A few weeks ago we spent some time sitting around tasting different alkali metal salts with Adam Mastroianni. 

To us, the difference between NaCl (sodium chloride, also known as normal table salt) and KCl (potassium chloride) seems very obvious, but Adam said they taste about the same to him.

Since this can be tested empirically, we ran some quick studies to learn more. Inspired by the design of R. A. Fisher’s lady tasting tea experiment, we decided to test batches of 8 samples at a time in randomized, single-blind designs. 

Study 1

For the first study, eight cups were prepared, labeled A through H. We tested four samples of ¼ tsp dry NaCl and four samples of ¼ tsp dry KCl, randomly assigned to the eight cups. Testers were always blind to what salt was in what cup, but the experimenter was not blind.

Adam started by tasting the eight cups in order, and guessing which salt was in each cup. After tasting, he guessed that his accuracy was between 4/8 and 8/8. 

In fact, Adam’s accuracy was 6/8 — he incorrectly identified Cup D as KCl when it was actually NaCl, and he incorrectly identified Cup H as NaCl when it was actually KCl. Otherwise he correctly identified which salt was in which cup (Table 1).

Following this, the eight cups were emptied, cleaned, and re-stocked with four of each NaCl and KCl again, in a new random order.

Then, one of the SMTM authors (Raccoon #3) tasted the eight cups in order, guessing which salt was in each cup. After tasting, the author estimated their accuracy was either 7/8 or 8/8. 

In fact, Raccoon #3 got 8/8 correct, always identifying KCl as KCl and NaCl as NaCl (Table 2). If Fisher were there, he would have rejected the null hypothesis. 

Study 2

The second study was a replication of the first, except that the ¼ tsps of salt were each dissolved in one cup of lukewarm water, yielding 8 cups of two different salt solutions, four of each, in a random order, single-blind.

Again, Adam went first. Adam started by tasting the eight cups in order and guessing which salt was in each cup. After tasting, he estimated that his accuracy was between 4/8 and 7/8. 

In fact, Adam’s accuracy was once again 6/8 — he incorrectly identified Cup C as NaCl when it was actually KCl, and he incorrectly identified Cup F as KCl when it was actually NaCl. Otherwise he correctly identified which salt solution was in which cup (Table 3).

At this point one of our friends, who we will identify as RG, arrived at the apartment and also wanted to try the solutions. Since she was not present while Adam was tasting, we figured she could try his brines.

RG tasted the eight cups in order, guessing which salt was in each cup, and at the end estimated that her accuracy was 3 or 4 out of 8. In reality, her accuracy was 5/8. Like Adam, she misidentified Cups C and F, and she also misidentified Cup A as NaCl instead of KCl (Table 3).

Following this, the eight cups were emptied, cleaned, and re-stocked with NaCl and KCl solution, again four of each in a new random order.

Raccoon #3 tasted the eight cups in order. They guessed that the first three cups were NaCl, but when they reached the fourth cup, they commented that they must have been wrong, that Cup D was NaCl and the first three had to have been KCl. Based on this inference, at the end this author guessed an accuracy of 5/8. 

This one is hard to score. On the one hand, as written their accuracy was indeed 5/8. However, they realized their mistake on the first three cups as soon as they reached Cup D but before being unblinded, so you could also rate their accuracy as 8/8. See Table 4 for details.