegar on soil and watching for a fizz, mixing baking soda with soil and watching for a fizz, and reading the colour of a plant-juice indicator such as red cabbage juice or turmeric. This paper states the claims made for them plainly and computes, from stated reaction and calibration assumptions, the widest claim each test can honestly carry. For a 10 g sample, the vinegar test evolves 24.0 mL of carbon dioxide per percent of carbonate present, so a clear fizz certifies only that free carbonate at or above about 0.42 percent (calcium carbonate equivalent) is present; it responds to carbonate, not to pH, and used in the popular inverted form ("no fizz means acid, add lime") it wrongly flags 65.9 percent of soils that do not need lime against a pH 6.0 threshold, in the stated population scenario. The baking-soda test can evolve only small gas volumes from exchangeable acidity: under the stated buffer model a clear fizz requires pH below 4.23, a scored fizz corresponds to an effective cut-off of pH 5.61, and that cut-off wanders between pH 4.78 and 6.14 across the stated sensitivity range, so the test misses 25.7 percent of soils needing lime at a pH 6.0 threshold and 44.2 percent at a pH 6.5 threshold. Red cabbage juice resolves pH in steps of 1.0 pH unit; against a lime threshold of 6.0, which falls in the middle of one colour step, it misclassifies 12.5 percent of soils even with a perfect observer and 15.0 percent with a stated observer step error, while turmeric, a single transition at about pH 7.8, misclassifies 44.9 percent. Each test has an honest certification — carbonate present or absent above a stated content, strongly acid or not, pH within a stated one-unit band, pH above or below about 7.8 — and none of them is the certification printed on the packet, which is a lime decision. This study is simulated, not measured.