5

Statistically, how to describe a theory making from one's own experience and haven't been tested by others?

  • A sample of one?
  • A one-person sample?

Googling both phrases doesn't yield any meaningful result.

Example: Regardless of how much it helps me to be more stable, confident and sensitive than before, it's just _____________.

  • The definition of sample in experimentation involves the notion "representative subset" or "representative piece". Except for product samples (here, try this new shampoo) there is no experimental scenario where a single thing being tested would be called a sample unless it were being considered part of a larger set of things being tested, and in that case it would not be called "a sample of one", except sarcastically. – Tᴚoɯɐuo Jul 23 '18 at 11:00
  • And if it were a piece of something, like skin tissue, it wouldn't be called "a sample of one". – Tᴚoɯɐuo Jul 23 '18 at 13:15
  • 3
    The phrase a sample of one is perfectly idiomatic, but it's also very informal. You'd say this in conversation, especially if you're trying to emphasize that you're not trying to claim that this solution you're talking about is universal. You would not use it in any scholarly context when talking about actual statistics. – Canadian Yankee Jul 23 '18 at 17:16
  • @Tᴚoɯɐuo - Though in an English language context, you may well say "a sample size of one", intentionally, knowingly, in order to highlight that it isn't a very scientific test (e.g. in a novel or magazine article). – Nigel Touch Jul 23 '18 at 18:18
  • @Nigel Touch: a sample size and a sample are not at all the same. – Tᴚoɯɐuo Jul 23 '18 at 18:24
24

We use the expression "anecdotal evidence". This doesn't have to mean a sample size of one (but it could) it does mean that the sampling was informal, and the data was based on personal testimony.

Mary said that using baking soda had cleared up her acne, but that's just anecdotal evidence. We organised a clinical trial to test the effectiveness.

Regardless of how much it helps me to be more stable, confident and sensitive than before, it's just anecdotal evidence.

This carries the general sense that you want to convey, but not the particular point about a sample of exactly one. That fact could be implied by the context.

  • 5
    This Q&A reminds me of the aphorism “The plural of anecdote is not data.” – Matthew Leingang Jul 23 '18 at 14:21
7

If you want a non-pejorative term that is used in academia/science for this kind of evidence, you could use case study. It's a common feature of, for example, medical, psychology, and business research literature. A definition from Merriam-Webster:

an intensive analysis of an individual unit (such as a person or community) stressing developmental factors in relation to environment

The "individual unit" is your "sample of one".


Note that case studies can be either qualitative or quantitative, depending on the method of inquiry into the particular case, though they may be especially popular among qualitative researchers (and may be denigrated by some more quantitative-focused researchers).

6

In stats, the expression that I've often heard being used is 'A sample size of...'.

In this case 'A sample size of one' sounds better to me.

2

In statistics, sample size is represented by n.

In medicine some studies are so narrow the sample size can only be one. For example, some genetic abnormalities are so rare a patient may be the only test subject.

This is referred to as an n of 1 study.

More can be found on wikipedia.

  • If it is a study yes, if it is only from personal experience like in the question the accepted answer fits better – Gimli Jul 23 '18 at 13:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.