This is a remarkable article. Is there a way to use this to overcome others' bias that passes ethical muster as not dark arts?

 

(HT +Tony Sidaway on G+.)

New Comment
13 comments, sorted by Click to highlight new comments since:
[-]Pfft100

I don't think the suggestions in the article seem all that Dark Arty. Presenting data as diagrams rather than text to make it easier to interpret is standard advise in scientific communication. In the example, the diagram contained much more information than the text, as well as being easier to process for humans.

And presenting evidence in a way which is not threatening to the listener's sense of identity is standard advice for all communication. (See in particular all the Less Wrong articles about atheism). If it turns out that asking someone about their personal strengths makes it easier for them to evaluate evidence neutrally, such a mind hack would be a very light art!

Two ideas are present here, one good, and one especially bad.

The good: You can be more open minded with self-affirmations before looking at politically controversial issues.

The bad: You can maneuver people into agreeing with you by having them go through self-affirmations and then presenting your argument. Note in the original article where they mention previous research doing this to "both" sides regarding the death penalty*. This is bad because it's a fully generalizable rhetorical technique that relies on filtering arguments (yes, a perfect reasoner would just adjust for the filtering, as the current sequence reruns emphasize, but if humans were perfect reasoners, the affirmations would not have any effect in the first place). The practicality of this attack is pretty questionable in most circumstances**, but it's something to watch out for.

* The first cite on the topic says that fabricated evidence was presented, but this doesn't seem to be the case for the second

** Beyond some obvious difficulties, and that this is a probably a marginal effect, they write "However, unlike previous studies, we find little evidence that affirmation increases the persuasive power of corrective information."

"Filtering" isn't a problem with an argument technique, it's a problem with argument. Every argument is a filtered argument, which means that anything that makes an argument more persuasive makes the filtering problem worse, whether it involves using self-affirmation or conveying ideas clearly. So the "bad" looks to be fully generalizable to all persuasive communication, making all arts dark.

I mostly agree but with severe caveats.

Piece of advise: If you find that one of your deeply held beliefs is being promoted through dark arts, you may want to have a crisis of faith to determine whether your own reasons for holding the belief are valid.

I think it's probably hard to find a belief with many adherents which nobody tries to promote using dark arts. If you find that one of your deeply held beliefs is more commonly promoted with dark arts than without, that should definitely raise alarms.

I don't see these as dark arts. Self-affirmation doesn't make people more likely to believe just any old thing, it makes them feel less threatened so they're more willing to consider something that they would otherwise reject out of defensiveness.

This doesn't seem to be dark arts to me. The use of graphs as a persuasive element seems to use the fact that graphs are visual and striking and so can overcome our cognitive defenses which prevent updating. This may connected to why graphs are such a useful mode of communication. Similarly, starting people off where they feel good about themselves may simply reduce the level which they feel a need to defend their beliefs as part of themselves.

So I don't see either of these as Dark Arts per se.

I am puzzled by the downvotes. I suspect that they may be coming from the title of the post which is unfortunate because people can read the link to see that the relevant methods worked on people across the political spectrum about issues they were otherwise unwilling to examine.

The use of graphs as a persuasive element seems to use the fact that graphs are visual and striking and so can overcome our cognitive defenses which prevent updating. This may connected to why graphs are such a useful mode of communication.

The impression I got from the description in the article is that without the graph subjects didn't believe that the data were accurate, with the graph they did. This, even though the existence of the graph doesn't provide any additional evidence about the accuracy of the underlying data.

If you'll indulge me some just-so-story type speculation, I suspect the reason striking visuals are more persuasive then words is that we evolved to expect others to manipulate us through language, we didn't evolve to expect to be manipulated through visuals.

Similarly, starting people off where they feel good about themselves may simply reduce the level which they feel a need to defend their beliefs as part of themselves.

Perhaps; on the other hand, this strikes me as a mild form of the infamous love-bombing technique.

If you'll indulge me some just-so-story type speculation, I suspect the reason striking visuals are more persuasive then words is that we evolved to expect others to manipulate us through language, we didn't evolve to expect to be manipulated through visuals.

That's a really interesting hypothesis. I would have guessed that it is just due to humans being very visual beings. I'd be really interested in seeing a way of testing your hypothesis. Another alternative hypothesis- people are more likely to dismiss simple assertions but graphs signal that someone has spent time and effort thinking about the issue in question. Moreover, knowing how to make a graph signals minimal intelligence so people are more likely to give credence?

Perhaps; on the other hand, this strikes me as a mild form of the infamous love-bombing technique.

That seems disconnected. Love-bombing involves making a set of emotional connections to the claims or people in question. This doesn't connect the emotion to the claims that are then evaluated later.

I would have guessed that it is just due to humans being very visual beings.

Well, we're also verbal beings, in fact most of our explicit rationality ability is verbal rather than visual.

I'd be really interested in seeing a way of testing your hypothesis.

I haven't thought of a good way to do this. However, one way to test your "it is just due to humans being very visual beings" is to take advantage of the fact that people differ greatly about how visual they are.

[-]Pfft30

I would have guessed that it is just due to humans being very visual beings.

The paper briefly considers whether the effect is due to graphs being easier to process, but apparently not (p.32):

These results do not appear to be driven by systematic differences how respondents processed Text or Graph – a post-treatment check of recall of a primary data source (NASA) found few significant differences between the treatments. Moreover, we observe no significant difference in the length of time respondents spent considering each treatment. These results suggest that the greater effectiveness of Graph is not simply a function of ease or depth of processing (results available upon request).

Another alternative hypothesis- people are more likely to dismiss simple assertions but graphs signal that someone has spent time and effort thinking about the issue in question.

This effect certainly exists, this is a special case of the reason the conjunction fallacy is frequently a good heuristic.