Data is just a bunch of anecdotes that we decided are similar enough to put in the same category.
Why do people insist on the opposite statement? “The plural of anecdote is not data.” Bah! That is the opposite of the truth. But I was given inspiration by a Jewish proverb, and I’ve figured out why people think this.
“For example” is not proof. -Jewish proverb
This is true. Anecdotes are not proof. By adding an assumption, we can see the logical progression of their thoughts.
Data is proof. (Assumption)
The plural of anecdote is not proof.
Therefore, the plural of anecdote is not data.
This is a valid syllogism, but it is not sound because the added assumption is not true. Yet we see people say this all of the time, “such and such has been proven by data” or “such and such has been proven by scientific research”. These statements make a fundamental category error.
Data cannot be used to “prove” things. They can only be used as evidence for belief. Statistics can only be used to make statistical statements, such as “99.9999999999% of the time if something leaves the ground, it returns to the ground.” This is the ur-experiment: throw a rock in the air one billion times, and it returns to the ground. But we cannot say it is therefore “proven” that everything must return to the ground. Something may achieve escape velocity and never return, although this is pretty rare.
The tacit assumption that data is proof betrays an ignorance of where data comes from. Perhaps people assume scientists receive their data from heaven as divine revelation? But no, they are actually just counting up carefully controlled anecdotes which they call “trials”.
Each of those times you throw the rock is an anecdote (or a “trial”). Taken together, they are data. The only remaining consideration is whether we think each anecdote was a “good” anecdote, worthy of inclusion in our “good” data. Maybe one time you throw the rock downward instead of upward, and I complain that this is a “bad” anecdote. Then I would not agree to include this trial in my dataset. Maybe you agree, maybe you disagree. Science cannot replace understanding or reason. It can only inform belief.