The Absence of Evidence

See Nonsense On Stilts.

The best way I can think of to describe the difference is the following:

If you wanted to determine whether someone was in a room, and someone poked their head through the doorway but kept their eyes closed, then reported that they didn’t see anybody, you wouldn’t take that as evidence that there was no one in there. That’s not evidence, that’s failing to check for evidence.

If someone poked in and looked around, and then said there was nobody there, that would be evidence that there wasn’t anyone there – in other words, evidence of absence. Or alternatively, evidence against presence.

The other way is the absence of evidence. It doesn’t mean anything at all.

If you haven’t made any observations, you have no positive evidence to support positions. That is not evidence against a particular position, just a reason not to assert one. To actively exclude a position, you need evidence against it.


9 Responses to “The Absence of Evidence”

  1. How about this scenario: I ask a number of people in a building if they have seen Mr. X. All respond no. It’s possible that none of them happened to go into room Z, where X actually is. But they also might not have seen him because he isn’t actually there. Absence of evidence is (weak) evidence of absence to the extent that we anticipated evidence of presence.

  2. From your perspective though, you would probably believe that P(he had his eyes open) > 0. In that case, him not seeing something would be evidence. evidence. I agree that absence of evidence is not necessarily evidence of absence, but it definitely can be, and often is.

  3. It’s evidence of absence if we would expect evidence in the event of the thing’s presence. It’s strong evidence, even.

  4. If lots of people report not seeing the guy, that’s not absence of evidence. That’s positive (although not very strong) evidence against his presence.

    If we then asked if anyone had looked in a certain room, and everyone said no, that wouldn’t be evidence he wasn’t there – UNLESS we presume he had to pass through rooms where he would be seen in order to get there.

    Of course, if you’re going to consider cases like teleportation or spontaneous vanishing, setting up logical arguments becomes quite difficult.

  5. If lots of people report not seeing the guy, that’s not absence of evidence. That’s positive (although not very strong) evidence against his presence.
    Then what counts as “absence of evidence”? If you look for evidence and don’t find it (it’s absent) then either that’s evidence of absence to some extent, or your search procedure has no value in detecting evidence. Simply having open eyes and ears can be considered a search procedure (though a quite weak one) and to the extent that you would have expected to be aware of evidence of fact F were it true, the fact that you are not yet aware can be considered (weak) evidence that it isn’t.

  6. Positive evidence that [Mr. X is not in the building] would be [seeing him outside the building]. Negative evidence for the claim that [Mr. X is not in the building] would be [not seeing him in the building] (because as P suggested, you would expect to see him if he were in fact in there).

    If you say positive evidence is NOT seeing something, then I think you are using the concept differently than how it is commonly used.

  7. Not seeing something isn’t evidence if you didn’t actually look.

    Think Trafalgar holding the telescope to his blind eye and saying “I see no ships.”

    It’s the truth, but it’s not evidence. (In the broader view, of course, his saying that is evidence of his intentions and the nature of his interactions. But in terms of whether there are ships there, it means nothing.)

    Only if, looking, nothing is seen, is that evidence. Just not very good evidence — it only acts against a very specific claim.

  8. Yeah, though sometimes negative evidence can sometimes be better than positive evidence.

    If you want to know more, I recommend a paper by Ulrike Hahn and Mike Oaksford. {Psychological Review, 2007, Vol. 114, No. 3, 704-732.} (what I’m talking about is on page 6/709) Unfortunately I don’t have permission to post it.

  9. Nick Tarleton Says:

    Suggestion: Stop using the unqualified phrase “absence of evidence”, since it sounds like the whole problem is confusion between absence-of-evidence-for and absence-of-evidence-either-way.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: